Consumer Energy, Etc.v.F. E. R. C

United States Court of Appeals, District of Columbia CircuitJan 29, 1982
673 F.2d 425 (D.C. Cir. 1982)

Nos. 80-2184, 80-2312.

Argued September 10, 1981.

Decided January 29, 1982.

Alan B. Morrison, Washington, D.C., with whom John Cary Sims, Washington, D.C., was on the brief for petitioners.

Larry L. Simms, Acting Asst. Atty. Gen., Dept. of Justice, of the bar of the Supreme Court of Vermont pro hac vice by special leave of Court, Washington, D.C., with whom Harold H. Bruff and Anthony J. Steinmeyer, Attys., Dept. of Justice, Washington, D.C., were on the brief for amicus curiae, U.S.

Jerome M. Feit, Deputy Sol., Federal Energy Regulatory Com'n, Washington, D.C., with whom Jerome Nelson, Acting Gen. Counsel, and Stephen R. Melton, Atty., Federal Energy Regulatory Com'n, Washington, D.C., were on the brief for respondent.

Edward J. Grenier, Jr., Washington, D.C., with whom Michael J. Shea, Glen S. Howard, Marilyn L. Muench and David A. Gross, Washington, D.C., were on the brief for intervenors, The Process Gas Consumers Group, et al.

C. William Cooper, Falmouth, Mass., with whom Richard M. Merriman, David G. Hanes, L. Peter Farkas, John D. McGrane, Peyton G. Bowman, III, and John R. Schaefgen, Jr., Washington, D.C., were on the brief for intervenors, United Distribution Companies.

Michael Davidson, Senate Legal Counsel, Washington, D.C., with whom M. Elizabeth Culbreth, Deputy Senate Legal Counsel, and Charles Tiefer, Asst. Senate Legal Counsel, Washington, D.C., were on the brief for amicus curiae, U.S. Senate, urging dismissal.

Eugene Gressman, Washington, D.C., for amicus curiae, Speaker of the U.S. House of Representatives, urging dismissal.

John A. Myler, Washington, D.C., was on the brief for intervenor, American Gas Ass'n.

Lawrence V. Robertson, Jr., Washington, D.C., was on the brief for intervenor, Interstate Natural Gas Ass'n of America.

R. Gordon Gooch, B. Donovan Picard, Thomas J. Eastment and John W. Leslie, Washington, D.C., were on the brief for intervenor, Petrochemical Energy Group.

Petitions for Review of an Order of the Federal Energy Regulatory Commission.

Before BAZELON, Senior Circuit Judge, and WILKEY and EDWARDS, Circuit Judges.

Opinion for the Court filed by Circuit Judge WILKEY.

WILKEY, Circuit Judge:

Petitioners Consumer Energy Council of America, Consumer Federation of America, and Public Citizen ("CECA") challenge the constitutionality of a legislative veto provision in the Natural Gas Policy Act of 1978 ("NGPA"). Title II of the NGPA directs the Federal Energy Regulatory Commission ("FERC") to implement an "incremental pricing" program, which shifts part of the price increase resulting from the deregulation of new natural gas from residential users to industrial users. Phase I of the program directs the Commission to promulgate within one year a rule applying only to "boiler fuel use of natural gas by any industrial boiler fuel facility." Phase II requires the Commission to issue within eighteen months a rule expanding the program to "any industrial facility which is within a category defined by the Commission" and not otherwise exempt. The statute provides, however, for the Phase II rule to take effect only if neither house of Congress adopts within thirty days a resolution disapproving the rule.

Pub.L. No. 95-621, 92 Stat. 3351 (codified at 15 U.S.C. §§ 3301-3342 (Supp. III 1979)).

15 U.S.C. § 3341(b) (Supp. III 1979).

Id. §§ 3342(a)(1), (b)(2).

Id. § 3342(c).

FERC issued its Phase II rule on 6 May 1980, three days before the deadline. Two weeks later the House of Representatives voted its disapproval. CECA then filed a petition for rehearing asking the Commission to make the rule effective in spite of the House's action, on the ground that the one-house veto provision was unconstitutional. Refusing to pass on the constitutional question, the Commission denied the petition and revoked the Phase II rule. CECA sought review in this court, and also petitioned FERC for rehearing on the revocation order. This second petition was denied, and CECA filed another petition in this court. The two petitions for review were consolidated and are now before us.

FERC has not taken a position on petitioners' constitutional claims, asserting that this court may dispose of the case on other grounds. It argues that the legislative veto provision is not severable from the provision authorizing the Commission to promulgate a Phase II rule, leaving petitioners with no claim for effective relief even if section 202(c) is declared unconstitutional. It also contends that the case is moot because the rule was properly revoked. The United States Senate and the Speaker of the House of Representatives have entered the case as amici curiae. In addition to supporting FERC's position, they contend that this court lacks jurisdiction to hear the case, a contention both FERC and CECA dispute, and that the legislative veto provision is constitutional. The United States has also appeared as amicus curiae, taking the opposing view that the veto provision is unconstitutional. Finally, several industry groups have filed briefs as intervenors, arguing generally in favor of the positions of both FERC and Congressional amici.

FERC's view is "that sound administrative practice requires the presumption of constitutional validity of the statutes entrusted to this Commission for implementation." FERC Order Denying Rehearing and Revoking Amendments Made by Order No. 80, at 3 (1 Aug. 1980) [hereinafter "Revocation Order"], reprinted in 45 Fed.Reg. 54741 (18 Aug. 1980), and Appendix A ("App. A") at 9.

Three sets of intervenors have appeared: (1) American Gas Association, Interstate Natural Gas Association of America, and United Distribution Companies ("Gas Suppliers"); (2) Process Gas Consumers Group, American Iron and Steel Institute, and Georgia Industrial Gas Group ("Gas Consumers"); and (3) Petrochemical Energy Group ("PEG"). Gas Suppliers address both the statutory and constitutional issues, whereas Gas Consumers and PEG address only the statutory issues.

Having determined that we have jurisdiction and that the constitutional issue is properly presented, we hold that the one-house legislative veto provision in section 202(c) of the NGPA is unconstitutional. Accordingly, we reverse and remand the Commission's orders and instruct it to reinstate the Phase II rule. The Commission is free, however, on its own motion and pursuant to the notice and comment requirements of the Administrative Procedure Act ("APA"), to consider amending the Phase II rule or changing its effective date.

5 U.S.C. §§ 553(b), (c) (1976).


A. Structure of the NGPA

Title I of the NGPA provides for the phased deregulation of natural gas prices. It sets allowable wellhead prices of gas considerably higher than under the costbased system of rate regulation, and provides that "new" natural gas prices will be decontrolled on 1 January 1985. Title II of the Act provides for incremental pricing as a means of easing the transition to deregulation. Incremental pricing is a "mechanism for passing through to end users some of the increased prices for natural gas," so that large industrial gas users will pay a disproportionate share of the increases in gas prices. The program's goals are "to restrain the prices paid by pipelines for natural gas supplies, particularly after deregulation, and to protect residential consumers from higher prices resulting from deregulation."

15 U.S.C. §§ 3311-3320 (Supp. III 1979) (wellhead price controls); id. § 3331 (price decontrol for new gas).

H.R. Rep. No. 1752, 95th Cong., 2d Sess. 68 (1978), U.S. Code Cong. Admin. News, pp. 8800, 8985 (Joint Explanatory Statement of the Committee on Conference) [hereinafter "Conference Report"].

Incremental pricing allocates a specified portion of the acquisition costs of natural gas to certain industrial users. Incremental pricing operates in this fashion: all the money that an interstate pipeline pays for different categories of natural gas in excess of certain specified "threshold" levels — which vary from category to category and increase from month to month — are added together and assigned to a special account. The amount of money in the account is recovered from the pipeline's distribution company customers by use of a surcharge mechanism according to the amount of gas the distribution companies sell to industrial customers subject to incremental pricing. The gas distribution company, in turn, is required to increase the rates to its industrial customers subject to incremental pricing, to recover its payment to the pipeline. The incremental pricing surcharge, however, may not increase the price of gas to industrial users above the alternative fuel price established by the Commission. If a gas distributor is unable to pass through his entire surcharge because his incrementally priced customers are already paying the alternative fuel price, the surcharge is reduced to that distributor and increased to the remaining distributors on the pipeline system whose industrial customers can absorb additional costs because they are paying less than the alternative fuel price. When all incrementally priced industrial facilities served by a pipeline reach equivalency with the alternative fuel price, amounts in excess of that necessary to keep those customers at equivalency may be allocated in whatever manner by which the pipeline and distribution companies are permitted to recover normal costs.

H.R. Rep. No. 938, 96th Cong., 2d Sess. 3-4 (1980) [hereinafter "Veto Report"].

Veto Report, supra note 9, at 4. The first goal was designed to facilitate an orderly transition from a regulated to a deregulated market. Congress feared that without a "market ordering" device such as incremental pricing, "the price of deregulated gas would be bid to excessively high levels" by pipelines which could continue to buy old gas at relatively low prices. Id. Accordingly,

Congress sought to eliminate this possible incentive for interstate pipelines to bid excessively high prices by initially increasing the price of gas delivered to large, price sensitive industrial customers to a level equal to the cost of their alternative fuel. Under this regime, Congress anticipated that pipelines, whose revenues depend upon volumetric throughput, would seek to minimize further increases in the cost of gas that might occur upon deregulation, for fear that industrial demand for natural gas (and therefore throughput revenues) would decline.

FERC Order No. 80, at 9 (6 May 1980) [hereinafter "Phase II Rule"], reprinted in 45 Fed.Reg. 31622, 31623 (13 May 1980), and Veto Report, supra, at 21.
The second goal was simply to shield high-priority users, such as residential gas consumers, from large and sudden price increases resulting from deregulation.

Title II provides for implementation of incremental pricing in two phases. Section 201 requires FERC within one year to issue an incremental pricing rule covering boiler fuel users. Section 202(a) provides that within eighteen months "the Commission shall, by rule, prescribe an amendment to the [Phase I] rule . . ." to extend incremental pricing to other industrial users. However, the Phase II rule "shall take effect only as provided under subsection (c) of this section," which states that the rule shall take effect after thirty legislative days "unless, during such 30 day period of continuous session of Congress, either House of the Congress adopts a resolution of disapproval." Thus, either house of Congress may veto FERC's Phase II rule by majority vote.

15 U.S.C. § 3341 (Supp. III 1979).

Id. § 3342(a)(1).

Id. § 3342(a)(2).

Id. § 3342(c)(1).

B. Legislative History of Title II

In April 1977 President Carter proposed to Congress a comprehensive energy program. In the congressional debate that followed a principal issue was whether to remove wellhead price controls on natural gas. There was widespread agreement that price controls should at least be loosened and that the dual gas market — in which the price of federally regulated interstate gas was lower than the price of intrastate gas, which was not subject to federal regulation — should be eliminated. But questions of specific tactics and timing were hotly disputed, with the primary debate centering on proposals for price deregulation. Two major concerns were the effect of deregulation on consumers and the problem of moving from a regulated to a deregulated market. Incremental pricing was proposed as a solution to both problems. President Carter's natural gas proposal and both the original bills passed in the House and the Senate contained provisions for incremental pricing.

In the House a special ad hoc energy committee reported out an energy bill rejecting natural gas deregulation and extending price regulation to intrastate gas, but raising the allowable price of gas by a substantial margin. The bill included a broad provision applying incremental pricing to interstate and intrastate pipelines and to local distribution companies. This meant that FERC would set both the price charged by pipelines to distribution companies and the price charged by distribution companies to local gas users. On 3 August 1977 the committee bill prevailed by a vote of 239-180. The bill did not provide for a legislative veto of FERC's incremental pricing regulations.

H.R. 8444, 95th Cong., 1st Sess. § 410, 123 Cong.Rec. 26169 (1977).

123 Cong.Rec. 27242 (1977).

In the Senate the President's bill was reported out of committee without recommendation. A substitute bill calling for deregulation ultimately prevailed on 4 October 1977 by a 50-46 vote. This bill applied incremental pricing only to charges by interstate pipelines to distribution companies, rather than also covering intrastate pipeline sales and sales from distribution companies to low-priority intrastate industrial users. Like the House bill, the Senate bill did not provide for legislative review of any incremental pricing rule.

Id. at 32306.

H.R. 5289, 95th Cong., 1st Sess. § 29, 123 Cong.Rec. 32308 (1977).

The two bills went to conference. It took the conference committee ten months to agree on a compromise bill calling for phased decontrol of natural gas pricing, along with a two-phase incremental pricing program which applied only to sales by interstate pipelines to distribution companies. The conference bill proved extremely controversial in both houses. After days of intense debate, the Senate passed the bill on 27 September 1978 by a vote of 57-42. The House ultimately considered all five energy bills, including the natural gas bill, in a single vote. The clear purpose of this unusual decision to consider five bills at once was to prevent certain defeat of the NGPA. The motion to suspend the rules to allow joint consideration was passed, after extremely bitter debate, on 13 October 1977 by a vote of 207-206. The House then approved the energy package on 14 October 1978 by a vote of 231-168. The bills were signed into law by the President on 9 November 1978.

124 Cong.Rec. 31847 (1978).

Id. at 36975.

Id. at 38503.

Remarks on Signing H.R. 4018, H.R. 5263, H.R. 5037, H.R. 5146, and H.R. 5289 into Law, 14 Weekly Comp.Pres.Doc. 1978 (9 Nov. 1978).

C. FERC's Incremental Pricing Rules

As required, the Commission promulgated Phase I regulations applying incremental pricing to large industries using natural gas as boiler fuel. The regulations were issued on 28 September 1979, and became effective 1 January 1980.

FERC Order No. 49, 44 Fed.Reg. 57726 (5 Oct. 1979) (codified in 18 C.F.R. §§ 154, 201, 204, 282 (1980)).

On 15 November 1979 the Commission issued a notice of proposed rulemaking providing for a broad Phase II rule that applied incremental pricing to all industrial users not specifically exempted by statute. On 6 May 1980, three days before the statutory deadline, the Commission adopted Phase II regulations based on its original November 1979 notice. During public hearings FERC had received many comments urging that incremental pricing not be expanded at all because market conditions had changed and because the incremental pricing concept itself was flawed. FERC rejected these suggestions:

Notice of Proposed Rulemaking and Public Hearings, FERC Docket No. RM 80-10, 44 Fed.Reg. 67170 (23 Nov. 1979).

The Commission believes that it was neither requested nor authorized to second-guess the social and economic judgments that the Congress made in enacting Title II. The role of the Commission under Section 202 is more limited. . . . It is up to the Congress to decide whether this Phase II submittal meets adequately the social and economic goals of the incremental pricing program or, indeed, whether those goals are still appropriate.

By virtue of the very review procedures built into section 202, it seems clear that the Congress sought to have this Commission develop a meaningful Phase II rule. The Congress would not have a meaningful choice if the Commission were to offer no rule, or a very narrow rule, for its review. . . .

The Commission believes that this Phase II rule presents a meaningful choice to the Congress.

Phase II Rule, supra note 10, at 5, reprinted in Veto Report, supra note 9, at 19.

D. The House Veto

The House Committee on Interstate and Foreign Commerce conducted hearings on the Phase II rule on 3 April and 6 May 1980. On 6 May, the same day FERC issued the final rule, the Subcommittee on Energy and Power reported favorably on a resolution of disapproval, and the next day the full committee did the same. The committee defined its task as determining "whether the risk of economic dislocation in requiring certain industrial customers at this time to shoulder increased gas costs is outweighed by the benefits in sheltering higher priority users from some cost increases and whether an expansion of incremental pricing at this time is consistent with current national priorities." In recommending disapproval of the rule, the committee emphasized that energy market conditions had changed drastically and in unanticipated ways since passage of the NGPA. Furthermore, uncertainty about the value of incremental pricing remained because there had not yet been sufficient time to evaluate the effects of Phase I. "Accordingly, the committee finds that current uncertainties with respect to phase II of incremental pricing must be substantially reduced before it would consider implementation of a phase II rule."

Veto Report, supra note 9, at 5.

Id. at 7.

Id. at 8.


Id. at 13-14. The committee also addressed the issue "whether section 204(e) of the NGPA requires that the Commission set a facility's maximum incremental pricing liability on the basis of its actual alternative fuel cost or whether, instead, section 204(e) merely gives the Commission a range of prices for establishing that facility's maximum incremental pricing liability." Id. at 10. The committee concluded: "The Commission's narrow interpretation of its authority under section 204(e) to set incremental pricing alternative fuel ceilings solely on the basis of installed alternative fuel capability and its perceived inability to tie delivered natural gas prices to natural gas wellhead market conditions appears to be a misreading of the statute." Id. at 11.

During the floor debate several reasons were given in support of the veto resolution. Most supporters declared that FERC had acted within its statutory mandate, but that changed circumstances and uncertainty about incremental pricing dictated that Congress drop the Phase II approach. Most congressmen appeared to agree with Representative Sharp's view that the congressional veto provision in section 202(c) was being used "not as a way to discipline [FERC], but as a way for us, as I think [was] intended by the conferees on the 1978 act, to have a second look at the policy that we could not be certain of[,] that we were not willing to mandate at that time." In addition, representatives who had opposed incremental pricing in 1978 argued for repeal of Phase I and declared their support for the veto because incremental pricing was unwise as a matter of policy. Other supporters of the resolution argued that FERC had violated the intent of Congress, by promulgating either too narrow a rule or too broad a rule. On 20 May 1980 the veto resolution passed the House by a vote of 369-34.

See, e.g., 126 Cong.Rec. H3839 (daily ed. 20 May 1980) (remarks of Rep. Sharp); id. at H3840 (remarks of Rep. Dingell); id. at H3843 (remarks of Rep. Preyer); id. at H3844 (remarks of Rep. Lundine); id. at H3850 (remarks of Rep. Kemp); id. at H3853 (remarks of Rep. Benjamin); id. at H3854 (remarks of Rep. Ashley). CECA and Congressional amici agree that the veto was based primarily on the determination that incremental pricing should not be expanded. See Brief for Congressional Amici at 47 n. 44; Reply Brief for Petitioners at 29.

126 Cong.Rec. H3841 (daily ed. 20 May 1980) (remarks of Rep. Sharp); see id. at H3849-50 (remarks of Rep. Levitas); id. at H3850 (remarks of Rep. Eckhardt).

"Every argument that has been made in committee, every argument that has been made on the floor against phase II is equally applicable to phase I." Id. at H3848 (remarks of Rep. Gramm); see id. at H3843 (remarks of Rep. Preyer); id. at H3845 (remarks of Rep. Brown); id. at H3851 (remarks of Rep. Kemp); id. at H3853 (remarks of Rep. Quayle).

See id. at H3847 (remarks of Rep. Markey). See also Veto Report, supra note 9, at 11 ("The Commission's narrow interpretation of its authority under section 204(e) . . . appears to be a misreading of the statute.").

See 126 Cong.Rec. H3849 (daily ed. 20 May 1980) (remarks of Rep. Stockman); id. at H3854 (remarks of Rep. Ashley).

H.R.Res. 655, 96th Cong., 2d Sess., 126 Cong.Rec. H3855 (daily ed. 20 May 1980).

E. Rehearing Proceedings before the Commission

On 5 June 1980 CECA petitioned for rehearing of the 6 May order, seeking elimination of the provision conditioning the rule's effectiveness on the failure of either house to pass a veto resolution within thirty days. The petition expressly challenged the constitutionality of section 202(c). On 1 August 1980 FERC denied the petition. It declined to rule on the constitutionality of the veto, finding "that sound administrative practice requires the presumption of constitutional validity of the statutes entrusted to this Commission for implementation."

Revocation Order, supra note 5, at 3, reprinted in App. A at 9.

The Commission then revoked the vetoed Phase II rule. It reasoned that if section 202(c) were declared unconstitutional the rule might take effect. This result was undesirable because "the Commission has not yet independently evaluated whether the Phase II rule meets the social and economic goals of the Title II incremental pricing program" and because "we might well have very serious reservations as to the wisdom of making the Phase II rule effective." Therefore, to ensure that it would have an opportunity to evaluate Phase I and Phase II if section 202(c) were held invalid, the Commission decided to "exercise its authority under sections 201 and 202 to amend the rule under section 201 to revoke the amendments made by [the Phase II order]."

Id. at 4-5, reprinted in App. A at 10-11.

Id. at 7, reprinted in App. A at 13.

Petitioners then sought rehearing of the revocation order. They challenged the Commission's authority to revoke the rule, and argued that the order was issued in violation of the APA's notice and comment requirements. The Commission denied the petition, finding that sections 201 and 202 of the NGPA permitted revocation by giving it power to amend the rule. It also found that the APA's requirements were met by the original notice of November 1979, and ruled in the alternative that there was "good cause" to ignore the APA in this case.

FERC Order Denying Rehearing on Revocation of Amendments in Order No. 80, at 2-3 (2 Oct. 1980) [hereinafter "Second Denial of Rehearing"], reprinted in 45 Fed.Reg. 71780 (30 Oct. 1980), and App. A at 22-23.

Petitioners sought judicial review of both the revocation order and the denial of the second rehearing petition, and this court consolidated the two petitions. We find that the revocation order was invalid, that section 202(c) is unconstitutional, and that the Phase II rule should become effective absent further Commission action to postpone or amend it. The Commission is free to conduct further proceedings not inconsistent with this opinion.


Only Congressional amici contest this court's jurisdiction. Their motion to dismiss, opposed by both petitioners and respondent FERC, asserts that this case should have been brought in the district court rather than the court of appeals. We hold that this case is properly in this court, and we deny the motion.

