dissenting:
Section 107(b) of the Natural Gas Policy Act of 1978, 15 U.S.C. § 3301 et seq. (1988) (“NGPA”), authorizes the Federal Energy Regulatory Commission (“FERC”) to set ceiling prices for high-cost natural gas in excess of the otherwise applicable maximum price “to the extent that such special price is necessary to provide reasonable incentives” for high-cost gas production. 15 U.S.C. § 3317(b) (1988). In Williams Natural Gas Co. v. FERC, 872 F.2d 438 (D.C.Cir.1989) (“Williams 7”), we recognized that, in order to meet these statutory criteria, an incentive price must both be “necessary to induce production (as evidenced by the ... [gas producer’s and purchaser’s] contractual agreement [to employ the incentive price]) and ... provide a reasonable incentive (as measured by prices for alternative fuels).” Id. at 448 (emphasis in original). We further noted that, in its 1983 Notice of Proposed Rule-making,1 FERC had expressed the “tentative conclusion that the incentive price established by Order No. 99[2] was no longer reasonable in comparison to the costs of alternative fuels,” a conclusion this court viewed as “tantamount to saying that the tight formation price was no longer authorized by the statute.” Id. at 447. Accordingly, we directed FERC to reconsider its decision to retain its incentive price for tight formation gas notwithstanding its own considerable doubts as to the continued propriety of that price.
On the record now before us, I do not believe that FERC’s orders in response to this court’s remand in Williams 73 afford a reasoned explanation for the agency’s ultimate decision not to reduce the incentive price for all tight formation wells spudded, recompleted or reworked after the NOPR’s publication in the Federal Register on February 22, 1983 (“post-NOPR wells”). Accordingly, I dissent from the majority’s decision to the extent that it affirms (i) FERC’s retention of its incentive price on a prospective basis for all post-NOPR wells spudded, recompleted or reworked on or before May 12, 1990, and (ii) the agency’s denial of refunds to pipelines for the above-market-rate payments they have already made for tight formation gas from postNOPR wells.
I.
In enacting the NGPA, Congress never intended that the regulated ceiling prices of gas would exceed market-clearing rates; as the Supreme Court has observed, the NGPA was intended to create price ceilings, not price supports, for gas producers. See FERC v. Martin Exploration Management Co., 486 U.S. 204, 210, 108 S.Ct. 1765, 1769, 100 L.Ed.2d 238 (1988) (“Not one participant in the legislative process suggested that producers should receive higher prices than deregulation would afford *397them.”); Williams I, 872 F.2d at 440-41 (“the Congress which passed the [NGPA] ... quite plainly supposed that the statutory ceiling prices would be lower than those which