Section 506 of the NGPA gives this court jurisdiction to review FERC orders and rules issued under the Act. Congressional amici insist, however, that petitioners do not seek review of either an order or a rule:

15 U.S.C. § 3416 (Supp. III 1979).

They instead challenge FERC's acknowledgement that by operation of law its proposed rule is not effective. In essence, petitioners seek to compel FERC to adopt the incremental pricing rule which the House of Representatives has rejected. The decision of an agency not to adopt a rule can be challenged only by an action for mandamus and similar means in the district court, not this Court, and this petition for review should be dismissed.

Motion of Congressional Amici to Dismiss for Lack of Jurisdiction at 3.

Congressional amici also contend that this case belongs in the district court because that court is better equipped to compile a record and accord effective relief. We are unpersuaded by either of these arguments.

The suggestion that challenges to an agency's refusal to adopt rules or conduct a rulemaking may be brought only in a district court is erroneous. In WWHT, Inc. v. FCC we found that this court had jurisdiction to review an allegation "that the FCC abused its discretion when it denied [a] request for rulemaking." Thus even if we analogized CECA's complaint here to a challenge to an agency refusal to conduct a rulemaking, we still would have no reason for finding a lack of jurisdiction.

656 F.2d 807 (D.C. Cir. 1981).

Id. at 815-16. See also National Organization For Reform of Marijuana Laws v. Ingersoll, 497 F.2d 654, 656 n. 3 (D.C. Cir. 1974) (denying petition for mandamus, but ordering that petition be considered as one for review, where petitioners challenge agency decision not to conduct rulemaking); Gage v. United States Atomic Energy Comm'n, 479 F.2d 1214, 1222 n. 27 (D.C. Cir. 1973) (Commission denial of petition for institution of a rulemaking proceeding "would constitute a final order reviewable by this court"). The specific holding in WWHT was that this allegation of abuse of discretion was cognizable under the APA, 5 U.S.C. § 706(2) (1976). Review was appropriate in the court of appeals because Congress so provided by statute. See 47 U.S.C. § 402(a) (1976); 28 U.S.C. § 2342(1) (1976).
Whether review belongs in the district court or court of appeals, then, is determined by reference to the jurisdictional provisions in the agency's governing statute. This points up the flaw in Congressional amici's reliance on Environmental Defense Fund v. EPA, 598 F.2d 62 (D.C. Cir. 1978), which held that only the district court had jurisdiction over the contention "that EPA should have engaged in an additional rulemaking proceeding." Id. at 91. This was based on a finding that the jurisdictional provisions of the statute at issue, the Federal Water Pollution Control Act, 33 U.S.C. §§ 1365(a), 1369(b) (1976), did not provide for review in the court of appeals of this type of agency action. The case does not stand for a general proposition that all denials of rulemaking petitions must be challenged in the district court. (Congressional amici also cite Natural Res. Defense Council, Inc. v. SEC, 606 F.2d 1031 (D.C. Cir. 1979), in which it was noted that the court earlier had dismissed a petition for review of an agency's refusal to propose rules "on the ground that the Commission's action was not final agency action subject to judicial review." Id. at 1037 n. 2. The court's reasoning is not clear since the decision was issued without opinion, and in any event such unpublished orders may not be cited as precedent. D.C. Cir.R. 8(f).)

It is clear, moreover, that petitioners do indeed seek review of both a rule and an order of the Commission. CECA raises constitutional challenges to the Phase II rule and statutory challenges to the Commission's decision to revoke the rule. FERC issued a final rule, the effectiveness of which was conditioned on the failure of either house of Congress to veto it. CECA petitioned for removal of that provision, but FERC refused. The rule remained a final agency rule, which CECA challenged on constitutional grounds. FERC then revoked the rule, an action which was "final in every sense of the word."

Response of FERC in Opposition to Motion to Dismiss at 5 n. 1.

We also reject the suggestion that the district court is in any way a more appropriate forum in this case. The rulemaking here has been completed. No further compilation of a record is necessary; the issues raised are purely legal and have been fully briefed on all sides. This court is competent to grant whatever relief might be necessary. The simple fact that petitioners desire a different outcome does not turn their claim into one for mandamus.

In sum, there is no reason in law or policy for this case to be brought in the district court. To accept Congressional amici's interpretation would be to create a standard whereby review of rules would be heard in one court and review of revocation of rules in another, even though the statute, record, and issues would be virtually identical. We decline to reach such a result because it would violate the congressional intent expressed in section 506 that review of final Commission decisions be conducted in the courts of appeals.

Congressional amici cite two statutes which expressly provide that agency refusals to implement a rule because of a congressional veto must be challenged in the district court. See 15 U.S.C.A. § 57a-1(f) (West Supp. 1981) (FTC rulemaking); Federal Insecticide, Fungicide, and Rodenticide Act Amendments of 1980, Pub.L. No. 96-539, § 4, 94 Stat. 3195 (to be codified at 7 U.S.C. § 136w(a)(4)) (EPA pesticide rulemaking). These statutes are irrelevant here, however, since Congress did not authorize the district court to hear such issues in FERC cases. Indeed, by directing the district court to certify to the court of appeals all questions of the constitutionality of the legislative veto, these provisions suggest that Congress intends, even where it directs suits to be brought in the district court, to have a court of appeals, rather than a single district court judge, ruling on the constitutionality of the legislative veto.

In Chadha v. Immigration Naturalization Serv., 634 F.2d 408 (9th Cir. 1980), consideration of juris, postponed until hearing on merits, ___ U.S. ___, 102 S.Ct. 87, 70 L.Ed.2d 80 (1981), the Ninth Circuit held that it had jurisdiction to review the House of Representatives' disapproval of the suspension of an alien's deportation. This was based on 8 U.S.C. § 1105a(a) (1976), which provides that review in the court of appeals "shall be the sole and exclusive procedure for the judicial review of all final orders of deportation." Congressional amici direct our attention to a recent Third Circuit decision interpreting the same jurisdictional provision, though not involving the legislative veto. The court found that it lacked jurisdiction to hear claims brought by non-immigrant Iranian students found deportable by the Immigration and Naturalization Service ("INS"), challenging the constitutionality of the regulation which led to the discovery of their illegal status. Discussing the holding in Chadha, the Third Circuit declared:

A faithful application of Cheng Fan Kwock [ v. INS, 392 U.S. 206, 88 S.Ct. 1970, 20 L.Ed.2d 1037 (1968),] compels the conclusion that, because the constitutionality of either the one-house disapproval or the Iranian-related regulation could not have been tested during administrative deportation proceedings, neither matter should be reviewable directly by a court of appeals under section 106(a).

Dastmalchi v. Immigration Naturalization Serv.,
660 F.2d 880, 888 (3d Cir. 1981).
We find neither of these cases relevant to the jurisdictional question before us. There is a considerable body of law interpreting the specific judicial review provisions of the Immigration and Nationality Act, centering on the meaning of "all final orders of deportation." In contrast, this case presents us with the more general issue of appellate review of a final agency rule. In particular, we note that the specific ground of disagreement stated by the Third Circuit was that the constitutionality of the legislative veto and the INS regulation could not have been determined during the administrative deportation proceedings. In this case, FERC refused to address the constitutional issue because it did not want to take a partisan position, not because it lacked the power to do so. See note 5 supra. Congressional amici do not claim otherwise. Thus, we are presented with a case in which the agency refused on policy grounds to address the issue, and then, in recognition of the importance of the issue, went on to revoke the rule. For us to refuse to review the constitutional issue would be tantamount to holding that a court of appeals may never decide constitutional challenges to any agency rule.


FERC, joined by Congressional amici and intervenors, argues that this court should not reach the constitutional issues because even if we found the legislative veto provision unconstitutional, petitioners would not be entitled to effective relief. The argument is that section 202(c) is inseverable from section 202(a) because Congress would not have authorized Phase II incremental pricing had the legislative veto provision not been included. Thus, if subsection (c) is unconstitutional all of section 202 must be struck down. FERC would be left without authority to promulgate any Phase II rule, and petitioners would not be entitled to reinstatement of the rule issued by FERC in May 1980.

Petitioners deny that section 202(c) is inseverable from the rest of section 202, but they go on to argue that if section 202 must stand or fall as one, then section 202 is inseverable from the rest of Title II, which in turn is inseverable from the NGPA. Striking down the entire NGPA arguably would provide petitioners with effective relief because it would reinstate the pre-existing stringent natural gas price control scheme. FERC denies that section 202 is inseverable from the remainder of the NGPA, arguing that Phase I and Title I could go into effect without change. Gas Consumers, on the other hand, agree that section 202 is inseverable from section 201, but contend that Title I may stand alone. We do not need to reach these secondary arguments, however, because we find that subsection (c) is severable from the remainder of section 202.

The parties appear confused as to whether the severability issue presents a question of standing or of mootness. Compare, e.g., Brief for Gas Suppliers at 13 (inseverability means petitioners lack standing), with id. at 2 (inseverability means case is moot). Since we find that subsection (c) is severable, we need not decide this question. We also need not decide whether severability presents a threshold question that must be resolved prior to a determination on the merits. Compare McCorkle v. United States, 559 F.2d 1258, 1261-62 (4th Cir. 1977) (refusing to determine constitutionality of legislative veto provision in Federal Salary Act because provision is inseverable from remainder of Act), cert. denied, 434 U.S. 1011, 98 S.Ct. 724, 54 L.Ed.2d 755 (1978), with Atkins v. United States, 556 F.2d 1028, 1057-71 (Ct.Cl. 1977) (per curiam) (holding legislative veto provision in Federal Salary Act constitutional, without deciding severability issue), cert. denied, 434 U.S. 1009, 98 S.Ct. 718, 54 L.Ed.2d 751 (1978), and id. at 1082 n. 16 (Shelton, J., concurring in part and dissenting in part) ("It is, of course, unnecessary for the majority to reach the question of severability since it holds the one-House veto to be constitutional.").

The presence of a severability clause, which expressly sets forth congressional intent that a statute stand in the event one of its provisions is struck down, makes it extremely difficult for a party to demonstrate inseverability. When there is no such clause, however, as in this case, the test is less certain. Petitioners argue that severability is presumed and that the proponents of inseverability must prove otherwise. This standard is derived from Tilton v. Richardson, which stated that "`[t]he cardinal principle of statutory construction is to save and not to destroy,'" and from Buckley v. Valeo, which held: "`Unless it is evident that the Legislature would not have enacted those provisions which are within its power, independently of that which is not, the invalid part may be dropped if what is left is fully operative as a law.'" Congressional amici and intervenors, on the other hand, contend that petitioners are required to overcome a presumption that section 202(c) is inseverable from section 202 as a whole. They cite the Supreme Court's statement in Carter v. Carter Coal Co. :

See, e.g., Electric Bond Share Co. v. SEC, 303 U.S. 419, 434, 58 S.Ct. 678, 82 L.Ed. 936 (1938); Chadha v. Immigration Naturalization Serv., 634 F.2d 408, 415-16 n. 3 (9th Cir. 1980), consideration of juris. postponed until hearing on merits, ___ U.S. ___, 102 S.Ct. 87, 70 L.Ed.2d 80 (1981).

403 U.S. 672, 684, 91 S.Ct. 2091, 2098, 29 L.Ed.2d 790 (1971) (plurality opinion) (quoting NLRB v. Jones Laughlin Steel Corp., 301 U.S. 1, 30, 57 S.Ct. 615, 620, 81 L.Ed. 893 (1937)).

424 U.S. 1, 108, 96 S.Ct. 612, 677, 46 L.Ed.2d 659 (1976) (per curiam) (quoting Champlin Refining Co. v. Corporation Comm'n, 286 U.S. 210, 234, 52 S.Ct. 559, 564, 76 L.Ed.2d 1062 (1932)).

In the absence of [a severability] provision, the presumption is that the legislature intends an act to be effective as an entirety — that is to say, the rule is against mutilation of a statute; and if any provision be unconstitutional, the presumption is that the remaining provisions fall with it.

See also Electric Bond Share Co. v. SEC, 303 U.S. 419, 434, 58 S.Ct. 678, 683, 82 L.Ed. 936 (1938) (severability clause "reverses the presumption of inseparability").

We think the question where the presumption lies is mostly irrelevant, and serves only to obscure the crucial inquiry whether Congress would have enacted other portions of the statute in the absence of the invalidated provision. This is fully in accord with United States v. Jackson, in which the Supreme Court refused to place significance on the absence of a severability clause: "[W]hatever relevance such an explicit clause might have in creating a presumption of severability, . . . the ultimate determination of severability will rarely turn on the presence or absence of such a clause." Rather, the question is whether Congress would have enacted the remainder of the statute without the unconstitutional provision. We do not view the imposition of any unspecified burden of persuasion on either side as beneficial to the inquiry.

See, e.g., Buckley v. Valeo, 424 U.S. 1, 108, 96 S.Ct. 612, 677, 46 L.Ed.2d 659 (1976) (per curiam); United States v. Jackson, 390 U.S. 570, 585, 88 S.Ct. 1209, 1218, 20 L.Ed.2d 138 (1968). See also Atkins v. United States, 556 F.2d 1028, 1083-85 (Ct.Cl. 1977) (Shelton, J., concurring in part and dissenting in part) (although presumption of severability was mentioned in early cases, present law is clear that there is no presumption), cert. denied, 434 U.S. 1009, 98 S.Ct. 718, 54 L.Ed.2d 751 (1978).

We are presented with two remarkably different interpretations of the legislative history and intent underlying section 202. Petitioners contend that incremental pricing was critical to the compromise that enabled the NGPA to pass. They point out that the legislative review provision was added only in conference, that the provision was not emphasized in either the conference report or the subsequent floor debates, and that a major concern of many congressmen was that broad price protection be provided for consumers. Respondent and its supporters, on the other hand, claim that the legislative veto provision was central to the compromise that divided incremental pricing into two stages. They point to assurances made during the congressional debates that Phase II did not have to go into effect, and contend that without the veto provision Phase II incremental pricing would not have been authorized.

Though the legislative history, as almost always is the case, contains contradictory comments about the importance of section 202 and subsection (c), we find that Congress would have enacted section 202 in the absence of the legislative review provision in subsection (c). At the outset, we reject intervenors' view that the conference report in any way "emphasized" that section 202 would not have been enacted without subsection (c). The report simply describes the working of that section, and thus states that the Phase II rule will go into effect only if neither house disapproves. No one disputes that this was the intent of section 202 as enacted. The question before us, however, is what Congress would have intended in the absence of section 202(c). On this question the conference report is silent. To the extent the report does aid the inquiry, it supports petitioners' argument for severability, as its summary of Title II fails to mention, much less emphasize the importance of, the legislative review provision in section 202. Combined with the fact that none of the incremental pricing provisions in earlier bills were made subject to congressional review, this lack of emphasis militates against the view that the veto provision was essential to the incremental pricing and NGPA compromises.

Conference Report, supra note 9, at 95-97.

Id. at 68-69.

The floor debate on the conference committee compromise reinforces our finding of severability. It is true that there were statements on the floor of each house explaining that Phase II could be rejected by legislative veto. In response to criticisms of incremental pricing, Senator Jackson noted that "this bill does not compel incremental pricing with respect to any industrial use other than in large boilers. Extension of incremental pricing beyond boiler use could be prevented by a majority vote of either House of Congress 2 years from now." In the House, Representative Dingell noted that the intent was not to "drive industrial users off natural gas and onto other fuels," and went on to describe "several statutory guarantees against such an unintended result," one of which was that "[a]ny broadening of incremental pricing to a broader universe of industrial users is subject to congressional review and single House veto."

124 Cong.Rec. 31821 (1978) (remarks of Sen. Jackson).

Id. at 38362 (remarks of Rep. Dingell).

Considered in isolation, these two statements tend to support respondent's view that section 202(c) was essential to the statutory scheme. But other remarks by these same congressmen indicate their belief that Phase II was more than a mere possibility, it was a desirable policy. Senator Jackson's primary response to concerns about the potential ill-effects of incremental pricing was not to point out that a Phase II rule could be vetoed, but to argue that it would succeed because it would not cause gas prices to rise higher than alternative fuel prices. Representative Dingell made the same point, and then went on to argue that "a broadening of incremental pricing to other industrial users will give added assurances against forced conversions." He thus saw Phase II as important in ensuring the overall success of incremental pricing.

Id. at 31821 (remarks of Sen. Jackson); see id. at 31843 (remarks of Sen. Jackson).

Id. at 38362 (remarks of Rep. Dingell).

More important, these two statements are the sole references to subsection 202(c) during the days of debate which focused quite extensively on incremental pricing. These debates leave absolutely no doubt that the controversy over incremental pricing did not hinge on whether it would be limited to boiler uses or extended to other industrial uses. Opponents of the Act's incremental pricing provisions stressed primarily the regional discrimination that would result from applying it only to interstate pipelines, although some attacked the concept itself. And the most pervasive defense of Title II was that it would provide strong protection against large price increases for residential consumers. Nothing in these dozens of remarks remotely indicates that these proconsumer arguments would have changed into opposing arguments had Phase II not been made subject to legislative review.

"Incremental pricing is imposed upon industrial users only in the interstate market. Thus, gas will be cheaper for industries in the producing States, further encouraging the flight of industries from the Northeast and Midwest to the Sunbelt." Id. at 38480 (remarks of Rep. Markey); see, e.g., id., at 38499 (remarks of Rep. Holt); id. at 38497 (remarks of Rep. Pressler); id. at 38396 (remarks of Rep. Gore); id. at 38391 (remarks of Rep. Coughlin); id. at 38376 (remarks of Rep. Stockman); id. at 38372 (remarks of Rep. Edwards); id. at 38353 (remarks of Rep. Horton); id. at 38353 (remarks of Rep. Moffett); id. at 38353 (remarks of Rep. Anderson); id. at 38482 (remarks of Rep. Brown); id. at 38126 (remarks of Sen. Hollings); id. at 30025 (remarks of Sen. Humphrey); id. at 30024 (remarks of Sen. Metzenbaum); id. at 30022 (remarks of Sen. Kennedy); id. at 30016 (remarks of Sen. Abourezk); id. at 28888 (remarks of Sen. Wallop); id. at 28876 (remarks of Sen. Weicker); id. at 28644 (remarks of Sen. Hansen).

See, e.g., id. at 38376 (remarks of Rep. Stockman); id. at 38372 (remarks of Rep. Edwards); id. at 28644 (remarks of Sen. Hansen).

"[This bill] reverses the historical unfairness in pricing between residential and industrial users. . . . Under the report, extra costs would be passed on to the industrial users." Id. at 31844 (remarks of Sen. Glenn); see, e.g., id. at 38498 (remarks of Rep. Wolff); id. at 38381 (remarks of Rep. Glickman); id. at 38361 (remarks of Rep. Dingell); id. at 31844-46 (remarks of Sen. Jackson); id. at 31840-41 (remarks of Sen. Muskie); id. at 31838 (remarks of Sen. Moynihan); id. at 31835 (remarks of Sen. DeConcini); id. at 31817 (remarks of Sen. Bumpers); id. at 30026 (remarks of Sen. Chiles); id. at 30024 (remarks of Sen. Heinz); id. at 28665 (remarks of Sen. Pearson).

To the contrary, it is noteworthy that Phase I was estimated to cover only 10% of interstate gas deliveries. See id. at 38362 (remarks of Rep. Dingell). Although the market ordering purpose conceivably might be served by Phase I alone, see id., broad consumer protection could not be provided by this limited coverage. Indeed, FERC subsequently found that the actual coverage of Phase I was only 7-8% of annual interstate volumes, see Phase II Rule, supra note 10, at 16, reprinted in Veto Report, supra note 9, at 25, and it concluded that "the objective of price shielding can be best advanced by a Phase II rule covering as broad a class of non-exempt users as possible." Id. at 70, reprinted in Veto Report, supra, at 49.

We are convinced, then, that Congress would have enacted both phases of incremental pricing without the veto provision. Title II of the NGPA set forth a congressional policy in favor of incremental pricing. Although the exact contours were left in doubt, the Congress made, through its delegation to FERC, at least a tentative decision in favor of extension of incremental pricing to nonboiler industrial uses. By finding the veto provision severable we do not destroy the distinction Congress made between Phase I and Phase II. The scope of Phase I was unambiguous, and the Commission was left with mostly technical discretion to formulate a workable rule. Phase II, however, was undefined; the Commission was directed only to come up with some extension of incremental pricing. The concerns about extending incremental pricing, especially the possibility that gas users would be forced to switch to other fuels, were left open for the Commission to consider. To some extent the Commission did exercise such discretion, as it set a moderate alternative fuel price level so that its extension of incremental pricing would not place great burdens on the uses to be covered by Phase II. Moreover, the Commission's stated reason for not considering a more limited Phase II rule was that it thought the veto provision left that determination to Congress. If the legislative veto had not been enacted, the scope of the Phase II rule would have been left to the Commission's discretion, which would include consideration of the effect of changed market conditions on the usefulness of incremental pricing. The distinction between Phase I and Phase II thus would have been preserved, and the same basic policy underlying Title II would have been enacted. The veto provision was not essential to the statutory policy, and it thus is severable.

The nature of the House's consideration of the disapproval resolution also indicates a recognition that incremental pricing generally was critical. The House Report recommending disapproval noted that "the incremental pricing concept was perceived to be a necessary prerequisite to increased wellhead prices for natural gas." Veto Report, supra note 9, at 6. And Representative Dingell began the hearings on the rule by stating: "Incremental pricing is one of the very important elements of the compromise that was forged during the 95th Congress which resulted in the enactment of the Natural Gas Policy Act." Phase II Incremental Pricing of Natural Gas: Hearings on H.R.Res. 655 Before the Subcomm. on Energy and Power of the House Comm. on Interstate and Foreign Commerce, 96th Cong., 2d Sess. 1 (3 Apr. 6 May 1980) [hereinafter "Phase II Hearings"] (statement of Rep. Dingell). During the hearings and debates, almost everyone indicated a belief that by considering a veto of the rule the House was engaging in a policy reassessment. The discussion focused almost exclusively on whether incremental pricing remained appropriate policy, especially in light of new market conditions. Representative Gramm noted that "every point for vetoing phase II is an equally good point for vetoing phase I." Id. at 381 (statement of Rep. Gramm). See generally pp. 437-438 supra.

See, e.g., 126 Cong Rec. H3854 (daily ed. 20 May 1980) (remarks of Rep. Ashley):

In my role as a conferee on the NGPA, I had occasion to engage in a colloquy with my colleague John Dingell during floor consideration of the bill. That colloquy made clear the conferees' understanding that a significant delegation of authority was being made to FERC under NGPA, with respect to phase II of incremental pricing. But the remarks also made very clear the intent of Congress that FERC exercise this authority cautiously, and that the Commission carefully consider the economic impact of incremental pricing in all industrial users (especially industrial process users) before promulgating a final rule.

See Phase II Rule, supra note 10, at 70-71, reprinted in Veto Report, supra note 9, at 49.

See id. at 5, reprinted in Veto Report, supra note 9, at 19; Phase II Hearings, supra note 66, at 383 ("I believe that the Congress cannot have it both ways. It cannot reserve [the judgment on the propriety of Phase II incremental pricing] to itself through the veto authority and then assert that it should have been exercised independently by the agency, free from congressional review or, more importantly, responsibility") (statement of FERC Chairman Curtis).

In McCorkle v. United States, 559 F.2d 1258 (4th Cir. 1977), cert. denied, 434 U.S. 1011, 98 S.Ct. 724, 54 L.Ed.2d 755 (1978), the Fourth Circuit refused to decide a constitutional challenge to the legislative veto provision in the Federal Salary Act on the ground that the provision was inseverable from the grant of power enabling the President to propose salary increases. The Court found:

Voiding the one-house veto as unconstitutional while leaving presidential authority intact would increase the President's power over salaries far beyond the intention of Congress. We are satisfied that the legislative history establishes that Congress would not have delegated authority to the President to establish salaries without the provision for the one-house veto.

at 1262. We think this holding is easily distinguishable from our finding of severability in this case. The Fourth Circuit found that the legislative history was clear that Congress would not have given the President blanket authority to set federal salaries without imposing a check. Here, in contrast, the delegation of rulemaking authority to FERC was entirely routine: it was required to issue a rule on a specifically defined subject, and to do so pursuant to normal administrative procedures. Such delegations are commonly made without any provision for legislative review, and we see nothing in the legislative history to indicate that Congress would have done differently in this case.
also stated that "the case against severance is strong" where the provision restricts a grant of power: "Otherwise, the scope of the power would be enlarged beyond the legislature's intent." Id. at 1261. We decline to adopt this as a general principle that would make all veto provisions prima facie inseverable. We think this statement does no more than restate the basic test that a court should determine whether the provision was so essential to the legislative purpose that the statute would not have been enacted without it. In McCorkle the restriction seemed essential to the Fourth Circuit, as the grant of power otherwise would have been much larger than Congress would have contemplated; in Chadha, on the other hand, the veto provision, although a restriction on a grant of power, was deemed not essential and therefore severable. See Chadha v. Immigration Naturalization Serv., 634 F.2d 408, 417 n. 5 (9th Cir. 1980), consideration of juris. postponed until hearing on merits, ___ U.S. ___, 102 S.Ct. 87, 70 L.Ed.2d 80 (1981).


FERC contends that its revocation order renders this case moot because there no longer is a rule for petitioners to challenge. Petitioners make two responses. First, they assert that FERC was not authorized to revoke the rule. Second, they contend that even if FERC had power to revoke the rule, the revocation order was invalid because FERC did not comply with the notice and comment requirements of the APA. FERC replies that it has revocation authority under sections 201(a) and 202 of the NGPA, that the original Phase II notice and opportunity to comment encompassed this revocation order as well, and that in any event this order fell within the "good cause" exception to the notice and comment requirements. We find that FERC's revocation order is invalid because it was issued in violation of the APA, and thus do not reach the issue whether FERC was authorized to revoke the rule.

FERC makes two arguments why it was not required to provide notice and comment prior to revoking the Phase II rule. First, it argues that the November 1979 notice and opportunity to comment provided prior to promulgation of the Phase II rule was sufficient to cover the order of revocation. FERC provides no support for this contention, but intervenors expand on it by asserting that petitioners had ample notice originally that one option to be considered by the Commission was to adopt no rule at all, and that this is all the APA requires. We disagree. Sections 553(b) and (c) set forth notice and comment requirements for "rule making," which is defined in section 551(5) to mean "agency process for formulating, amending, or repealing a rule." Thus, the APA expressly contemplates that notice and an opportunity to comment will be provided prior to agency decisions to repeal a rule. If the notice and comment provided prior to a rule's promulgation were meant to be sufficient to encompass any later repeal of the rule, simply because there was always a possibility that no rule would be adopted, the statute never would have included repeal of a rule within the definition of rulemaking.

Intervenors raise an additional ground not relied on by FERC. They assert that § 506 of the NGPA provides specific authorization for FERC to repeal its rule without any further notice or comment. Section 506(a)(3) provides: "At any time before the filing of the record of a proceeding in a United States Court of Appeals . . . the Commission may, after providing notice it determines reasonable and proper, modify or set aside, in whole or in part, any order issued under the provisions of this chapter." 15 U.S.C. § 3416(a)(3) (Supp. III 1979). Intervenors contend that this provision permitted the revocation order, since the Commission implicitly found that the "reasonable and proper" notice would in this case be no notice. They also find authority for the Commission's action in § 506(a)(2), which provides that upon the filing of a petition for rehearing, the Commission may "grant or deny the requested rehearing or modify the original order without further hearing." Id. § 3416(a)(2). Gas Consumers contend that since "the revocation of the proposed Phase II regulations was effected by FERC as the direct and acknowledged result of Petitioners' application for rehearing," FERC was authorized to revoke the rule without providing notice and comment. Brief for Gas Consumers at 47.
We note that FERC has not relied on these provisions as authority for their revocation order, either in response to the two petitions for rehearing or in argument to this court. This provides a strong indication that the agency itself does not view § 506 as providing authority summarily to revoke a disapproved final rule. More important, we are convinced that Congress did not intend § 506(a) to permit the Commission to use the occasion of a petition for rehearing to make any substantive change whatsoever in a rule, without providing some notice and comment. To so interpret the provision would effectively eviscerate the procedural protections provided by the APA, a result clearly not envisioned by the NGPA. We therefore decline to adopt intervenors' argument that the revocation order as carried out was authorized by § 506.
We also note, without deciding, the issue whether § 506(a) applies at all to rulemaking, since § 506(a) deals only with orders while § 506(b) deals with rules. But cf. ECEE, Inc. v. FERC, 611 F.2d 554, 559-66 (5th Cir. 1980) (holding that rehearing requirement in § 506(a) applies to rulemaking orders issued under § 506(b)).

5 U.S.C. §§ 553(b), (c) (1976).

Id. § 551(5) (emphasis added).

Intervenors, though not FERC, argue that notice and comment were not required because FERC did not "repeal a rule," but rather merely revoked "proposed regulations." They claim that § 553(b) applies only to final rules that are legally enforceable. See Brief for Gas Consumers at 52. We reject the view that FERC's Phase II rule was nothing more than a proposed regulation, see pp. 465-466 infra, but note that in any event the 6 May order was a final rule. The rulemaking process was at that point complete. The fact that the House of Representatives voted its disapproval did not make the rule any less final. The Commission's action in revoking the final rule, therefore, fit squarely within the category of actions subject to the APA's notice and comment requirements.

The value of notice and comment prior to repeal of a final rule is that it ensures that an agency will not undo all that it accomplished through its rulemaking without giving all parties an opportunity to comment on the wisdom of repeal. Such an opportunity was lacking here. The Commission consistently stated that it had no choice but to issue a broad Phase II rule. The Commission issued a final rule, and petitioners sought to amend the provision making the rule's effectiveness contingent on legislative review. Although the petition for rehearing said nothing about repealing the rule, and no other party requested such action, the Commission went ahead and rescinded it. The specific concerns that motivated this decision — the constitutionality of the legislative veto and the results that might follow from a judicial decision to strike down the veto — were different from those raised during the original rulemaking. All of these factors demonstrate that notice and comment would have been useful prior to repeal, and thus buttress further our conclusion that the Commission was required to follow section 553.

See Notice of Proposed Rulemaking, supra note 24, 44 Fed.Reg. at 67170-71; Phase II Rule, supra note 10, at 5, reprinted in Veto Report, supra note 9, at 19.

Spartan Radiocasting Co. v. FCC, 619 F.2d 314 (4th Cir. 1980), cited by intervenors for the proposition that an agency may modify a previous decision in light of a petition for rehearing without providing further notice or opportunity to comment, is inapposite. In modifying a previous order, the FCC adopted a position which was raised by the rehearing petition, which had long been argued before the FCC, and which had been brought up specifically in the preceding rulemaking proceeding; indeed, the Commission noted that there had been a full opportunity to debate this particular issue "in five separate pleading cycles." Id. at 321. In contrast, we are presented here with a situation in which the Commission made it known from the outset that it construed its mandate to preclude the possibility of issuing no rule or postponing the issuance of a rule. Moreover, the court in Spartan Radiocasting distinguished situations in which notice and comment were necessary to prevent unfairness, observing that "an agency decision which is made in reconsideration of a final order has been overturned for failure to give published notice of the fact of its renewed deliberations. Arlington Oil Mills, Inc. v. Knebel, 543 F.2d 1092, 1098-99 (5th Cir. 1976)." Id.
In Arlington Oil Mills the Department of Agriculture announced a decision on peanut support prices on 19 March, then revoked that decision on 6 July and reinstated the preexisting price levels. The Department provided no notice to the public that it was considering rescission of the 19 March order, nor did it solicit comments or conduct hearings. 543 F.2d at 1097. The court of appeals held that "the Secretary was required to provide adequate notice that reconsideration was underway and to give all interested persons a reasonable opportunity to participate and present their views before he arrived at the July 6 announcement." Id. at 1099. This is similar to what went on here. CECA's petition raised only the prospect of reconsideration of the provision making the rule's effectiveness contingent on legislative review. No party had any inkling that the Commission was considering a complete repeal of the rule. This is not a case in which "[a]ny surprise worked upon petitioner arises from the success of the opponent's arguments, rather than from any claimed failure of notice about their consideration." Spartan Radiocasting, 619 F.2d at 322. The Commission's decision clearly caught petitioners, indeed all parties, by surprise.

The Commission's second argument is that additional notice and comment, even if normally required, were not required here because the Commission found, in accordance with the exception contained in section 553(b)(B), that they were "unnecessary" in this case. FERC contends that it acted "to insure that defectively promulgated regulations will not become effective," citing its statements in the revocation order that it had not engaged in an independent determination of the validity of the policies embodied in the Phase II rule. This contention is easily rejected. Section 553(b)'s "good cause" exception requires the agency to state the reasons for finding "good cause" in its decision. Here, however, FERC merely asserted that further notice and opportunity to comment were unnecessary; only later did it argue that they were unnecessary because the regulations were defective.

5 U.S.C. § 553(b)(B) (1976) provides that the requirements of § 553 do not apply "when the agency for good cause finds (and incorporates the finding and a brief statement of reasons therefor in the rules issued) that notice and public procedure thereon are impracticable, unnecessary, or contrary to the public interest."

Brief for Respondent at 16.

The Commission's entire statement regarding its failure to provide notice and comment is as follows:

Nor do we believe that the notice and comment procedures of 5 U.S.C. § 553 require another result. Because the Commission provided full notice and opportunity to comment prior to issuing Order No. 80, further comment is unnecessary. Notice of Proposed Rulemaking, 44 Fed.Reg. 67170 (November 23, 1979). However, to the extent that 5 U.S.C. § 553 would otherwise require it, we find here that for good cause, further notice and public procedure are "unnecessary" in accordance with 5 U.S.C. § 553(b).

Further, regardless of the application of those procedures to regulations which become operative, they cannot be required where regulations are revoked because they have not been fully considered.

Second Denial of Rehearing, supra note 40, at 3, reprinted in App. A. at 23. It is clear that the Commission considered the question of defective promulgation to be different from the question of "good cause."
The Commission's argument that notice and comment requirements do not apply to "defectively promulgated regulations" is untenable because it would permit an agency to circumvent the requirements of § 553 merely by confessing that the regulations were defective in some respect and asserting that modification or repeal without notice and comment was necessary to correct the situation. Such a holding would ignore the fact that the question whether the regulations are indeed defective is one worthy of notice and an opportunity to comment.

Furthermore, the argument that repeal was required because the regulations were defective does not explain why notice and comment could not be provided. This court has held that "use of these exceptions by administrative agencies should be limited to emergency situations." The fact that FERC considered these regulations defective did not imply that an emergency existed. Were a court to strike down the legislative veto provision in section 202(c), the Commission would be able to assert the same modification and revocation authority it asserted in this case, and thus be able to consider altering the rule. Since no emergency existed, the Commission was not entitled to ignore section 553(b).

American Fed'n of Gov't Employees, AFLCIO v. Block, 655 F.2d 1153, 1156 (D.C. Cir. 1981).

Indeed, given that the Commission never explained why it could not postpone consideration of revocation until after a judicial ruling on the one-house veto, the only result of acceptance of the Commission's argument would be to prevent judicial review entirely. FERC justified its revocation on the need to assure that it had "an opportunity to evaluate Phase I, and the impacts of Phase II as proposed[,] in the event section 202(c) is held invalid." Revocation Order, supra note 5, at 7, reprinted in App. A at 13. Before this court, however, FERC argues not only that revocation was justified because of the possibility of a judicial ruling on the constitutionality of section 202(c), but that the revocation decision itself prevents this court from reaching the constitutional issue at all.


Petitioners raise four constitutional challenges to the legislative review provision in section 202(c). They claim that the veto provision violates the constitutional doctrine of separation of powers; that it deprives the President of his veto power under Article I, Section 7 of the Constitution (the Presentment Clause); that it violates Article I, Section 7's requirement of bicameralism; and that it delegates legislative authority without standards or provision for judicial review. FERC, Gas Consumers, and PEG have taken no position on these claims. Congressional amici and Gas Suppliers, however, contest each of these arguments, and assert generally that Congress is authorized to use the one-house veto device under the Necessary and Proper Clause of the Constitution.

We hold that section 202(c) is unconstitutional. The primary basis of this holding is that the one-house veto violates Article I, Section 7, both by preventing the President from exercising his veto power and by permitting legislative action by only one house of Congress. In addition, we find that the one-house veto contravenes the separation of powers principle implicit in Articles I, II, and III because it authorizes the legislature to share powers properly exercised by the other two branches. Because we find these bases sufficient to resolve the issue, we do not reach the undue delegation of powers issue raised by petitioners.

The doctrine of undue delegation prohibits standardless delegation of powers from Congress to the Executive or administrative agencies. See, e.g., A. L. A. Schechter Poultry Corp. v. United States, 295 U.S. 495, 55 S.Ct. 837, 79 L.Ed. 1570 (1935); Panama Refining Co. v. Ryan, 293 U.S. 388, 55 S.Ct. 241, 79 L.Ed. 446 (1935). Although the doctrine has seldom been used, and indeed has often been declared deceased, there is evidence that the Supreme Court has not written it off. See, e.g., National Cable Television Ass'n, Inc. v. United States, 415 U.S. 336, 342, 94 S.Ct. 1146, 1149, 39 L.Ed.2d 370 (1974) (construing statute narrowly to avoid delegation doctrine problems); Industrial Union Dept. AFL-CIO v. American Petroleum Inst., 448 U.S. 607, 671-88, 100 S.Ct. 2844, 2878-2887, 65 L.Ed.2d 1010 (1980) (Rehnquist, J., concurring in judgment); California Bankers Ass'n v. Schultz, 416 U.S. 21, 90-91, 94 S.Ct. 1494, 1531-1532, 39 L.Ed.2d 812 (1974) (Douglas, J., dissenting); id. at 91-93, 94 S.Ct. at 1532-1533 (Brennan, J., dissenting). Moreover, a growing number of other judges and commentators have urged revival of the doctrine. See, e.g., J. Ely, Democracy and Distrust 131-34 (1980); T. Lowi, The End of Liberalism 146, 154-55, 297-99 (1969); Gewirtz, The Courts, Congress, and Executive Policy-Making, Law Contemp. Prob., Summer 1976, at 46, 49-65; McGowan, Congress, Court, and Control of Delegated Power, 77 Colum.L.Rev. 1119, 1127-30 (1977); Wright, Book Review, 81 Yale L.J. 575, 582-87 (1972). Although we recognize the importance of the concerns about broad delegations, see note 172 infra, we note that acceptance of petitioners' argument would require us not only to pronounce a revival of the delegation doctrine, but also to extend the doctrine beyond the administrative setting to encompass congressional delegations of power to parts of Congress itself. We do not suggest that the breadth of such a holding necessarily demonstrates its implausibility. See, e.g., Dixon, The Congressional Veto and Separation of Powers: The Executive on a Leash?, 56 N.C.L. Rev. 423, 448-51 (1978) (recognizing plausibility of delegation doctrine argument as applied to legislative veto); McGowan, supra, 1159-60 (same). But given the sufficiency of the other constitutional claims, we believe consideration of this possibility is neither necessary nor appropriate in this case, especially in light of the short shrift given the argument by petitioners themselves.

A. Federal Rulings on the Legislative Veto

Although Congress and the Executive have been at odds for decades over the constitutionality of the legislative veto, the Supreme Court has never ruled on the issue. Three federal courts presented with the issue have disposed of the cases without addressing the merits. Two other federal courts, however, have ruled on the constitutionality of two different legislative review provisions, one provision being struck down and the other being upheld.

One Supreme Court Justice has expressed the view that the legislative veto is constitutional. In a separate opinion in Buckley v. Valeo, 424 U.S. 1, 257, 96 S.Ct. 612, 744, 46 L.Ed.2d 659 (1976) (White, J., concurring in part and dissenting in part), Justice White briefly addressed the constitutionality of a one-house veto provision in the Federal Election Campaign Act of 1971, 2 U.S.C. § 438(c) (1976). First, he likened a one-house veto to disapproval of proposed legislation, meaning that the requirements of Article I, Section 7 do not apply. 424 U.S. at 284-85, 96 S.Ct. at 757-58. Moreover, he argued that the principal purpose of the President's veto power, to provide the Executive Branch "with some bargaining and survival power against what the Framers feared would be the overweening power of legislators," id. at 285, 96 S.Ct. at 757, is not implicated by review of agency rulemaking, which he considered legislative in nature. He concluded:

I would be much more concerned if Congress purported to usurp the functions of law enforcement, to control the outcome of particular adjudications, or to pre-empt the President's appointment power; but in the light of history and modern reality, the provision for congressional disapproval of agency regulations does not appear to transgress the constitutional design, at least where the President has agreed to legislation establishing the disapproval procedure or the legislation has been passed over his veto.

at 285-86, 96 S.Ct. at 757-58.
The majority in Buckley did not address this issue, disposing of the case on other grounds. Id. at 140 n. 176, 96 S.Ct. at 692 n. 176 (per curiam).

See McCorkle v. United States, 559 F.2d 1258 (4th Cir. 1977) (severability), cert. denied, 434 U.S. 1011, 98 S.Ct. 724, 54 L.Ed.2d 755 (1978); Clark v. Valeo, 559 F.2d 642 (D.C. Cir.) (en banc) (per curiam) (ripeness), aff'd mem. sub nom. Clark v. Kimmitt, 431 U.S. 950, 97 S.Ct. 2667, 53 L.Ed.2d 267 (1977); Citronelle-Mobile Gathering, Inc. v. Gulf Oil Corp., 420 F.Supp. 162 (S.D.Ala. 1976) (standing), remanded, 578 F.2d 1149 (5th Cir. 1978), cert. denied, 444 U.S. 879, 100 S.Ct. 167, 62 L.Ed.2d 108 (1979).
Judge MacKinnon's dissenting opinion in Clark v. Valeo expressed the view that the legislative veto provision at issue, the same one involved in Buckley v. Valeo, was unconstitutional. His opinion was primarily a response to Justice White's position in Buckley, see note 83 supra, that Article I, Section 7 does not apply because regulations subject to congressional disapproval either become effective by the nonaction of Congress or are disapproved by action that is equivalent to the rejection of proposed legislation. Judge MacKinnon's view was that Congress most definitely does "act" when it fails to veto an agency regulation because "it may act affirmatively by a majority vote to approve or the resolution to veto may not carry by the required majority. Either procedure would constitute an act or vote of approval for that house." 559 F.2d at 687 (MacKinnon, J., dissenting) (emphasis in original). Since Congress "acts" regardless how it votes on the resolution, it clearly engages in legislative action which must conform to the Article I, Section 7 requirements of presentation to the President and passage by both houses of Congress. See id. at 685-90. Like Justice White, Judge MacKinnon did not discuss the question whether the legislative veto violated the doctrine of separation of powers.

A federal district court in Montana recently upheld a statute empowering a House committee to declare an "emergency situation" with regard to public lands and to direct the Secretary of the Interior to withdraw the lands from mineral leasing activities. The court construed the statute narrowly as leaving to the Secretary the decision on the scope and duration of the withdrawal, however, so as to avoid problems under the separation of powers doctrine. Pacific Legal Foundation v. Watt, 529 F.Supp. 982 (D.Mont. 1982).
Impressively, in the past two years three state supreme courts have declared legislative veto schemes to be in violation of state constitutions. See Opinion of the Justices, 431 A.2d 783 (N.H. 1981) (violates state constitutional requirements of bicameral action by a majority of both legislative houses and presentation to executive for approval or veto); State ex rel. Barker v. Manchin, 279 S.E.2d 622 (W.Va. 1981) (violates state separation of powers doctrine); State v. A.L.I.V.E. Voluntary, 606 P.2d 769 (Alas. 1980) (violates state constitutional requirement of presentation to executive).

In Atkins v. United States the Court of Claims upheld (4-3) the one-house veto provision of the Federal Salary Act of 1967, under which the President's recommendations for salaries of certain government officials became law unless either house of Congress disapproved them within thirty days. In March 1974 the Senate vetoed the President's recommendations for judicial salary increases, and 140 federal judges brought suit claiming inter alia that the veto provision was unconstitutional. In rejecting this claim the court emphasized that it was focusing only "on this specific mechanism in this specific statute — how it works, what it involves, what values and interests are implicated — not on an overarching attempt to cover the entire problem of the so-called legislative veto, or even a large segment of it."

556 F.2d 1028 (Ct.Cl. 1977) (per curiam), cert. denied, 434 U.S. 1009, 98 S.Ct. 718, 54 L.Ed.2d 751 (1978).

2 U.S.C. § 359(1) (1976). This provision was amended in 1977 to provide that the recommendations take effect only if approved by both houses. 2 U.S.C. § 359(1) (Supp. III 1979).

The Atkins majority began by determining that Congress had power under the Necessary and Proper Clause to enact a veto provision. It then found that the Act's legislative review provision did not violate either the principle of bicameralism, because "the one-House veto does not alter the existing law in any fashion, but only preserves the legal status quo"; the Presentment Clause, because the President was permitted to recommend whatever he wished, including no recommendation at all; or the separation of powers doctrine, because the veto provision is an aid to legislative policymaking and does not interfere with any essential executive function. The three judges in dissent disagreed completely, arguing that the veto provision violated all of these constitutional doctrines.

Id. at 1063.

Id. at 1065.

Id. at 1069-70.

Id. at 1075-82 (Shelton, J., concurring in part and dissenting in part).

The other federal court ruling is Chadha v. Immigration Naturalization Service, in which the Ninth Circuit struck down a provision of the Immigration and Nationality Act allowing either house of Congress to veto a suspension of deportation granted by the Attorney General. The Immigration and Naturalization Service ("INS") ruled that Chadha could remain in the United States even though he was deportable, but the House vetoed this decision and the INS issued a final order of deportation. Chadha petitioned for appellate review. A unanimous panel declared the veto provision violative of the doctrine of separation of powers.

634 F.2d 408 (9th Cir. 1980), consideration of juris. postponed until hearing on merits, ___ U.S. ___, 102 S.Ct. 87, 70 L.Ed.2d 80 (1981).

The court held that such a violation occurs when there is "an assumption by one branch of powers that are central or essential to the operation of a coordinate branch, provided also that the assumption disrupts the coordinate branch in the performance of its duties and is unnecessary to implement a legitimate policy of the Government." It then stated that however the congressional review power was characterized, its use in the immigration context was unconstitutional: if used as a "correction of judicial or executive misapplication of the statute," it interfered with a core function of the Judiciary; if used as a "means for sharing the administration of the statute with the Executive on an ongoing basis," it interfered with a core function of the Executive; and if used as an "exercise of a residual legislative power," it violated the constitutional requirement of bicameralism. Like the majority in Atkins, the Chadha court sought to minimize the reach of its holding:

Id. at 429; see id. at 429-31.

Id. at 429; see id. at 431-33.

Id. at 429; see id. at 433-35.

[W]e are not here faced with a situation in which the unforeseeability of future circumstances or the broad scope and complexity of the subject matter of an agency's rulemaking authority preclude the articulation of specific criteria in the governing statute itself. Such factors might present considerations different from those we find here, both as to the question of separation of powers and the legitimacy of the unicameral device.

Id. at 433 (footnotes omitted).

This background makes clear that this case is truly one of first impression. The Supreme Court has never ruled on the legislative veto, and neither of the one-house veto decisions deals with legislative review of agency rulemaking. We therefore cannot rest our decision on any precedent but must consider carefully the applicability of the constitutional provisions relied on by both sides to the dispute.

B. Political Question

Congressional amici argue that this court should not decide the constitutionality of the one-house veto because this is a political question. They contend that until recently "the political branches resolved their disputes in this context solely by political means, without any intervention by the courts," and that "a stalemate or impasse on general principles necessitating judicial intervention is lacking." Since the Senate supports the legality of the House's action in this case, and since the Executive apparently agrees with the House on the undesirability of expanding incremental pricing, all that is presented here is "an abstract constitutional position." Citing this court's en banc decision in Clark v. Valeo, in which we refused to reach the merits of a challenge to a one-house veto provision, Congressional amici argue that this issue should be left to the politically representative branches.

Brief for Congressional Amici at 51-52.

Id. at 52.

559 F.2d 642 (D.C. Cir.) (en banc) (per curiam), aff'd mem. sub nom. Clark v. Kimmitt, 431 U.S. 950, 97 S.Ct. 2667, 53 L.Ed.2d 267 (1977).

Congressional amici raise the political question argument only with regard to the Presentment Clause, but address their arguments generally to whether this court should decide the case as a whole. We address the political question issue as if it had been raised in defense to all of petitioners' constitutional claims.

This reliance on Clark v. Valeo is misplaced. The Clark majority held that "the matter before us does not present a ripe `case or controversy' within the meaning of Article III," while specifically stating that "we need not address those [questions] pertaining to standing or political question, because the unripeness of the action is so pervasive." The case lacked ripeness because the legislative review provision at issue had not been invoked: "Until Congress exercises the one-house veto, it may be difficult to present a case with sufficient concreteness as to standing and ripeness to justify judicial resolution of the pervasive constitutional issue which the one-house veto provision involves." Here, of course, the House of Representatives did exercise the disapproval power conferred by section 202(c), thus presenting us with a concrete case in which to adjudicate the constitutionality of the one-house veto.

Id. at 649.

The court also stated: "Were this court to decide the threshold `case or controversy' questions differently, it would nevertheless refuse to reach the merits of the unicameral veto under the doctrine of judicial prudence enunciated in Samuels v. Mackell, 401 U.S. 66, 73, 91 S.Ct. 764, 768, 27 L.Ed.2d 688 (1971)." Id. at 650 n. 10. This "doctrine of judicial prudence," however, is limited to the declaratory judgment context. Samuels v. Mackell, a companion case with Younger v. Harris, 401 U.S. 37, 91 S.Ct. 746, 27 L.Ed.2d 669 (1971), held that a district court should refuse to grant relief by way of declaratory judgment to a plaintiff who raises an issue that is the subject of a pending state criminal prosecution against that plaintiff.
The court's obvious concern was that it not "strain to exercise its jurisdiction to resolve this momentous political as well as legal problem." 559 F.2d at 650. Significantly, its reasons for refusing to hear the case do not apply here. The court stated that "[a] constitutional decision might better await an attack upon legislative review by the entity subject to such review, which could be informed by the testimony of those who felt constrained to withhold proposals because of veto indications from the reviewing authorities." Id. at 650 n. 10. But here FERC decided that "it would be wholly inappropriate for the Commission to be a partisan in any potential conflict between Congress and the Executive." Brief for Respondent at 13 n. 14. Moreover, the question whether the agency felt constrained in its rulemaking is irrelevant here because the rule was vetoed. The Clark majority declared that judicial consideration of "legislative review mechanisms ought at an absolute minimum to be informed by experience and not depend solely on abstract analysis or speculation," so that the issue would be adjudicated in "a concrete setting, of regulations rejected or clearly trimmed in advance of submission." 559 F.2d at 650 n. 10. These conditions are satisfied by the House veto. Far from having merely a "bare `case or controversy,'" id., we have a live dispute that presents us with "a proper record compiled in an appropriate adversary proceeding with briefing of the issue worthy of its importance." Id. at 649 n. 8.

Baker v. Carr set forth the relevant political question considerations, which, as recently summarized by Justice Powell, lead to "three inquiries: (i) Does the issue involve resolution of questions committed by the text of the Constitution to a coordinate branch of Government? (ii) Would resolution of the question demand that a court move beyond areas of judicial expertise? (iii) Do prudential considerations counsel against judicial intervention?" The first two issues are easily resolved. Nothing in Article I, Section 7 suggests that Congress is permitted to decide when a legislative action must be performed by both houses and presented to the President; indeed, the inclusion of Clause 3 of that section was intended to prevent Congress from subverting the presentation requirement by merely restyling the nature of its action. Both requirements are specific limitations on congressional power, and it would be exceedingly anomalous to hold that the Framers meant for Congress to determine when the limitations apply. Moreover, no special nonjudicial expertise is needed in determining their applicability. Rather, these issues require "no more than an interpretation of the Constitution."

Goldwater v. Carter, 444 U.S. 996, 998, 100 S.Ct. 533, 62 L.Ed.2d 428 (1979) (Powell, J., concurring).

See note 128 infra.

Baker listed three prudential considerations that might lead a court to refuse to resolve a constitutional issue: "the impossibility of a court's undertaking independent resolution without expressing lack of the respect due coordinate branches of government; or an unusual need for unquestioning adherence to a political decision already made; or the potentiality of embarrassment from multifarious pronouncements by various departments on one question." The latter two are plainly inapplicable here, so the contention must rest on the need to respect the political branches. Congressional amici contend that a judicial resolution would violate that respect because the Legislature and the Executive have been able to resolve the dispute over the legislative veto "solely by political means" and because the dispute remains relatively young. They also contend that there actually is no "real dispute" over the propriety of the Phase II rule, and that combined with the lack of a stalemate on the veto issue, this is reason enough for this court to defer resolution. We disagree.

Brief for Congressional Amici at 51.


Whether there is a "real dispute" over the policy choice not to implement the Phase II rule is irrelevant. The critical dispute is whether the procedure by which that choice was made is constitutional. To accept Congressional amici's implicit principle we would have to hold that we will not hear any Executive challenge to the constitutionality of a legislative action, no matter how obviously unconstitutional, if we find that the Executive supports the policy embodied in that legislative act. Moreover, Congressional amici's emphasis on the policy views of the President ignores the fact that petitioners are private parties with an important stake in the resolution of this issue.

Congressional amici also state: "The Senate's support for the House in this brief highlights the abstractness of petitioners' asserting a right of the Senate to check the House." Id. at 52 n. 48. Congressional amici completely misread the purpose of the bicameralism requirement when they assert that so long as the Senate and House agree that one house may act alone there is no constitutional problem. If petitioners are correct, the veto represents a legislative action that the Constitution intended both houses to approve; the two houses' agreement to violate that intent cannot stand. Such a revision of constitutional intent can be accomplished only by constitutional amendment, not by mere agreement embodied in a statute.

Cf. Goldwater v. Carter, 444 U.S. 996, 1004, 100 S.Ct. 533, 62 L.Ed.2d 428 (1979) (Rehnquist, J., concurring in judgment) (in discussing why Youngstown Sheet Tube Co. v. Sawyer, 343 U.S. 579, 72 S.Ct. 863, 96 L.Ed. 1153 (1952), did not present a political question, emphasizing that "private litigants brought a suit contesting the President's authority under his war powers to seize the Nation's steel industry, an action of profound and demonstrable domestic impact").

We also reject the contention that "the President and Congress have not yet had time to reach a solution or else settle into a stalemate." Legislative review of independent agency rulemaking may be relatively new, but for sixty years there has been "a long tug of war between the Executive and Legislative Branches of the Federal Government" over the constitutionality of legislative review devices. Congressional amici themselves note that Congress has enacted over 200 such provisions since 1932. Legislative review provisions or similar measures have been attacked as unconstitutional by the administrations of Presidents Wilson, Hoover, Roosevelt, Truman, Eisenhower, Kennedy, Johnson, Nixon, Ford, Carter, and now, as exemplified by the position of the United States as amicus curiae, Reagan. A common form of objection has been to express reservations when signing bills into law, but Presidents have also vetoed bills on the ground that legislative veto provisions were unconstitutional. This historical record shows that the dispute is neither new nor temporary, but rather represents a clear disagreement between the political branches as to the meaning of the Constitution. A judicial resolution is therefore appropriate.

Brief for Congressional Amici at 53. Even if there were a long history of the use of the legislative veto, it does not automatically follow that the courts would refuse to intervene. See Powell v. McCormack, 395 U.S. 486, 546-47, 89 S.Ct. 1944, 23 L.Ed.2d 491 (1969) ("That an unconstitutional action has been taken before surely does not render that same action any less unconstitutional at a later date.")

Buckley v. Valeo, 424 U.S. 1, 140 n. 176, 96 S.Ct. 612, 692 n. 176, 46 L.Ed.2d 659 (1976) (per curiam).

Brief for Congressional Amici at 50. See generally Watson, Congress Steps Out: A Look at Congressional Control of the Executive, 63 Calif.L.Rev. 983, 1089-92 (1975) (listing statutes).

See Watson, supra note 118, at 988. n. 9. There has also been a considerable number of congressmen opposed to the legislative veto on constitutional grounds. See, e.g., id. at 988 n. 11. See generally id. at 1002-29.

The most famous example is President Roosevelt's decision to sign the Lend-Lease Act of 1941, but to submit to his Attorney General a memorandum, to be published when to do so would not offend political allies, stating his view that the Act's provision allowing the Act's authorization to be terminated by concurrent resolution was unconstitutional. See Jackson, A Presidential Legal Opinion, 66 Harv.L.Rev. 1353 (1953). For other examples, see Dixon, supra note 82, at 429-32 nn. 23-24, 29 (Ford and Carter); Watson, supra note 118, at 1016 n. 160 (Eisenhower). To some extent, presidential opposition appears to have been tempered by the presidential desire to obtain statutory authority. See J. Bolton, The Legislative Veto 10-13 (1977).

See, e.g., Dixon, supra note 82, at 428 n. 21, 429 n. 24, 432 n. 29 (vetoes by Nixon, Ford, and Carter). The Ford and Carter administrations testified strongly against the constitutionality of the legislative veto in various legislative hearings. See Improving Congressional Oversight of Federal Regulatory Agencies: Hearings on S. 2258, et al., Before the Senate Comm. on Government Operations, 94th Cong., 2d Sess. 124 (18-25 May 1976) (statement of Assistant Attorney General Antonin Scalia); Letter from Assistant Attorney General Patricia M. Wald to Rep. Peter W. Rodino, Jr. (5 May 1977) (letter prepared in response to congressional request for Justice Department opinion), cited in McGowan, supra note 82, at 1141-42. President Carter also issued a personal message to Congress arguing the unconstitutionality and undesirability of legislative veto mechanisms. Message to the Congress, 14 Weekly Comp.Pres.Doc. 1146 (21 June 1978).
In 1980 Secretary of Education Hufstedler, acting pursuant to an opinion by Attorney General Civiletti, refused to revise a regulation vetoed by Congress and declared that the regulation was still in effect. The regulation was withdrawn in 1981 by the new Secretary, thus ending the confrontation. See Nathanson, Separation of Powers and Administrative Law: Delegation, the Legislative Veto, and the `Independent' Agencies, 75 Nw.U.L.Rev. 1064, 1078 n. 55 (1981).

Congressional amici rely on this court's statement in United States v. American Tel. Tel. Co., 567 F.2d 121 (D.C. Cir. 1977), that "[t]he framers, rather than attempting to define and allocate all governmental power in minute detail, relied, we believe, on the expectation that where conflicts in scope of authority arose between the coordinate branches, a spirit of dynamic compromise would promote resolution of the dispute in the manner most likely to result in efficient and effective functioning of our governmental system." Id. at 127. They neglect to cite, however, the statement more relevant to the situation here: "Where the dispute consists of a clash of authority between two branches, however, judicial abstention does not lead to orderly resolution of the dispute." Id. at 126. Cf. Chadha, 634 F.2d at 419 ("It would stand the political question doctrine on its head to require the Judiciary to defer to another branch's determination that its acts do not violate the separation of powers principle.").
We also find it ironic that Congressional amici urge the political question doctrine so strongly upon us, given that a recent statute calling for congressional review of independent agency rulemaking specifically requests the courts to determine the statute's constitutionality as soon as possible:

(f)(1) Any interested party may institute such actions in the appropriate district court of the United States, including actions for declaratory judgment, as may be appropriate to construe the constitutionality of any provision of this section. The district court immediately shall certify all questions of the constitutionality of this section to the United States court of appeals for the circuit involved, which shall hear the matter sitting en banc.

(2) Notwithstanding any other provision of law, any decision on a matter certified under paragraph (1) shall be reviewable by appeal directly to the Supreme Court of the United States. Such appeal shall be brought not later than 20 days after the decision of the court of appeals.

(3) It shall be the duty of the court of appeals and of the Supreme Court of the United States to advance on the docket and to expedite to the greatest possible extent the disposition of any matter certified under paragraph (1).

15 U.S.C. § 57a-1(f) (West Supp. 1981) (FTC rulemaking). Congress' apparent desire to obtain an immediate ruling on the constitutionality of the FTC congressional veto contrasts sharply with its argument here that the issue is nonjusticiable because it is properly resolved by the political branches.

C. The Necessary and Proper Clause

Article I, Section 8, Clause 18 of the Constitution gives Congress power "To make all Laws which shall be necessary and proper for carrying into Execution the foregoing Powers, and all other Powers vested by this Constitution in the Government of the United States, or in any Department or Officer thereof." Since McCulloch v. Maryland, the Supreme Court has construed this authorization to permit Congress to exercise broad legislative powers not expressly enumerated in the Constitution. Congressional amici urge that under this principle the legislative veto must be upheld as a modern technique necessary for preserving the flexibility of congressional action. The Court of Claims in Atkins adopted this argument:

17 U.S. (4 Wheat.) 316, 4 L.Ed. 579 (1819).

In particular we note that the necessary and proper clause, which has sanctioned the massive delegation of legislative functions over the past century, provides a firm grounding for this veto. Congress plainly felt the need for this veto device, instead of relying solely on the power to override presidential recommendations by a full-fledged statute. In McCulloch's phrases, Congress exercised "its best judgment" in the selection of this measure and sought to "accommodate its legislation to circumstances."

Congressional amici conclude: "The Necessary and Proper Clause is the Constitution's expression that Congress must not be denied the flexible means needed to deal with exactly the kind of extraordinarily divisive issues resolved in the conference compromise that created § 202(c)."

Brief for Congressional Amici at 37. We reject the notion that the "extraordinarily divisive issues" presented in this case provide some special justification for use of the veto here. See id.; Gas Suppliers at 51 ("these unique circumstances"). Although many congressmen may have considered the NGPA and the comprehensive energy package to be two of the more difficult legislative decisions they had ever faced, there is nothing particularly unique about the use of the veto in this situation. Rather it is a typical provision authorizing one house of the Congress to veto a rule promulgated by an agency. We see no principled way of distinguishing the use of the veto in cases involving "extraordinarily divisive issues" from those involving "routine" issues. Cf. Buckley v. Valeo, 424 U.S. 1, 132, 96 S.Ct. 612, 688, 46 L.Ed.2d 659 (1976) (per curiam) ("We see no reason to believe that the authority of Congress over federal election practices is of such a wholly different nature from the other grants of authority to Congress that it may be employed in such a manner as to offend well-established constitutional restrictions stemming from the separation of powers.").

In our view, the Necessary and Proper Clause fails to advance the argument on behalf of the one-house veto. We do not understand either petitioners or the United States to contend that, in the absence of the asserted constitutional violations, the legislative veto provision would be unconstitutional because Congress does not have the authority to enact it. Rather, the claim is that however "necessary" Congress found this particular veto provision, it is not "proper" because it has other constitutional infirmities. The Necessary and Proper Clause does not override other provisions of the Constitution. McCulloch stated the test as follows: "Let the end be legitimate, let it be within the scope of the constitution, and all means which are appropriate, which are plainly adapted to that end, which are not prohibited, but consistent with the letter and spirit of the constitution, are constitutional." Plainly there is dispute whether the veto is "consistent with the letter and spirit of the Constitution," and reference to the Necessary and Proper Clause will not provide the answer. As the Ninth Circuit noted in Chadha, the clause "authorizes Congress to `make all laws,' not to exercise power in any way it deems convenient. That a power is clearly committed to Congress does not sustain an unconstitutional form in the exercise of the power."

17 U.S. (4 Wheat.) at 421.

634 F.2d at 433. See also Buckley v. Valeo, 424 U.S. 1, 134-35, 96 S.Ct. 612, 689-90, 46 L.Ed.2d 659 (1976) (per curiam) (rejecting argument that Necessary and Proper Clause provides authority to enact statutes violating other constitutional provisions).
Congressional amici suggest that Congress' power to employ the veto comes not from the first half of the Necessary and Proper Clause, which deals vertically with Congress' power in relation to the states, but rather from the second half, which deals horizontally with Congress' power over the Executive and the Judiciary. See Brief for Congressional Amici at 36 n. 33. The majority opinion in Atkins made the same point, but, like Congressional amici, did not explain its significance, instead merely citing to an article written by Professor Van Alstyne. See 556 F.2d at 1069. The thesis of this article is that this largely ignored second half of the clause actually is the source of exclusive congressional control over all nonessential powers exercised by the other two branches: "I will contend that this clause assigns to Congress alone the responsibility to say by law what additional authority, if any, the executive and the courts are to have beyond that core of powers that are indispensible, rather than merely appropriate, or helpful, to the performance of their express duties under articles II and III of the Constitution." Van Alstyne, The Role of Congress in Determining Incidental Powers of the President and of the Federal Courts: A Comment on the Horizontal Effect of the Sweeping Clause, Law Contemp. Prob., Spring 1976, at 102, 107.
The potential breadth of this argument is staggering. This view of the horizontal effect of the Necessary and Proper Clause calls into question the correctness of Supreme Court decisions recognizing the existence of executive privilege, see United States v. Nixon, 418 U.S. 683, 94 S.Ct. 3090, 41 L.Ed.2d 1039 (1974), executive power of removal of executive officers, see Myers v. United States, 272 U.S. 52, 47 S.Ct. 21, 71 L.Ed. 160 (1926), the supervisory power of the courts, and the power of the judiciary to provide for damage actions to vindicate constitutional rights, see Bivens v. Six Unknown Named Agents, 403 U.S. 388, 91 S.Ct. 1999, 29 L.Ed.2d 619 (1971). See Van Alstyne, supra, at 107-31. In other words, acceptance of the view that the second half of the clause places stringent limits on the inherent powers of the Executive and the Judiciary could work a radical transformation in the distribution of constitutional powers resulting from two centuries of judicial review. The author himself concedes that his article might be regarded "as an heretical and `revisionist' treatment of the constitutional principle of separated powers." Id. at 118. Plainly, such a change should come, if at all, only from the Supreme Court.
Moreover, we think the question of the inherent powers of the Executive and Judiciary is not central here. The Executive does not contend that it has inherent power to conduct rulemaking in the absence of a delegation from Congress. Rather, the claim is that when Congress does enact a law requiring the Executive or an independent agency to act, Congress may not then reserve for itself the right to nullify, by other than legislation passed by both houses and presented to the President, the decisions made by the administrator in carrying out the law. In other words, the question before us is not whether some power is implicit in the Constitution's grant of powers to the Executive; it rather is whether Congress may veto decisions the Executive has made pursuant to a legislative delegation. See Dixon, supra note 82, at 477-79. This question does not involve the Necessary and Proper Clause, which is properly read as "a means of making the exercise of powers by the various branches effective, not as a means of shifting powers between the branches of government." Special Prosecutor: Hearings Before the Senate Comm. on the Judiciary, 93d Cong., 1st Sess. 453 (1973) (testimony of Acting Attorney General Bork). The second half of the Necessary and Proper Clause no more empowers Congress to contravene constitutional requirements than does the first half.

D. The Constitutional Lawmaking Process

Article I, Section 7, Clause 2 provides that "Every Bill which shall have passed the House of Representatives and the Senate, shall, before it becomes a Law, be presented to the President of the United States." Clause 3 of the same section similarly provides that "Every Order, Resolution, or Vote, to Which the Concurrence of the Senate and House of Representatives may be necessary (except on a question of Adjournment) shall be presented to the President of the United States." In each case the President may give either his approval, in which case the action of Congress becomes law, or his disapproval, in which case the action may not become law unless it is repassed by two-thirds of both the House and the Senate. These provisions set up the fundamental prerequisites to the enactment of federal laws: bicameral passage of the legislation, and presentation for approval or disapproval by the President.

Clause 3 was inserted to prevent Congress from evading the requirements of Clause 2 by merely calling its action something other than a "bill." The Convention enacted this additional clause by a 9-1 vote. See 2 M. Farrand, The Records of the Federal Convention of 1787, at 301-02, 304-05 (rev'd ed. 1937).

Article I, Section 1 also makes clear that legislation shall be enacted by a bicameral Congress: "All legislative Powers herein granted shall be vested in a Congress of the United States, which shall consist of a Senate and House of Representatives."

Congressional amici make three basic arguments in support of their contention that a legislative veto may be exercised without regard to these two requirements. First, they assert that the one-house veto is similar to other types of legislative action which historically have been undertaken by one house alone and which have not been presented to the President. Second, they assert that an agency rule subject to congressional veto is merely a proposal which can be rejected by either house of Congress in the same way that proposed legislation may be rejected by either house. Finally, they analyze the purposes of the bicameralism and presentation requirements, and conclude that these purposes do not mandate application of the requirements to the exercise of the one-house veto. We reject each of these arguments, and hold that congressional disapproval of final agency rules must be conducted in accordance with the requirements of Article I, Section 7.

1. Exceptions to the requirements of Article I, Section 7

Congressional amici emphasize that Article I, Section 7, Clause 3 provides that the presentation requirement applies to actions "to which the Concurrence of the Senate and House of Representatives may be necessary," and assert that "[t]he term `may' has a meaning here consistent with its ordinary and usual construction; some actions of the Congress may require concurrence and presentation, while others may not." They cite "four categories of actions, having the force of law and affecting persons outside the Congress," that historically have been conducted by only one house or have not been presented to the President, or both, and assert that this demonstrates that Article I, Section 7 should be flexibly construed.

Brief for Congressional Amici at 56 (emphasis in original).

Id. at 57.

We agree that not every action taken by Congress automatically must receive the concurrence of both houses and be presented to the President for approval, but we reject the implication that these requirements are often and lightly ignored. On the contrary, the exceptions are clearly narrow and provide no support for finding Article I, Section 7 inapplicable to one-house consideration of an agency rule.

a. Legislative initiation of agency investigations

This category actually refers only to statutory authority which at one time permitted either house or the President to require the Federal Trade Commission ("FTC") to investigate prescribed subjects. The critical difference between investigation and rulemaking was identified in Buckley v. Valeo. The Supreme Court declared that powers "of an investigative and informative nature" can be performed by a commission constituted of legislatively appointed officers, because the powers fall "in the same general category as those powers which Congress might delegate to one of its own committees." Rulemaking, on the other hand, "represents the performance of a significant governmental duty exercised pursuant to a public law," and thus cannot be conducted by such a commission. This FTC provision, therefore, tells nothing about whether Article I, Section 7 applies to a resolution disapproving a rule. A resolution enacted under this statute does not affect substantive law, as occurs when Congress vetoes an agency rule, but only instructs the FTC to conduct an investigation. Congress' powers with regard to investigations are unique and irrelevant to its power to make law.

15 U.S.C. § 46(d) (1976) gives the FTC power "Upon the direction of the President or either House of Congress to investigate and report the facts relating to any alleged violations of the antitrust Acts by any corporation." This is modified by id. § 46a, which states that "After June 16, 1933, no new investigations shall be initiated by the Commission as the result of a legislative resolution, except the same be a concurrent resolution of the two Houses of Congress."

Id. at 137, 96 S.Ct. at 690.

Id. at 141, 96 S.Ct. at 692.

In Atkins the Court of Claims noted that "Congress is acknowledged to have ... the power to carry on investigations, even though this . . . is not expressly provided for in the Constitution," 556 F.2d at 1062, and used this as evidence that Congress may deviate from the requirements of Article I, Section 7 even where not expressly provided in the Constitution. The Supreme Court has held, however, that the investigatory power is "an essential and appropriate auxiliary to the legislative function," and therefore is implicit in Article I, Section I. McGrain v. Daugherty, 273 U.S. 135, 174, 47 S.Ct. 319, 328, 71 L.Ed. 580 (1927). The power therefore is constitutionally based. Article I, Section 7 does not apply because investigations are an incident to the legislative power, not an exercise of it that produces binding substantive rules of conduct.

b. Presidential plans for executive reorganization

For much of this century, the President has been authorized to formulate executive reorganization plans which will take effect unless either house votes a resolution of disapproval. Since these provisions raise the same unresolved constitutional questions presented in this case, they cannot be viewed as support for the contention that Article I, Section 7 may be ignored. It is true that the Executive has generally supported the constitutionality of this basic statutory scheme, but only on limited and narrow grounds. While this apparent accommodation between the Executive and the Congress on use of the veto in reorganization statutes might be relevant to a challenge to those statutes, it does not bear on the veto's constitutionality in rulemaking or other contexts, where the Executive has strongly opposed the one-house disapproval device.

The current authorization is at 5 U.S.C. § 906 (Supp. III 1979).

See, e.g., 43 Op. Att'y Gen. 2 (31 Jan. 1977) (Attorney General Bell) [hereinafter Bell Opinion] (expressing "opinion that the [disapproval] procedures provided in ... the reorganization statute are constitutionally valid," but emphasizing this "is to be taken in no manner as approving the constitutionality of the procedure of congressional disapproval of executive action by resolution in other statutes"). But see 37 Op. Att'y Gen. 56, 63-64 (1933) (Attorney General Mitchell) ("The attempt to give to either House of Congress, by action which is not legislation, power to disapprove administrative acts, raises a grave question as to the validity of the entire provision in the [statute] for Executive reorganization of governmental functions.").
The view that reorganization statutes are a legitimate exception to the requirements of Article I, Section 7 is based on a "reverse legislation" theory, which holds that when Congress authorizes the President to make any proposal (including none at all) on a given subject matter, free from all procedural constraints but subject to a one-house veto, the basic requirements of Article I, Section 7 are fulfilled. One elaboration of the theory is as follows:

In more precise constitutional terms, the essential aspects of a potential "reverse legislation" exception to the general objection to one-house vetoes are the following: (1) viewing such statutes as the Salary and Reorganization Acts as calling for a discrete sequence of legislative choices (even though loosely bound together by an initial statute) that are not intertwined with program administration, article II considerations are not reached; (2) the President's article I role is preserved because instead of having a veto power at the end, he has a proposal power at the beginning; (3) the congressional role is preserved because unless both houses concur, by withholding a negative, the status quo remains unchanged; and (4) it is critical to the safeguarding of the President's role in this context that there be no selective or partial veto power in Congress. If Congress has a power to modify, that is, to negate part of the President's proposal but not all of it, with no opportunity for the President to say whether he would agree to severability, then his constitutional parity with Congress in lawmaking is undermined.

Dixon, supra note 82, at 485. See also Bell Opinion, supra, at 2-5. Under the reverse legislation theory, the Atkins case was properly decided, even though it did not expressly rely on this rationale. See, e.g., Dixon, supra note 82, at 471. But cf. Bell Opinion, supra, at 5 (one-house veto "is uniquely appropriate to executive reorganization," which is "a matter in which the President has a peculiar interest and special responsibility.").
Many commentators have challenged the viability of the reverse legislation theory. See, e.g., Nathanson, supra note 121, at 1090; McGowan, supra note 82, at 1141 n. 93, 1159 n. 176; Scalia, The Legislative Veto: A False Remedy For System Overload, Regulation, Nov./Dec. 1979, at 19, 22. See also Separation of Powers: Hearings Before the Subcomm. on Separation of Powers of the Senate Comm. on the Judiciary, 90th Cong., 1st Sess. 250-51 (1920 July, 2 Aug. 13-15 Sept. 1967) [hereinafter " Separation of Powers Hearings"] (statement of Alexander Bickel) (if reorganization act veto is constitutional, all legislative vetoes are constitutional). A major problem is that it concedes that the original congressional delegation is standardless and that the real congressional action in authorizing presidentially formulated policy comes from the failure of either house to disapprove. As the Ninth Circuit observed in Chadha, there are grave problems with such a theory of approval by inaction — "an inaction that could equally imply endorsement, acquiescence, passivity, indecision, or indifference." 634 F.2d at 435. And, as we discuss at note 167 infra, the entire theory breaks down if the scheme is changed to disapproval by concurrent resolution. Finally, although the theory purports to be a limited exception to Article I, Section 7, it actually could support "a regime under which any presidential proposal on a given subject would become part of the United States Code if no house of Congress objected within, say, sixty days." Chadha, 634 F.2d at 435.
As some observers have noted, a true reverse legislation scheme would be where presidential proposals took effect only if approved by both houses of Congress. The roles of each branch would be reversed in time, but not in substance. See Watson, supra note 118, at 1072. This is how the Federal Salary Act, which formerly authorized the one-house veto device at issue in Atkins, now stands. 2 U.S.C. § 359(1) (Supp. III 1979).
In any event, whatever the validity of the reverse legislation hypothesis, it clearly is inapplicable here because (1) FERC is an independent agency and it, not the President, formulated the Phase II rule; (2) FERC was required, not simply permitted, to issue such a rule; and (3) the veto disrupted an ongoing administrative program.

c. Presidential foreign affairs and national defense decisions

Congress has often combined its delegation of foreign affairs authority to the Executive with provisions for disapproval of actions by concurrent resolution. As with the veto in reorganization statutes, the constitutionality of these provisions has not been resolved, and they therefore cannot serve as solid support for finding exceptions to Article I, Section 7. And, although a number of these schemes have been enacted, the Executive has never completely accepted their constitutionality. In any event, the foreign affairs veto presents unique problems since in that context there is the additional question whether Congress or the President or both have the inherent power to act.

See, e.g., 50 U.S.C. § 1544(c) (1976) (War Powers Resolution) (presidential authority to use armed forces without declaration of war or specific statutory authorization may be terminated by concurrent resolution); id. § 1622(a)(1) (presidential declaration of national emergency may be terminated by concurrent resolution).

In normal contexts, the Supreme Court has held that concurrent resolutions have the force of law only if presented to and approved by the President. See United States v. California, 332 U.S. 19, 28, 67 S.Ct. 1658, 1663, 91 L.Ed. 1889 (1947) (vetoed joint resolution "does not represent an exercise of the constitutional power of Congress to dispose of public property under Article IV, § 3, Cl. 2"); United States ex rel. Levey v. Stockslager, 129 U.S. 470, 475, 9 S.Ct. 382, 383, 32 L.Ed. 785 (1889) (joint resolution "was approved by the President, and took effect only on such approval"). See also Watts v. United States, 161 F.2d 511, 513 (5th Cir.) ("A joint resolution of the House and Senate, when approved by the President, has the effect of law."), cert. denied, 332 U.S. 769, 68 S.Ct. 81, 92 L.Ed. 354 (1947). Thus, although it is true that "Congress can express its opinion through resolutions which are not subject to presidential veto, even though such power is not expressly granted in the Constitution," Atkins, 556 F.2d at 1062, this point is irrelevant to our analysis, because such "opinion" resolutions have no substantive effect.

For example, the War Powers Resolution, see 50 U.S.C. §§ 1541- 1548 (1976), was passed over President Nixon's veto, which was based in part on his belief that § 5(c) ( 50 U.S.C. § 1544(c)), requiring immediate withdrawal of troops if Congress enacts a joint resolution not presented to the President, was unconstitutional. The President's Message to the House of Representatives Returning H.J.Res. 542 Without His Approval, 9 Weekly Comp.Pres.Doc. 1285, 1286 (24 Oct. 1973). And when President Ford signed the International Security Assistance and Arms Export Control Act of 1976, Pub.L. No. 94-329, 90 Stat. 729 (1976) (codified in scattered sections of 22 U.S.C.), he expressed his reservations about the constitutionality of section 211(a) ( 22 U.S.C. § 2776(b) (1976)), which permits Congress to veto proposed arms sales by concurrent resolution. See Statement by the President on Signing H.R. 13680 Into Law, 12 Weekly Comp.Pres.Doc. 1104, 1105 (1 July 1976).

Presidents have long maintained that congressional authorizations of the use of military force are unnecessary. See, e.g., S.Rep. No. 797, 90th Cong., 1st Sess. 22 (1967) (statement of President Johnson, 18 Aug. 1967) ("`We stated then, and we repeat now, we did not think the [Gulf of Tonkin] resolution was necessary to do what we did and what we're doing.'"] No such claim is made with regard to rulemaking authorizations.

d. Congressional proposals for constitutional amendments

The final category cited by Congressional amici is the only one in which there is a judicial decision holding that Article I, Section 7 did not entirely apply. In Hollingsworth v. Virginia the Supreme Court ruled that proposals for constitutional amendments, although requiring the concurrence of both houses, do not require presentation to the President. The Court did not explain this holding, but Justice Chase stated in a footnote to the Attorney General's brief: "The negative of the president applies only to the ordinary cases of legislation; he has nothing to do with the proposition or adoption of amendments to the constitution." By not mentioning presidential participation, Article V, which sets forth the procedure for amending the Constitution, makes clear that proposals for constitutional amendments are congressional actions to which the presentation requirement does not apply. No constitutional provision does the same with regard to resolutions disapproving agency rules. The reasons for not presenting a proposed amendment to the President therefore are irrelevant.

3 U.S. (3 Dall.) 378, 1 L.Ed. 644 (1798).

Id. at 381 n."a" (Chase, J.). See also id. (argument of the Attorney General) ("[T]he case of amendments is evidently a substantive act, unconnected with the ordinary business of legislation, and not within the policy or terms of investing the president with a qualified negative on the acts and resolutions of congress.").

There are other express constitutional exceptions to one or both requirements of Article I, Section 7. See U.S. Const. art. I, § 2, cl. 5 (House has "sole Power of Impeachment"); id. art. I, § 3, cl. 6 (Senate has "sole Power to try all Impeachments"); id. art. I, § 5, cl. 1 ("Each House shall be the Judge of the Elections, Returns, and Qualifications of its own Members"); id. art. I, § 5, cl. 2 ("Each House may determine the Rules of its Proceedings, punish its Members for disorderly Behaviour, and, with the Concurrence of two thirds, expel a Member"); id. art. II, § 2, cl. 2 (Senate has power to concur in treaties and give advice and consent to presidential appointments).

Congressional amici assert that "actions of the Houses accord with the Presentation Clause when they draw their validity from a larger procedure: for legislative review, the enactment of the original legislation delegating authority; for constitutional amendments, the process of state ratification." Brief for Congressional Amici at 60-61. This proposed test makes no sense. A congressional proposal for a constitutional amendment draws its validity from the Constitution, not the process of state ratification. As Justice Chase said, the President simply has no formal role in the process of proposing and ratifying constitutional amendments. Furthermore, we disagree that the delegation of veto authority in the original legislation is sufficient to comply with requirements of the Presentment Clause. See pp. 468-469 infra.

Thus, we agree with the main point of Congressional amici — that Article I, Section 7 does not apply rigidly to every single congressional activity — but we resist their implication that law or history supports a relaxed construction of the presentation and bicameralism requirements. Only with regard to constitutional amendments is there a Supreme Court ruling that the Presentment Clause does not apply, and that is clearly an exceptional case. The other alleged exceptions either do not represent substantive lawmaking, as with initiation of agency investigations, or represent limited areas of Executive-Legislative accommodation in the exercise of special powers, an accommodation that nonetheless remains clouded by constitutional controversy. The question remaining, then, is whether there is some reason why the decision of one house to veto a rule should be deemed different from the kind of legislative decisions the Constitution requires to be made by both houses and presented to the President.

In Atkins the majority noted that many "policy choices" do not require adherence to Article I, Section 7: decisions "not to introduce a bill, or not to vote on a bill once introduced," or committee decisions "which have great policy significance," or individual decisions to make comments which "can often result in changed policies, not only throughout the Federal Government, but in the private sector." Atkins, 556 F.2d at 1062 n. 24. These statements are of course true, but they prove only the obvious, and indeed trivial, point that not every word uttered or action taken within a Congress of 535 members must receive the concurrence of both houses and be presented to the President. None of these actions purport to have the force of law, and they thus are irrelevant to the question whether adherence to Article I, Section 7 is required when Congress, or one house thereof, formally adopts a resolution of disapproval which clearly does affect the substantive law.

2. Purposes of the lawmaking procedures a. Presentation to the President

The primary reason for the presidential veto power conferred by Article I, Section 7 was to give the President a defensive weapon against legislative intrusions on the powers of the Executive. The importance of this concern was emphasized by Alexander Hamilton in The Federalist:

The propensity of the legislative department to intrude upon the rights and to absorb the powers of the other departments has been already suggested and repeated; the insufficiency of a mere parchment delineation of the boundaries of each has also been remarked upon; and the necessity of furnishing each with constitutional arms for its own defense has been inferred and proved. From these clear and indubitable principles results the propriety of a negative, either absolute or qualified, in the Executive upon the acts of the legislative branches. Without the one or the other, the former would be absolutely unable to defend himself against the depredations of the latter. He might gradually be stripped of his authorities by successive resolutions, or annihilated by a single vote. And in the one mode or the other the legislative and executive powers might speedily come to be blended in the same hands. If even no propensity had ever discovered itself in the legislative body to invade the rights of the Executive, the rules of just reasoning and theoretic propriety would of themselves teach us that the one ought not to be left to the mercy of the other, but ought to possess a constitutional and effectual power of self-defense.

The Federalist No. 73, at 489-90 (P. Ford ed. 1898) (A. Hamilton).

This concern is reflected in the Convention debates, which centered largely on whether the veto should be absolute or qualified and whether the Judiciary should share the veto power with the Executive.

See, e.g., 1 M. Farrand, supra note 128, at 97-104, 138-40, 144; 2 M. Farrand, supra, at 73-80, 301-02, 304. See generally Watson, supra note 118, at 1044-48.

In addition, however, the value of a check against ill-advised legislation was recognized and urged as an important justification for the negative. Hamilton, after discussing the veto as a means of protecting presidential authority, wrote:

Although the debates indicate that use of a negative as a means of protecting the Executive and Judiciary from legislative encroachment was the primary concern, the veto's potentially more general use was clearly contemplated. See, e.g., 1 M. Farrand, supra note 128, at 97-98 ("Mr. Gerry doubts whether the Judiciary ought to form a part of [the proposed Council of Revision], as they will have a sufficient check agst. encroachments on their own department by their exposition of the laws, which involved a power of deciding on their Constitutionality.... It was quite foreign from the nature of ye. office to make them judges of the policy of public measures."); id. at 104 ("Mr. Gerry observed that a power of suspending might do all the mischief dreaded from the negative of useful laws; without answering the salutary purpose of checking unjust or unwise ones."); id. at 139 (Mr. Madison observed that "whether the object of the revisionary power was to restrain the Legislature from encroaching on the other co-ordinate Departments, or on the rights of the people at large; or from passing laws unwise in their principle, or incorrect in their form, the utility of annexing the wisdom and weight of the Judiciary to the Executive seemed uncontestable"); 2 M. Farrand, supra at 73 (Mr. Wilson stated that "this power of the Judges did not go far enough. Laws may be unjust, may be unwise, may be dangerous, may be destructive; and yet not be so unconstitutional as to justify the Judges in refusing to give them effect. Let them have a share in the Revisionary power, and they will have an opportunity of taking notice of these characters of a law, and of counteracting, by the weight of their opinions the improper views of the Legislature."); id. ("Mr. Ghorum did not see the advantage of employing the Judges in this way. As Judges they are not presumed to possess any peculiar knowledge of the mere policy of public measures.... He thought it would be best to let the Executive alone be responsible."); id. at 74 (Mr. Madison said combining the Judiciary with the Executive in the Revisionary power "would moreover be useful to the Community at large as an additional check agst. a pursuit of those unwise unjust measures which constituted so great a portion of our calamities."); id. at 78 ("Col Mason observed that the defence of the Executive was not the sole object of the Revisionary power. He expected even greater advantages from it." Since the Legislature "must be expected frequently to pass unjust and pernicious laws," the "restraining power was essentially necessary.")
The proposal to include the Judiciary in the veto power was defeated, but not due to opposition to the existence of any policy check on the Legislature. If that were the case, the President's veto power would also have been defeated, as Madison noted: "If such a Union [between the Executive and the Judiciary] was an improper mixture of powers, or such a Judiciary check on the laws, was inconsistent with the Theory of a free Constitution, it was equally so to admit the Executive to any participation in the making of laws; and the revisionary plan ought to be discarded altogether." Id. at 77. The Convention did not discard the plan, nor did it limit the reach of the revisionary power; it simply refused to include the Judiciary.

But the power in question has a further use. It not only serves as a shield to the Executive, but it furnishes an additional security against the enaction of improper laws. It establishes a salutary check upon the legislative body, calculated to guard the community against the effects of faction, precipitancy, or of any impulse unfriendly to the public good which may happen to influence a majority of that body.... The primary inducement to conferring the power in question upon the Executive is to enable him to defend himself; the secondary one is to increase the chances in favor of the community against the passing of bad laws through haste, inadvertence, or design.

The Federalist No. 73, at 490-91 (P. Ford ed. 1898) (A. Hamilton). See also Sherman, A Citizen of New Haven, II, in Essays on the Constitution of the United States 240 (P. Ford ed. 1892) (published 25 Dec. 1788) ("The united wisdom and various interests of a nation should be combined in framing the laws.... [T]he partial negative vested in the President by the new Constitution on the acts of Congress and the subsequent revision, may be very useful to prevent laws being passed without mature deliberation.").

There can be no doubt, then, that policy use of the presidential veto was within the contemplation of those who drafted and ratified Article I, Section 7.

See A. Kelly W. Harbison, The American Constitution 338-39 (4th ed. 1970); J. Story, I Commentaries on the Constitution of the United States §§ 885-886 (5th ed. 1905); Clineburg, The Presidential Veto Power, 18 S.C.L.Rev. 732, 737 (1966); Federal Administrative Law Developments — 1975, 1976 Duke L.J. 285, 290.

Early Presidents used the veto rarely, and generally believed that the power was properly employed only against bills that encroached on presidential authority or that were considered unconstitutional. Given the unrestricted text of the Presentment Clause, as well as the inherent difficulty of distinguishing one veto purpose from another, these limitations could not hold up. Andrew Jackson was the first President openly to declare that he was entitled to exercise the veto on policy grounds, and since then the veto has been used predominantly to disapprove legislative policy. Though there were isolated attempts to strip the President of power to exercise policy vetoes, these never were seriously debated, and by and large such vetoes have been popular with the public. The Supreme Court has several times stated that the Presentment Clause permits the President to veto legislative decisions he considers unwise.

As Professor Black notes, vetoing a bill on grounds of unconstitutionality is a different, and broader, rationale than vetoing a bill to protect executive authority. Despite the absence of discussion on the former during the debates, there has been little controversy about the validity of this purpose. See Black, Some Thoughts on the Veto, Law Contemp. Prob., Spring 1976, at 87, 89; Zinn, The Veto Power of the President, 12 F.R.D. 207, 230 (1952).

For example, two of Madison's vetoes have been interpreted either as defensive or as policy based. Compare Black, supra note 153, at 90-91, with E. Corwin, The President 279 (4th ed. 1957).

Naturally the veto power did not escape the early talent of Americans for conjuring up constitutional limitations out of thin air. The veto was solely a self-defensive weapon of the President; it was the means furnished him for carrying out his oath to `preserve, protect and defend the Constitution' and was not validly usable for any other purpose; it did not extend to revenue bills, never having been so employed by the King of England; it did not extend to `insignificant and trivial' matters like private pension bills; it was never intended to give effect merely to presidential desires, but its use must rest on considerations of great weight, and so on and so forth. Although efforts of this sort to forge shackles for the power derived a certain specious plausibility from the rarity of the veto's use in English history, they met with failure from the first.

E. Corwin, supra note 154, at 279 (footnote omitted).

See A. Kelly W. Harbison, supra note 152, at 338-39.

See E. Corwin, supra note 154, at 279 (failed attempts to amend the Constitution to strip President of veto power); Black, supra note 153, at 91 (attempt to impeach President Tyler for misuse of veto power).

See H. Hockett, The Constitutional History of the United States 1826-1876, at 84-85 (1939); Black, supra note 153, at 91 (citing W. Binkley, President and Congress 97-98 (1947)).

See, e.g., Youngstown Sheet Tube Co. v. Sawyer, 343 U.S. 579, 587, 72 S.Ct. 863, 866, 96 L.Ed. 1153 (1952) ("The Constitution limits [the President's] functions in the lawmaking process to the recommending of laws he thinks wise and the vetoing of laws he thinks bad."); The Pocket Veto Case, 279 U.S. 655, 678, 49 S.Ct. 463, 466, 73 L.Ed. 894 (1929) ("And it is just as essential a part of the constitutional provisions, guarding against ill-considered and unwise legislation, that the President, on his part, should have the full time allowed him for determining whether he should approve or disapprove a bill, and if disapproved, for adequately formulating the objections that should be considered by Congress, as it is that Congress, on its part, should have an opportunity to re-pass the bill over his objections.").

In sum, there simply can be no serious argument today that it is unconstitutional for the President to use a veto to further his policy views. That this purpose was originally considered secondary makes it no less entitled to consideration. And that there was limited controversy over the purpose's legitimacy is far outweighed by the consistent practice for 150 years that policy vetoes are the rule, defensive vetoes the exception. It makes no sense to argue that the primary and undeniably legitimate use of a constitutional power may be dropped from consideration entirely when deciding whether that provision is applicable to a new situation. To hold that the presidential veto power does not have a policy purpose in the legislative review context would logically require that the same holding be applied to all legislation, a result that would alter the balance of Executive-Legislative power so radically that no court could conceive of adopting it. In construing the validity of the failure to present the disapproval resolution to the President, therefore, we must consider both the effect on the President's ability to protect his authority from encroachment and the effect on his ability to check unwise legislation.

It may be true that the original understanding was that "the veto would be used only rarely, and certainly not as a means of systematic policy control over the legislative branch," Black, supra note 153, at 90, but this is because of the belief that the President would not feel secure enough to exercise the veto frequently, not because of any general opposition to policy vetoes. In any event, it makes no sense to conclude that since the veto power became more powerful than expected, it should not be applied to congressional disapproval resolutions. This confuses two inquiries: (1) whether the veto power should be limited because its use has exceeded original expectations, and (2) whether a disapproval resolution is sufficiently like those congressional actions to which the Presentment Clause, whatever its scope, applies, so that the resolution should be presented to the President. It is noteworthy that Professor Black concludes from his analysis not that courts should seek to limit the power of the veto, but that Congress should use concurrent resolutions, which do not have the force of law, to express its policy views where those views otherwise are effectively blocked by the existence of the presidential veto. See id. at 99-101.

Accord Atkins, 556 F.2d at 1065. We note that the constitutional analysis of the need for presentation of disapproval resolutions does not depend on whether the resolution is simple (one-house) or concurrent, even though the Presentment Clause states that only actions of both houses need be presented. As one commentator has noted, "It verges on irrationality to maintain that action by concurrent resolution, whereby Congress is at least held in check by its own structure, is invalid because the veto clause so states, but that the invalidity of a simple resolution, wherein a single House acts without check, is more in doubt." Watson, supra note 118, at 1066 n. 428.

b. Bicameralism

There is little doubt as to Article I, Section 1's purpose in vesting all legislative powers "in a Congress" consisting of "a Senate and House of Representatives." Perhaps the greatest fear of the Framers was that in a representative democracy the Legislature would be capable of using its plenary lawmaking power to swallow up the other departments of the Government. The Supreme Court noted in Buckley that "the debates of the Constitutional Convention, and the Federalist Papers, are replete with expressions of fear that the Legislative Branch of the National Government will aggrandize itself at the expense of the other two branches." A corollary fear was that national majorities would be able to exercise unconstrained power, thereby destroying the power of state governments. The answer to both concerns proved to be bicameralism. Requiring the concurrence of two branches in enacting laws would help prevent a single majority from undertaking to control the Federal Government. And the Great Compromise, dividing the Congress into one branch directly representing the people and one branch representing the states, was devised as a means of preserving state power. The overriding objective of bicameralism, then, is to constrain the exercise of the federal legislative power by making sure that the Legislature can act only where representatives of two different constituencies are in agreement.

424 U.S. at 129, 96 S.Ct. at 687. See also Chadha, 634 F.2d at 433 (citing various provisions of Convention debates and The Federalist).

See, e.g., 1 M. Farrand, supra note 128, at 336-38.

See generally Watson, supra note 118, at 1030-37.

As Madison put it,

In republican government, the legislative authority necessarily predominates. The remedy for this inconveniency is to divide the legislature into different branches; and to render them, by different modes of election and different principles of action, as little connected with each other as the nature of their common functions and their common dependence on the society will admit.

The Federalist No. 51, at 345 (P. Ford ed. 1898) (J. Madison).
The scope and force of bicameralism are not diminished by the Convention's rejection of language that would have specified that each house "shall in all cases have a negative on the other." 2 M. Farrand, supra note 128, at 197. The debates make clear that there was concern that this language might call into question the specific exceptions to bicameral action, such as appointments and treaties. As a solution, "Mr. Madison moved to strike out the words `each of which shall in all cases, have a negative on the other[']; the idea being sufficiently expressed in the preceding member of the Article; vesting the `legislative power' in `distinct bodies'; especially as the respective powers and mode of exercising them were fully delineated in a subsequent article." Id. at 197 (footnote omitted). Madison's motion was adopted 7-3. Quite clearly all it accomplished was the avoidance of confusion whether certain stated exceptions to bicameral action were valid; it in no way suggested that the significance or scope of the bicameralism requirement was diminished.

3. The one-house veto under Article I, Section 7

What emerges from our analysis of the purposes of the lawmaking restrictions in Article I is that the Framers were determined that the legislative power should be difficult to employ. The requirements of presentation to the President and bicameral concurrence ultimately serve the same fundamental purpose: to restrict the operation of the legislative power to those policies which meet the approval of three constituencies, or a supermajority of two. If the legislative veto represents an exercise of the legislative power, then, it must be exercised only in compliance with these constitutional requirements.

We agree with Congressional amici that the defensive purpose of the presidential veto was fulfilled by the President's opportunity to veto the NGPA in 1978. If the President believes the legislative veto is an unconstitutional encroachment on his authority, the appropriate time to object by veto is when the power is created — here, when section 202(c) of the NGPA was passed. That this veto may be overriden does not make it meaningless. The Framers specifically rejected proposals to give the President an absolute negative on Congress. If two-thirds of the House and Senate disagree that the legislative veto is unconstitutional, they may enact it over the President's veto.
We recognize the substantial obstacles to the exercise of a presidential veto against a one-house veto provision. For example, having lobbied so hard for national energy legislation, President Carter would have been hard pressed to veto the NGPA because of its legislative review provisions, which were an extremely minor part of the statute. But Article I, Section 7's defensive purpose does not require more than that the President have an opportunity to veto legislation he deems unconstitutional.
This does not mean, however, that the President in any way waives his right to object to the legislative veto if he fails to veto it or if his veto is overridden. Such a "waiver" argument has been read into Justice White's statement that "the provision for congressional disapproval of agency regulations does not appear to transgress the constitutional design, at least where the President has agreed to legislation establishing the disapproval procedure or the legislation has been passed over his veto." Buckley, 424 U.S. at 286, 96 S.Ct. at 758 (White, J., concurring in part and dissenting in part). See also Note, Congressional Oversight of Administrative Discretion: Defining the Proper Role of the Legislative Veto, 26 Am.U.L. Rev. 1018, 1032 (1977). We prefer to read this passage as simply stating the obvious point that an attempt by one house to disapprove rules in the absence of statutory authorization would present a different issue from a one-house veto passed pursuant to effective legislation. To adhere to the waiver argument would require that one "believe that the Congress could pass a bill stripping the President of all his veto power over any subsequent legislation — which bill, if passed by two-thirds after presidential disapproval, would be entirely constitutional since, after all, the President had his opportunity to veto." Oversight and Review of Agency Decisionmaking, 28 Ad.L.Rev. 569, 691 (1976) (statement of Antonin Scalia). Such a result would ignore the fact that the presidential veto power ultimately inures to the benefit of the citizenry by acting as a check on the exercise of the plenary legislative power. See Gewirtz, supra note 82, at 78 n. 129 ("[I]f a statute dilutes the President's veto power has not something gone awry in the constitutional system which no concurrence among the occupants of two political branches should be able to insulate from challenge?").
Judicial review of the constitutionality of legislative acts is independent of the presidential veto power. The Supreme Court has never hesitated to find an unconstitutional encroachment on presidential authority simply because the President signed the statute at issue. See, e.g., National League of Cities v. Usery, 426 U.S. 833, 841 n. 12, 96 S.Ct. 2465, 2469 n. 12, 49 L.Ed.2d 245 (1976) (noting that in Buckley v. Valeo and Myers v. United States the Court had struck down laws as unconstitutional encroachments on executive power even though the President had failed to veto them). The President remains free to challenge the constitutionality of any statute he or his predecessor signs, as, of course, are private parties.

Congressional amici argue that Article I, Section 7 does not apply because FERC's Phase II rule was never an effective law, but merely a proposal to be accepted or rejected by Congress through the one-house disapproval mechanism. Since one house of Congress undoubtedly may refuse to enact proposed legislation, and may do so without the participation of the President, acceptance of this argument would end all constitutional inquiry. The policy purposes of Article I, Section 7 would be irrelevant. We emphatically reject this position, however, and hold that the veto of the Phase II rule effectively changed the law by altering the scope of FERC's discretion and preventing an otherwise valid regulation from taking effect. Accordingly, the Senate's concurrence and presentation to the President were necessary prerequisites to the effectiveness of the disapproval resolution.

The contention that the Phase II rule was a mere legislative proposal is easily refuted by comparing the status of the rule to that of a bill introduced in Congress. If neither house of Congress acts to approve a bill, the bill dies; but when neither house acts to disapprove an agency rule, the rule becomes law. Clearly the agency rule has some legal force that a proposed bill does not. If Congress had wanted FERC merely to propose an extension of incremental pricing to be accepted or rejected by Congress, it could have so provided. FERC's Phase II proposal would then have become law only if approved by both houses of Congress and the President, or by two-thirds of each house. Instead, Congress provided that the Phase II rule could take effect as a valid agency action pursuant to a congressional delegation, though subject to disapproval by majority vote of one house.

Indeed, there is a clear logical flaw in the argument that the effect of a veto "is as though a bill passed in one House and failed in another," Buckley, 424 U.S. at 285 n. 30, 96 S.Ct. at 757 n. 30 (White, J., concurring in part and dissenting in part), which is made evident by considering the case of a two-house veto. If only one house disapproves a rule under that scheme, the rule takes effect. Yet if the rule were merely a proposal whose underlying policy was never set by Congress, then one house's rejection should be sufficient to kill it. Clearly, then, the rule under a two-house veto scheme has more legal force than a legislative proposal. Cf. Chadha, 634 F.2d at 435 n. 41 (argument that nonobjection of both houses changes the law is not applicable where review provision requires disapproval by both houses). And surely a rule issued pursuant to a one-house veto scheme cannot have a different legal status from a rule issued pursuant to a two-house veto scheme. We cannot believe, for example, that Congress intended rules promulgated by the FTC to have presumptive legal effect because subject to a two-house veto, see 15 U.S.C.A. § 57a-1 (West Supp. 1981), while intending FERC rules, which are promulgated pursuant to the same administrative process but subject instead to a one-house veto, to have no effect at all. This would imply that the two-house veto is more constitutionally suspect than the one-house veto because the former works to overturn an authorized rule whereas the latter merely rejects a legislative proposal. The obvious anomaly of this argument makes clear that it is untenable. See Kilbourn v. Thompson, 103 U.S. 168, 182, 26 L.Ed. 377 (1881) ("neither branch of Congress, when acting separately, can lawfully exercise more power than is conferred by the Constitution on the whole body").
In response it has been argued:

[S]uch syllogistic reasoning ignores the basis for the constitutional distribution of legislative power. The Framers feared control of Congress by parochial legislative interests. When the proposals emanate from the President or his agents, however, that fear is unfounded: the President is presumed by the Constitution to act in the national interest. This presumption permits an executive proposal made pursuant to a congressional delegation of power, without a legislative veto provision, to become law even though the proposal might not have the majority support of either House of Congress. The source of the proposals, then, provides the safeguard against the dangers that bicameralism was instituted to avoid. Thus, a statute requiring one-House approval or two-House disapproval of executive proposals in a postenactment setting is functionally equivalent to the constitutional scheme.

Javits Klein, Congressional Oversight and the Legislative Veto: A Constitutional Analysis, 52 N.Y.U.L.Rev. 455, 493-94 (1977) (footnote omitted).
It can safely be said that such an argument would have left most of the Framers stupefied. The idea that a "presumption" that the President will act beneficently is equivalent to the fundamental check of bicameralism conflicts totally with the pervasive theme that the only "presumption" is that there will be "a clashing of interests." 3 M. Farrand, supra note 128, at 400 (statement of Mr. Dayton in the Senate, 24 Nov. 1803). Although the Framers feared legislative dominance, they labored under no belief that an unrestrained President would be any less dangerous. On the contrary, their decisive rejection of proposals to give him an absolute negative was based on the same view of human nature that led to restricting the Legislature to bicameral action subject to presidential veto.
It is also noteworthy that this argument on its own terms does not apply to independent agency rulemaking, as we have here, since the "proposal" does not come from the President. See Javits Klein, supra, at 490.

Such a delegation is permissible because it was created by a statute enacted pursuant to Article I, Section 7. The delegation may be quite broad, so long as Congress performs "[t]he essentials of the legislative function," which are "the determination of the legislative policy and its formulation and promulgation as a defined and binding rule of conduct." In other words, "there is no forbidden delegation of legislative power `if Congress shall lay down by legislative act an intelligible principle' to which the official or agency must conform." An agency rule issued pursuant to such a delegation is binding and effective without action by either Congress or the President.

Yakus v. United States, 321 U.S. 414, 424, 64 S.Ct. 660, 667, 88 L.Ed. 834 (1944) (Stone, C.J.).

Amalgamated Meat Cutters Butcher Workmen v. Connally, 337 F.Supp. 737, 746 (D.D.C. 1971) (three-judge court) (quoting J. W. Hampton, Jr., Co. v. United States, 276 U.S. 394, 409, 48 S.Ct. 348, 352, 72 L.Ed. 624 (1928) (Taft, C.J.)).

We recognize that in practice Congress will usually take some "action" when it determines not to veto a rule, as Judge MacKinnon argued in Clark, 559 F.2d at 685-90 (MacKinnon, J., dissenting). Judge MacKinnon's view appears to be that under a congressional veto scheme, the President is entitled to presentation of both a decision to veto, see id. at 689, and a decision not to veto, see id. at 690 (emphasis in original) ("In approving regulations submitted by the FEC the concurrence of both houses is necessary, since if one house disagrees, the regulations are killed. Hence, such regulations must be presented to the President.").
We disagree on this latter point. If Congress tries but fails to pass a statute to overturn an effective regulation, it certainly engages in "action." Yet there can be no argument that Congress' failure to pass a statute must be presented to the President for veto. In the same way, its failure to disapprove an agency rule does not constitute formal and official action invoking Article I, Section 7. Congress (or one house) "acts" for purposes of this provision when it affirmatively passes some bill, order, vote, or resolution that alters the substantive law; it does not act when such proposals are not adopted and the substantive law — i.e., the promulgated regulations — remain unchanged. To adopt Judge MacKinnon's alternative view would be to suggest that regulations promulgated under a two-house veto scheme with provision for presentation to the President would become effective only by the failure of both houses to disapprove, and only after the President had an opportunity to veto the regulations. Such a holding would empower the President to veto both regulations and disapproval of regulations, an unwarranted and illogical extension of his power under Article I, Section 7. The proper analysis, in our view, is that promulgated regulations become effective without further action by either Congress or the President.

An effective delegation, therefore, must embody some basic congressional policy decision. That policy decision, plus what the agency does in accordance with that decision, constitutes the law on the subject matter. We recognize that, due to the dormancy of the delegation doctrine, courts have frequently upheld extremely broad delegations giving little policy guidance to the administrators, but the Supreme Court has never held that Congress may abdicate entirely its responsibility to make a policy decision when it authorizes agency action. Arguments that Congress actually decided nothing whatsoever about the extension of incremental pricing, therefore, would raise serious questions as to the validity of the Phase II delegation itself.

See generally K. Davis, Administrative Law Treatise §§ 3.1-3.8 (2d ed. 1978).

Congressional amici and intervenors argue strenuously that Congress actually decided nothing whatsoever about the extension of incremental pricing. See Brief for Congressional Amici at 65-66; Brief for Gas Suppliers at 46 ("[S]ection 202 contemplated merely a proposal which was to be reviewed by Congress"); Brief for Gas Consumers at 41 ("[N]o serious reading of the statute can support a conclusion that Congress pre-judged the issue"); Brief for PEG at 9 n. 8 ("The decision on extending this rule under Section 202 was left to another day"). It unquestionably is true that Congress sought to avoid a definitive decision on extension of incremental pricing. But if, as these parties suggest, Congress was utterly neutral on whether such extension should occur at all, then a serious question would arise whether the delegation could be effective at all, unless we were willing, as we are not, to accept the view that the existence of a legislative veto can serve to validate an otherwise too broad delegation. Cf. Chadha, 634 F.2d at 431 ("We would be reluctant to view the statutory criteria of section 244 as so broad and indefinite as to permit constant legislative redefinition that at the same time does not amount to an effective amendment, for such a view would raise serious delegation problems.") Moreover, acceptance of Congressional amici's view would necessarily lead to the conclusion that the Phase II rule would have been a new policy that could take effect without affirmative approval by either house and without any presidential participation. Since the Constitution clearly envisages all three entities combining to set national policy, such a construction would be just as constitutionally dubious.

Moreover, in this case there is no question that Congress did set forth an original policy on incremental pricing, one which was reconsidered by the House in disapproving the Phase II rule. It may be true that Congress contemplated that it would make the final decision on extending incremental pricing after FERC issued a rule, but it is indisputable that Congress wanted FERC to do more than merely propose legislation. Section 202 of the NGPA represents an authorization, approved by both houses and the President, giving FERC discretion to issue an incremental pricing rule that would attain binding legal effect without further congressional action. The supporters of incremental pricing believed it an essential component of the gas deregulation compromise.

See pp. 443-444 notes 60-67 supra.

The 1980 debates in the House make clear that the policy embodied in Title II and in the Phase II rule was being re-evaluated in light of the limited experience under Phase I and the economic conditions that had developed since 1978. Representative Preyer commended the "vigorous rejection of a theory overrun by reality," Representative Stockman stated that "our purpose today is simply to lay to rest by the agency of legislative veto a dubious regulatory experiment that has been overtaken by the rush of events and unanticipated developments," and Representative Gramm added that "every argument that has been made . . . against phase II is equally applicable to phase I." Congressional amici themselves admit that in vetoing the rule the House was making "a social policy judgment," an "assessment that is peculiarly legislative in nature" because it is a "predictive assessment, using public participation and collective decisionmaking, with democratic accountability as the final test of judgment." The only thing missing from this description is the crucial fact that the Congress had already made one judgment on this policy problem, and had agreed to allow FERC to formulate a rule that would go into effect without further congressional action. In taking its "second look" at the problem, Congress undeniably engaged in a reconsideration of its previously enacted policy.

126 Cong.Rec. H3843 (daily ed. 20 May 1980) (remarks of Rep. Preyer); see id. at H3845 (remarks of Rep. Brown).

Id. at H3848 (remarks of Rep. Stockman); see id. at H3849-50 (remarks of Rep. Levitas); id. at H3844 (remarks of Rep. Lundine).

Id. at H3848 (remarks of Rep. Gramm).

Brief for Congressional Amici at 43-45.

126 Cong.Rec. H3841 (daily ed. 20 May 1980) (remarks of Rep. Sharp) ("We are using a legislative veto not as a way to discipline that agency, but as a way for us, as I think [was] intended by the conferees on the 1978 act, to have a second look at the policy that we could not be certain of[,] that we were not willing to mandate at that time.").

This is precisely the kind of decision that the Constitution envisions will be made only by both houses with the participation of the President through his veto power. The President and both houses of Congress agreed on a policy when they took their "first look." Undoing this policy requires adherence to the same procedure. The Senate's views on incremental pricing expansion were as critical as those of the House. If the Senate disagreed that the rule was undesirable, the two houses would have been in conflict whether incremental pricing should be halted, and the original authorization should have gone forward.

The same is true of presidential participation. Clearly the President's first opportunity to veto was not sufficient to cover the disapproval resolution; he had an opportunity to approve incremental pricing, to the extent adopted by FERC, but he did not have an opportunity to approve the abolition of extended incremental pricing, which is what the disapproval resolution effectively accomplished. If, as Congressional amici contend, section 202(c) assigned to Congress "a single, discrete, well-defined role" of "deciding whether to expand a major experiment," the President was entitled to participate in that decision. Otherwise, Congress has created a device which effectively expands the reach of its legislative power by permitting it to determine conclusively whether effective law is created or not.

Those who argue that an opportunity for presidential veto of the regulations is equivalent to an opportunity to veto the disapproval resolution, see, e.g., Javits Klein, supra note 167, at 483, 486-88, miss the crucial point that the presumption under a rulemaking statute is that the agency rule will take effect. The agency is authorized to formulate such a rule as it sees fit, in accordance with the statutory delegation. The President's power to "veto" a proposed rule of a subordinate official is thus inherent in the delegation itself. Once the agency decides on a rule, however, the rule may become effective law. A congressional attempt to stop the rule thus amends the original delegation. The President's opportunity to veto this disapproval — in effect to preserve the original delegation and prevent a change in the law — is a quite different thing from an opportunity to veto the rule, and one which the Presentment Clause seeks to protect by requiring concurrence and presidential approval or concurrence of a supermajority.
Here, in any event, the rule was issued by an independent agency, so the President was not given a chance to "veto" it.

Brief for Congressional Amici at 47.

It is true that a resolution of disapproval does not amend the statutory language and thus, to that extent, differs from enactment of a statute which obviously must comply with Article I, Section 7. But there is no question that the effect of a congressional veto is to alter the scope of the agency's discretion. In this case, the practical effect probably was to withdraw the discretion altogether; Representative Brown seemed justified in saying, "The overwhelming vote to disapprove phase II . . . should be a clear message to the FERC not to send up another incremental pricing rule." In other cases, exercise of the legislative veto may enable one house of Congress effectively to dictate that a specific type of rule be promulgated. In either case, one house of Congress is enabled to enact a policy to which the other house and the President did not agree originally, a result that violates the requirements of Article I, Section 7.

126 Cong.Rec. H3845 (daily ed. 20 May 1980) (remarks of Rep. Brown).

It has been suggested that the nonobjection of either house is merely a condition precedent to the effectiveness of the rule, such that the rule never gains legal effect until the condition is satisfied. This may describe the intent of Congress, but we do not find it relevant to a determination of the veto's constitutionality. Merely styling something as a condition on a grant of power does not make that condition constitutional. Otherwise Congress could, for example, provide that all rights and duties established by legislation are conditioned on the vote of either house of Congress to eliminate them, thus enabling instant repeal of all statutes by simple resolution. In effect, Congress could use "conditions" as a means of circumventing completely the stringent restrictions on the legislative power in Article I.

See, e.g., Atkins, 556 F.2d at 1068; Schwartz, The Legislative Veto and the ConstitutionA Reexamination, 46 Geo.Wash.L.Rev. 351, 364 n. 109 (1978); Abourezk, The Congressional Veto: A Contemporary Response to Executive Encroachment on Legislative Prerogatives, 52 Ind.L.J. 323, 336-38 (1977). Primary reliance is placed on Currin v. Wallace, 306 U.S. 1, 59 S.Ct. 379, 83 L.Ed. 441 (1939), in which the Supreme Court upheld a statute giving the Secretary of Agriculture power to designate tobacco auction markets and inspect tobacco, on the condition that the affected farmers voted in a referendum to approve the program, and on United States v. Rock Royal Cooperative, Inc., 307 U.S. 533, 59 S.Ct. 993, 83 L.Ed. 1446 (1939), in which the Court upheld a statute authorizing the Secretary to define milk marketing areas and propose marketing orders, which could go into effect only if approved by the affected parties.
It is argued that conditioning administrative decisions on approval of private groups certainly cannot be any more constitutionally authorized than conditioning administrative decisions on congressional approval. As one commentator has noted, "This is a non sequitur.... A delegation which disperses power is not necessarily constitutionally equivalent to one which concentrates power in the hands of the delegating agency." Watson, supra note 118, at 1067 n. 430.
More important, the issues raised concerning the legislative veto simply would not be involved in a decision upholding a delegation to a private party:

A court decision to allow congressional delegation of power to private groups on grounds so unstructured yet conclusive, however bad as public policy, involves in constitutional terms solely a question of delegation doctrine. No other constitutional considerations come into play. Such a ruling could furnish no constitutional warrant for the quite different situation that is reached when, by a congressional veto device imposed on administrative authority, the Congress tries to blur or wipe out the constitutional prescriptions of article I, section 7 (regarding the scope of the congressional role in policymaking), and of article II (regarding the role of the Executive in executing statutes).

Dixon, supra note 82, at 455. These prior decisions regarding conditions placed on either statutory or administrative policies, therefore, are not relevant to this case, because they are of a different kind altogether.
In this regard, it is interesting that some supporters of the constitutionality of the one-house veto believe that the committee veto, whereby disapproval power is conferred on a single committee of one house, is unconstitutional. See, e.g., Schwartz, supra, at 364 n. 109. Yet if the constitutionality of the one-house veto is demonstrated by its status as a "condition," then the same would be true of the committee veto. The only reason the two possibly are different is that some conditions are unconstitutional. And that, of course, demonstrates why restyling the legislative veto as a condition is unhelpful. To accept the Atkins majority's finding that "[t]he doctrine of unconstitutional conditions is in no way pertinent to this case," on the ground that power over federal salaries is legislative rather than executive, see 556 F.2d at 1068, would be to make every conceivable delegation of disapproval power to some part of the Congress — even to a single congressman — per se constitutional.

It is worth recalling here that Clause 3 of Article I, Section 7 was adopted precisely to avoid congressional circumvention of Clause 2 by calling legislative action something other than a bill. See note 128 supra. It would thus be ironic if the decisive factor in determining the constitutionality of the legislative veto were whether it can fairly be termed a "condition." See Separation of Powers Hearings, supra note 138, at 252 (statement of Alexander Bickel) (Congress "has simply gone the long way around, and put the thing in another form of words about conditions and so forth, but it has done nothing different.").

It suffices here that in practical effect the House's veto changed national incremental pricing policy. FERC's Phase II rule was authorized to take effect without additional congressional action. The absence of such action cannot be equated with affirmative congressional action to alter the law. Congress attempted to do by one house what the Constitution requires be done only by both houses and the President. Section 202(c) is therefore unconstitutional.

The article I authorization to make law does not permit positive law which alters the substantive legal rights of individuals to be enacted by a mere executive recommendation, which is not a final exercise of specifically delegated power to alter these legal rights, followed by legislative inaction — an inaction that could equally imply endorsement, acquiescence, passivity, indecision, or indifference.

634 F.2d at 435.

E. Separation of Powers

The one-house veto's violation of Article I, Section 7 is a sufficient basis for its unconstitutionality. Presumably, a legislative review mechanism permitting a rule to be repealed by a joint resolution presented to the President would present no constitutional problems. Even though such a device would still differ from enactment of a statute — since the statutory language would remain the same although the specific action was forbidden, and since a veto resolution is easier to adopt than an affirmative bill — the essential elements of the constitutional lawmaking process would participate. There would be neither an increase in total federal power nor a violation of separation of powers. Congress would be exercising power in a manner consistent with Article I.

We note, however, that the analysis might well differ in the case of agency adjudication rather than rulemaking. Legislative disapproval of an individual adjudication, even if voted by both houses with the President's approval or by two-thirds of each house, could create bill of attainder, due process, or equal protection problems. See Chadha, 634 F.2d at 435 n. 42; cf. Pillsbury v. FTC, 354 F.2d 952 (5th Cir. 1966) (congressional interference with adjudication violates due process).

When Congress seeks to enable one house to exercise effective control over administrative decisions, however, the separation of powers principle comes into play. Petitioners contend that the one-house veto intrudes on the functions of the Executive and the Judiciary by permitting a single house of Congress to invalidate a rule on the ground that the exercise of discretion was either unwise or unlawful. Congressional amici respond that the House's veto did not intrude on the judicial power because the veto was based primarily on policy objections rather than a belief that FERC had violated its mandate. They also assert that there was no intrusion on the executive power because FERC is an independent agency not subject to executive control and that review of rule-making is legislative rather than executive or judicial in nature.

We conclude that this one-house veto violated the separation of powers doctrine. The fact that FERC is not subject to presidential control does not entitle Congress to direct FERC's administrative decisions. The Supreme Court has held that rulemaking is substantially a function of administering and enforcing the public law. As such, Congress may not create a device enabling it, or one of its houses, to control agency rulemaking. Congress' duty to oversee agency action is connected with its ultimate power of revising the laws under which the agency operates. The creation of further congressional power violates the Constitution.

1. Meaning of separation of powers

The most recent comprehensive statement on the separation of powers doctrine by the Supreme Court was in the Buckley v. Valeo per curiam opinion, the relevant portion of which was joined by seven Justices. The doctrine of separation of powers "is at the heart of our Constitution." There is common "recognition of the intent of the Framers that the powers of the three great branches of the National Government be largely separate from one another." The fundamental purpose of this separation is to check the extent of power exercisable by any one branch of Government in order to protect the people from oppression. Justice Brandeis' famous quotation still stands as the most concise statement of this objective:

424 U.S. at 118-43, 96 S.Ct. at 681-93. This portion of the opinion was joined by Chief Justice Burger and Justices Brennan, Stewart, Marshall, Blackmun, Powell, and Rehnquist. Justice White joined "much of" this part, but wrote separately. Id. at 257, 96 S.Ct. at 744. (White, J., concurring in part and dissenting in part). Justice Stevens did not participate.

Id. at 119, 96 S.Ct. at 682.

Id. at 120, 96 S.Ct. at 682.

The doctrine of the separation of powers was adopted by the Convention of 1787, not to promote efficiency but to preclude the exercise of arbitrary power. The purpose was, not to avoid friction, but, by means of the inevitable friction incident to the distribution of governmental powers among three departments, to save the people from autocracy.

Myers v. United States, 272 U.S. 52, 293, 47 S.Ct. 21, 84, 71 L.Ed. 160 (1926) (Brandeis, J., dissenting). The Ninth Circuit in Chadha argued that a second purpose of separation of powers was to promote efficiency by serving "to facilitate administration of a large nation by the assignment of numerous labors to designated authorities." 634 F.2d at 423. See also Brief for United States at 23-24. Such a "division of labor" certainly tends to enhance the efficiency with which various tasks are performed, though we have no doubt that the Framers intended this consideration to be subordinate to the concern for dividing powers to prevent tyranny. In any event, there is considerable doubt whether any efficiency is gained by enabling one house of Congress to strike down a rule promulgated pursuant to an orderly administrative process and to insert nothing in the rule's place. See p. 475 infra.

It is also clear from Buckley v. Valeo that "total separation" was not contemplated: the Framers "viewed the principle of separation of powers as a vital check against tyranny," but "they likewise saw that a hermetic sealing off of the three branches of Government from one another would preclude the establishment of a Nation capable of governing itself effectively." As a result, courts correctly have abstained from rigidly applying this principle to every apparent intermingling of power. It is false to reason from this, however, as have some supporters of the legislative veto, that separation of powers is a mere theoretical construct with little or no practical significance. As the Court held, "The principle of separation of powers was not simply an abstract generalization in the minds of the Framers: it was woven into the document that they drafted in Philadelphia in the summer of 1787." Accordingly, the "Court has not hesitated to enforce the principle of separation of powers."

424 U.S. at 121, 96 S.Ct. at 683.

See, e.g., Miller Knapp, The Congressional Veto: Preserving the Constitutional Framework, 52 Ind.L.J. 367, 390 (1977) ("It is doubtful that the concept of separation of powers can really have any objective meaning.").

Buckley, 424 U.S. at 124, 96 S.Ct. at 684.

Id. at 123, 96 S.Ct. at 684.

2. Intrusion into the executive sphere a. Significance of FERC's independence

Congressional amici argue that the House's veto of the Phase II rule by definition did not interfere with the Executive's administration of the statute because FERC is not part of the Executive Branch. The Commission is an independent regulatory agency "within the Department [of Energy]" for purposes of its budget, but functionally independent because its members "shall hold office for a term of four years and may be removed by the President only for inefficiency, neglect of duty, or malfeasance in office." Since the Supreme Court has upheld the constitutionality of such agency independence, Congressional amici argue that it is clear that the President can have no claim to participation in the making of FERC's rules: "FERC performs only those functions as assigned to it by Congress; it derives no independent authority from the Constitution, and separation of powers principles do not apply to it."

42 U.S.C. §§ 7171(a), (b) (Supp. III 1979). The Commission is also authorized to represent itself in court, as it has done here. Id. § 7171(i).

See Humphrey's Executor v. United States, 295 U.S. 602, 55 S.Ct. 869, 79 L.Ed. 1611 (1935).

Brief for Congressional Amici at 40-41.

The contention that the separation of powers doctrine does not apply to independent agencies is manifestly groundless. In Buckley v. Valeo the Supreme Court applied the doctrine to the Federal Election Commission ("FEC"), an independent agency. It is true that the President, as representative of the Executive, does not have a claim to control the decisionmaking of independent agencies. But it is an enormous, and unwarranted, jump from this to the conclusion that Congress may itself interfere with an independent agency's decisions without regard to separation of powers. Although FERC is substantially independent of the Executive, it nonetheless performs executive functions. The constitutionality of agency independence has not turned on a determination that certain agency functions are properly legislative rather than executive in nature. Buckley held directly the contrary, finding that the FEC's functions represented "the performance of a significant governmental duty exercised pursuant to a public law.... These administrative functions may therefore be exercised only by persons who are `Officers of the United States.'"

424 U.S. at 141, 96 S.Ct. at 692.

We find, therefore, that the constitutionality of the one-house veto does not depend on whether it is used against an executive agency or an independent agency. There has been a general breakdown in any distinction between the functions of the two types of agency, and in practice the interference by Congress is identical in either case. Indeed, it is ironic that Congressional amici attempt to place great significance on the Commission's independence and on the need for having a politically accountable check on the agency's decision. The fundamental justification for making agencies independent is that since they exercise adjudicatory powers requiring impartial expertise, political interference is undesirable. By then turning around and asserting that this independence is a justification for the one-house veto, Congress attempts simultaneously to decrease the power of the Executive and increase its own power.

Executive agencies perform the same adjudicatory and rulemaking functions as do independent agencies. See, e.g., 29 U.S.C. § 655 (1976) (Secretary of Labor shall by rule promulgate occupational safety and health standards). See generally K. Davis, supra note 171, at §§ 2:7-2:8; Nathanson, supra note 121, at 1099-109.

Cf. Watson, supra note 118, at 1081 n. 463 ("That a commission acts as an agent of the legislative department makes more emphatic the necessity that these bodies not be used as tools to give Congress a lawmaking role independent of the President.").

b. Congressional disruption and control of the administrative process

All parties agree that the particular decision made in this case by the House is best characterized as a general legislative policy decision. The key issue was whether incremental pricing should be extended at all, not whether the specific extension in the Phase II rule was the best one. Nonetheless, some congressmen were concerned that the specific rule issued by FERC was inappropriate, and, more generally, it is clear that one-house review mechanisms such as section 202(c) may give Congress the power to control the agency's discretion in considerable detail. We therefore must consider whether the one-house veto permits Congress to interfere unconstitutionally with FERC's administrative discretion.

See, e.g., 126 Cong.Rec. H3840 (remarks of Rep. Dingell) ("[T]he Commission's rule is seriously flawed"); id. at H3847 (remarks of Rep. Markey) ("FERC has caved in to industry pressure to write a rule of little effectiveness."); id. at H3854 (remarks of Rep. Ashley) ("I am convinced that the promulgation of the rule as currently constituted would be a serious mistake in public policy.").

Many proponents of the constitutionality of the legislative veto concede that the one-house review mechanism may in some instances raise serious problems of interference with executive functions. Justice White, for example, stated that he "would be much more concerned if Congress purported to usurp the functions of law enforcement, to control the outcome of particular adjudications, or to pre-empt the President's appointment power," than when Congress provides for legislative review of agency rulemaking. This and similar views rest on a belief that rulemaking is inherently a legislative, rather than executive, function. Seven Justices in Buckley, however, rejected this position in unequivocal terms:

Buckley, 424 U.S. at 285, 96 S.Ct. at 757 (White, J., concurring in part and dissenting in part).

Javits Klein, supra note 167, at 476-77; Nathanson, supra note 121, at 1087-88; Note, The Constitutionality of the Legislative Veto, 23 Wm. Mary L.Rev. 123, 133-34 (1981).

[The Commission's rulemaking, advisory opinion, and eligibility determination] functions, exercised free from day-to-day supervision of either Congress or the Executive Branch, are more legislative and judicial in nature than are the Commission's enforcement powers, and are of kinds usually performed by independent regulatory agencies or by some department in the Executive Branch under the direction of an Act of Congress. Congress viewed these broad powers as essential to effective and impartial administration of the entire substantive framework of the Act. Yet each of these functions also represents the performance of a significant governmental duty exercised pursuant to a public law. While the President may not insist that such functions be delegated to an appointee of his removable at will, Humphrey's Executor v. United States, 295 U.S. 602, 55 S.Ct. 869, 79 L.Ed. 1611 (1935), none of them operates merely in aid of congressional authority to legislate or is sufficiently removed from the administration and enforcement of public law to allow it to be performed by the present Commission. These administrative functions may therefore be exercised only by persons who are "Officers of the United States."

424 U.S. at 140-41, 96 S.Ct. at 692-93 (emphasis added) (footnotes omitted). Justice White characterized the FEC's rulemaking function as "purely legislative," but nonetheless concluded that this function had to be performed by Article II officers: "I know of no authority for the congressional appointment of its own agents to make binding rules and regulations necessary to or advisable for the administration and enforcement of a major statute where the President has not participated either in the appointment of each of the administrators or in the fashioning of the rule or regulations which they propound." Id. at 281, 96 S.Ct. at 755 (White, J., concurring in part and dissenting in part) (emphasis added). The tension in this argument is extreme. If rulemaking is an Article I function there is no reason why it must be performed by Article II officers. The only solution lies in Buckley's determination that rulemaking is substantially an executive function.
This holding in Buckley also demonstrates the error in the Atkins majority's assertion that "in exercising the delegated functions, the executive officer merely acts as an agent of the legislative branch of the Government." 556 F.2d at 1068. When the Executive acts to carry out laws passed by Congress, it remains a coordinate, not a subordinate, branch.

In contrast, the Court did hold that powers "of an investigative and informative nature, falling in the same general category as those powers which Congress might delegate to one of its own committees," could be exercised by a Commission whose members were appointed by Congress.

424 U.S. at 137, 96 S.Ct. at 690.

It of course is true that the specific holding of Buckley concerned the Appointments Clause of Article II, Section 2, Clause 2, and that here there is no specific clause forbidding the Congress from exercising rulemaking powers. Nonetheless, as the Buckley Court emphasized, Article II, Section 3 entrusts to the President, not to Congress, "the responsibility to `take Care that the Laws be faithfully executed.'" If rulemaking is sufficiently an executive function so that only Article II officers may conduct it, it would seem a fortiori that Congress is prohibited from substantial interference in the rulemaking process. It would be anomalous in the extreme to hold that Congress may not appoint the officials who make rules, but may enact a mechanism permitting effective congressional control over these officials' decisions.

Id. at 138, 96 S.Ct. at 691.

Congressional veto proponents contend, however, that the one-house veto is simply one of many techniques Congress may use in carrying out its responsibility to oversee the administration of the laws. The error in this argument is that the one-house veto provides Congress with a fundamentally different kind of authority over administrative rulemaking from what it otherwise would have. When Congress conducts investigations or hearings, or enacts a "report and wait" requirement, or threatens to reduce appropriations, or imposes reporting requirements, or engages in other modes of oversight, its ability to influence the agency derives almost entirely from its ability to pass a statute requiring a different agency action or reducing the agency's appropriations. These oversight methods enable Congress to inquire "into past executive branch action in order to influence future executive branch performance." Congress' supervisory power thus comes directly from its legislative power.

Under such a procedure, also known as the "laying system," the agency action must lay before Congress for a specified period before going into effect. The purpose is to give Congress an opportunity to pass a statute. Since there is no disturbance of the constitutional lawmaking process, the provisions present no apparent constitutional problems. See Clark v. Valeo, 559 F.2d at 648-49 nn. 5-6. See generally Watson, supra note 118, at 1060-61.

Javits Klein, supra note 167, at 461 (emphasis in original).

Even passage of an opinion resolution declaring congressional intent is persuasive ultimately because of the implied threat of a statutory amendment.

The one-house veto, on the other hand, effectively enables Congress "to participate prospectively in the approval or disapproval of . . . law `enacted' by the executive branch pursuant to a delegation of authority by Congress." In effect, Congress is able to expand its role from one of oversight, with an eye to legislative revision, to one of shared administration. This overall increase in congressional power contravenes the fundamental purpose of the separation of powers doctrine. Congress gains the ability to direct unilaterally, and indeed unicamerally, the exercise of agency discretion in a specific manner considered undesirable or unachievable when the enabling statute was first passed. Not only does this expand the congressional power, but it may also expand the total national power. Because of the veto, the rulemaking agency is given greater power than Congress might otherwise delegate, and Congress normally will let rules take effect unless so clearly undesirable that a veto is deemed warranted.

Javits Klein, supra note 167, at 462.

See Bruff Gellhorn, Congressional Control of Administrative Regulation: A Study of Legislative Vetoes, 90 Harv.L.Rev. 1369, 1427-28 (1977).

See Watson, supra note 118, at 1073-74; Scalia, supra note 138, at 25-26. Alexander Hamilton, though not aware of the legislative veto possibility, succinctly described the pernicious results of allowing national policy to be easily made and easily altered:

The internal effects of a mutable policy are still more calamitous. It poisons the blessing of liberty itself. It will be of little avail to the people that the laws are made by men of their own choice if the laws be so voluminous that they cannot be read, or so incoherent that they cannot be understood; if they be repealed or revised before they are promulgated, or undergo such incessant changes that no man who knows what the law is to-day can guess what it will be to-morrow. Law is defined to be a rule of action; but how can that be a rule which is little known and less fixed?

The Federalist No. 62, at 415 (P. Ford ed. 1898) (A. Hamilton).

Fundamentally, the argument for the legislative veto as a means of oversight comes down to a belief that the existence of powerful and unaccountable rulemaking bodies demands that Congress have an efficient and effective means of control, i.e., that far from contravening the separation of powers doctrine, the one-house veto helps implement it. The empirical evidence suggests that legislative vetoes are not all that efficient in practice, however, and this case presents an example of extreme disruption of the normal administrative process. FERC claims that it did not exercise its policymaking discretion because it believed that Congress wanted only a proposal to evaluate. Yet Congress appears to have delegated the Phase II issue in large part precisely to make use of FERC's policy expertise. The result was the agency assuming the bizarre position that it failed to exercise reasoned decisionmaking when it issued the rule.

See, e.g., Bruff Gellhorn, supra note 210, at 1432-33 ("One obvious but important practical effect of the combination of congressional and judicial review is the increased potential for impasse.... Legislative vetoes, by placing Congress and the courts in similar roles, make possible a three-sided interaction, with heightened potential for delay, disruption, and unforeseeable results."); id. at 1438-39 ("To the extent that legislative veto authority allows special interests to achieve their ends by pressuring administrative agencies through congressional committees, such authority works to promote, rather than negate, faction.").

When the agency has difficulty in determining its mandate, the value of the administrative process, with its guarantee of equal opportunity to participate by all interested parties, diminishes sharply. The agency may tend to ignore parties' submissions since it knows Congress ultimately will review the rule prior to its implementation. As a result, the attempt to enhance the political accountability of agency rulemakers may actually backfire as administrators view their roles as proposers of appropriate policy rather than final decisionmakers who must listen to and attempt to reconcile competing interests. Cf. Chadha, 634 F.2d at 431 n. 32 (emphasizing "one danger inherent in departures from the rule of separation of powers: when a governmental power is exercised by a branch other than that ordinarily responsible, the specific guarantees of governmental regularity applicable to the ordinarily responsible branch are in effect short-circuited."); Bruff Gellhorn, supra note 210, at 1438 ("The congressional review process adds a second stage to rulemaking, one in which not all interested persons now participate, and in which not all interests receive equal attention.").
The one-house veto also increases sharply the negotiating power of individual congressmen or committees in seeking to direct administrative outcomes. The threat of a veto is much more plausible than the threat of a statutory amendment, making administrators more likely to succumb to congressional interference. Thus whatever the legal status of informal congressional interference with administrative decisionmaking, see D.C. Federation of Civic Ass'ns v. Volpe, 459 F.2d 1231, 1248-49 (D.C. Cir. 1971) (administrative exercise of discretion must be based on merits in accord with the statutory mandate, not on extraneous congressional pressure), cert. denied, 405 U.S. 1030, 92 S.Ct. 1290, 31 L.Ed.2d 489 (1972), the one-house veto is significantly worse in this respect. See Clark v. Valeo, 559 F.2d at 682 (MacKinnon, J., dissenting).

Moreover, if the legislative veto is an especially effective means of congressional control of administrative decisionmaking, then the separation of powers problem looms even larger. In arguing for the legislative veto, former Senator Javits wrote:

In most situations, even when a clear consensus is present, it takes a long time for Congress to work its legislative will, owing, in large measure, to the various forms a single, affirmative, legislative remedy may take. A simple and unamendable resolution of approval or disapproval adopted pursuant to a legislative veto provision incorporated in an earlier statute, however, avoids the institutional delays and permits expedited postenactment review. In this characteristic, the legislative veto stands alone among congressional oversight techniques.

Javits Klein, supra note 167, at 462.

This argument seems correct, but it fails to explain the need for a one-house veto. The delays in enacting an affirmative and amendable legislative remedy are a result of nonconstitutional congressional rules of procedure which can be changed by Congress acting alone. If Congress chooses to use an unamendable resolution of disapproval as a means of expediting action, it may do so, if it acts by both houses and presents the resolution to the President. This would add little delay so long as there is a "clear consensus." To the extent there is not a consensus, the failure to act is not an undesirable "delay" but rather exactly the outcome of the legislative process envisioned by the Framers. The bicameralism and presentation requirements in Article I, Section 7 are not unfortunate by-products of a poorly designed scheme but rather carefully constructed impediments to the Legislature's exercise of power.

See Myers v. United States, 272 U.S. 52, 293, 47 S.Ct. 21, 84, 71 L.Ed. 160 (1926) (Brandeis, J., dissenting), cited at p. 471 supra. See also Gewirtz, supra note 82, at 59 ("If Congress cannot agree on what it wants to accomplish, and cannot hammer out an agreement on the basic choices involved in an area, then there probably should be no law.").

The fundamental problem of the one-house veto, then, is that it represents an attempt by Congress to retain direct control over delegated administrative power. Congress may provide detailed rules of conduct to be administered without discretion by administrative officers, or it may provide broad policy guidance and leave the details to be filled in by administrative officers exercising substantial discretion. It may not, however, insert one of its houses as an effective administrative decisionmaker.

See Rehnquist, Committee Veto: Fifty Years of Sparring between the Executive and the Legislature 7 (12 Aug. 1969) (remarks of Assistant Attorney General before Administrative Law Section of American Bar Association) ("[T]he vice is that Congress not having chosen to [designate specific terms], and having delegated its authority to the Executive, it may not then control Executive exercise of this delegated power").

One final argument in favor of the legislative veto is that this increase in congressional power is necessary if "unbridled executive discretion is to be avoided, and the potential abuses of unfettered presidential power contained." On this view, courts must recognize "that with the rise of the `imperial presidency' some alternative must be devised if accountability on executive power is to be effected." We refuse to accept the view that as a matter of constitutional law the Executive is too powerful and the Congress too weak. The great irony in this view is that there has been no usurpation of rulemaking power by either the President or unelected independent administrators. Their powers have come solely from congressional authorizations. If Congress has given away too much power, it may by statute take it back or may in the future enact more specific delegations. It is one thing to agree that "[d]elegation of lawmaking power is a categorical imperative of modern government," and quite another to conclude that because Congress fails to constrain its delegations sufficiently to produce accountability, it may therefore insert itself into the administrative process. The power to cure the perceived problem lies entirely within congressional control both before it delegates power at all and after the administrators exercise their discretion.

Javits Klein, supra note 167, at 497.

Miller Knapp, supra note 191, at 387.

Congress also has numerous informal methods of overseeing agency action. See generally Kaiser, Congressional Action to Overturn Agency Rules: Alternatives to the "Legislative Veto", 32 Ad.L.Rev. 667 (1980).

Schwartz, supra note 182, at 353. It has been argued that "consistent application of the view that the legislature cannot be empowered to disapprove a rule because disapproval is the exercise of a power to make laws would destroy the rulemaking powers of executive and administrative agencies themselves." Id. at 374. This is incorrect because it views the invalidity of the legislative veto as based on a "rigid separation-of-powers approach," id., when in fact the crucial deficiency is not that Congress is engaged in lawmaking but that it does so by one house and without presentation to the President. And the problem with the one-house veto under the separation of powers doctrine is not that there is some intermingling of powers but that Congress substantially increases its total power by assuming roles that make the work of the other branches subordinate or irrelevant. Delegation of rulemaking power does not present this difficulty because the rulemaking entity is always subject to both the Judiciary and the Congress. If in practice the agencies have assumed a power that seems disproportionately large, then it may be time for the courts and Congress to reassert their unquestioned powers, respectively, to strike down undue delegations and to provide more specific delegations.

Congressional unwillingness to use its constitutional powers cannot be deemed a sufficient reason for inventing new ways for it to act. As Justice Douglas wrote, "Legislative action may indeed often be cumbersome, time-consuming, and apparently inefficient," but that is not reason for interpreting the Constitution as conferring legislative authority on the President. Similarly, the inefficiency of the Article I lawmaking process is no excuse for interpreting the Constitution to permit legislative participation in the administrative process. We indeed "pay a price for our system of checks and balances, . . . a price that today may seem exorbitant to many." But if change is necessary, it must come either from a congressional reassertion of its right, and indeed responsibility, to provide meaningful standards for administrative action or from an amendment to the Constitution. The Constitution, with its principle of separation of powers, may be flexible, but it may not be changed by a judicial determination that its established procedures are no longer adequate.

Youngstown Sheet Tube Co. v. Sawyer, 343 U.S. 579, 629, 72 S.Ct. 863, 886, 96 L.Ed. 1153 (1952) (Douglas, J., concurring).

Id. at 633, 72 S.Ct. at 888.

3. Intrusion into the judicial sphere

A final use of a legislative veto is to prevent the effectiveness of agency actions that exceed the statutory mandate. This was not a primary justification of the Phase II veto, but both the House's veto report and certain congressmen urged approval of the veto resolution on the ground that FERC's rule violated the congressional intent underlying section 202. Petitioners urge that to the extent the veto was based on this ground, it unconstitutionally intruded upon the exercise of judicial powers by the courts. We agree.

See, e.g., Veto Report supra note 9, at 11 ("The Commission's narrow interpretation of its authority under section 204(e) to set incremental pricing alternative fuel ceilings solely on the basis of installed alternative fuel capability and its perceived inability to tie delivered natural gas prices to natural gas wellhead market conditions appears to be a misreading of the statute."); 126 Cong.Rec. H3847 (daily ed. 20 May 1980) (remarks of Rep. Markey) ("It is the clear intention of Congress in the incremental pricing provisions of the Natural Gas Policy Act to shield residential gas users and other high-priority customers from rapidly escalating gas prices for as long as possible. The rule before us today falls seriously short of this mandate."); id. at H3849 (remarks of Rep. Stockman) ("Clearly we now have before us a regulation, a proposed rule or program that was not intended, even by the authors of this provision when it was adopted in 1978,"); id. at H3854 (remarks of Rep. Ashley) ("[T]he thrust of the rule runs counter to the legislative history of the NGPA and to the intent of Congress.").
There is no doubt that legislative vetoes sometimes are based on a finding of violation of statutory intent. See, e.g., Chadha, 634 F.2d at 430 n. 31 ("It appears that the House in this instance did view its action as a correction of a misapplication of the statutory criteria governing eligibility for exercise of the Attorney General's discretion."). Indeed, Congress has provided in some statutes that a rule may be vetoed only on this ground. See 20 U.S.C. § 1232(d)(1) (1976 Supp. III 1979) (regulations issued by Secretary of Education may be vetoed by concurrent resolution only if "the final regulation is inconsistent with the Act from which it derives its authority"). But cf. Bruff Gellhorn, supra note 210, at 1429 ("Although congressional review is often ostensibly based on legality, consideration of policy in a legislative body is inevitable.").

The function of courts in reviewing agency action is to interpret the statutory delegation and determine whether the administrative decision is in compliance with that delegation. The problems that arise when one house of Congress assumes this role are clear. If one house vetoes a rule, the courts are prevented from exercising review, even though under prior decisions on the same statute, or on analogous statutes, they might have upheld the agency's exercise of discretion. And if the rule is not vetoed, the courts are presented with a difficult question of how much weight, if any, to give to the implicit congressional finding that the rule represents a proper exercise of statutory discretion.

As the Chadha court observed, the effect may be to render judicial interpretations of the statute mere advisory opinions. 634 F.2d at 430 n. 31.

In some statutes, though not in section 202(c) of the NGPA, Congress has attempted to eliminate this problem by providing that failure to veto a final regulation "shall not represent . . . an approval or finding of consistency with the Act from which it derives its authority for any purpose, nor shall such failure to adopt a concurrent resolution be construed as evidence of an approval or findings of consistency necessary to establish a prima facie case, or an inference or presumption, in any judicial proceeding." 20 U.S.C. § 1232(d)(1) (1976 Supp. III 1979) (Department of Education regulations). See also 15 U.S.C.A. § 57a-1(d) (West Supp. 1981) (similar provision for failures to disapprove FTC rules). It is not at all clear, however, why failure to veto is not some evidence of compliance with congressional intent, if a decision to veto becomes a conclusive determination of a violation of congressional intent. In essence, Congress is conceding that its review cannot be systematic or thorough. Apparently it will veto only where it considers the statutory violation abundantly clear. Yet those cases are the ones where the courts should have the least difficulty making the same finding. This demonstrates a peculiarity inherent in the "statutory intent" justification for the legislative veto: it is unclear why it is necessary at all, since the courts are presently performing the precise function Congress now seeks to perform. See Chadha, 634 F.2d at 431 ("There is no indication that the Judiciary is incapable of determining whether the Attorney General has abused his discretion or applied a statutory standard improperly."). It is no answer to argue, as one former senator has, that the congressional veto is necessary to enable Congress to avoid "that institutional weakening which would result from constantly relying on the judicial process to protect its prerogatives." Abourezk, supra note 182, at 333. The limitation on Congress that results from restricting to the Judiciary the power to strike down executive actions is not something to be mourned, but rather is the very essence of separation of powers.

Either way, the congressional review is entirely standardless and may be conducted without equal participation of interested parties. If there is a dispute between the houses as to the statutory intent, the rule is defeated, and the interested parties are prevented from gaining a judicial interpretation. Moreover, the reviewing body — which in practice may well be a congressional committee — will inevitably look not primarily to the objective legislative intent at the time the statute was enacted, but rather to the "intent" at the present. This permits Congress effectively to alter the meaning of a statute as circumstances and the composition of Congress change over time. Assumption of this power diminishes the role of the Judiciary and expands that of Congress. Accordingly, it violates the separation of powers doctrine.

This problem is clearly more severe in cases adjudicating rights of particular individuals, see Chadha, 634 F.2d at 430-31, 435, but applies to rulemaking as well. See Bruff Gellhorn, supra note 210, at 1438; McGowan, supra note 82, at 1148-49.

See Bruff Gellhorn, supra note 210, at 1429-30 ("[W]hen Congress is unclear initially in forming legislative history, it is unlikely to contribute more than the agencies or the courts in later attempts to reconstruct it. Furthermore, the validity of a legislative purpose that is purportedly part of the law but not reflected in ordinary sources of legislative history diminishes as time passes and a new Congress with new members convenes."); Watson, supra note 118, at 1067.

F. Conclusion

We are aware that our decision today may have far-reaching effects on the operation of the National Government. Yet this cannot deter us from finding the one-house veto unconstitutional. Congressional amici would have us, under the principles of flexibility and practicality in constitutional adjudication, approve an institutional structure whereby the administrative "experts" make policy and the people's representatives (without the President) are reduced to exercising a negative. This contravenes the constitutional procedures for making law. The genius of our Constitution, its adaptability to changes in the nature of American society, depends ultimately on the steadfastness with which its basic principles and requirements are observed. Otherwise its critical protections against governmental tyranny would quickly become meaningless, as the Government in power could shape it to suit whatever purposes seem sound at the present. The Article I restrictions on the exercise of the legislative power, as well as the principle of separation of powers, are fundamental to the constitutional scheme, and because section 202(c) attempts to evade them it cannot stand.


Intervenors urge that in event we declare section 202(c) unconstitutional, the Phase II rule issued by FERC should not be ordered to go into effect. This is based on the view, argued by FERC itself, that the rule was not a product of reasoned decisionmaking and that FERC might well come up with a different rule after taking a close look at the situation. We have held, however, that FERC's attempted revocation of the rule, based on the alleged failure to engage in reasoned decisionmaking, was invalid because FERC did not provide notice and an opportunity to comment. The rule therefore shall take effect. To give the Commission an opportunity to reconsider the rule, however, if it decides this is necessary or if any party requests it to do so, we delay the effectiveness of the rule for thirty days from the date of our decision. During this period the Commission may consider whether to amend the substance of the rule or, in order to provide time for sufficient consideration, whether to postpone the effective date further. If the proceedings are reopened, petitioners and all other interested parties must be provided with notice and an opportunity to comment on all issues raised. We express no views as to the precise scope of the Phase II delegation or the extent, if any, of the Commission's power to delay, amend, or revoke the rule.

See pp. 445-448 supra.

Cf. Buckley, 424 U.S. at 143, 96 S.Ct. at 693 (providing 30-day stay of Court's judgment to "afford Congress an opportunity to reconstitute" the unconstitutional Federal Election Commission).

So Ordered.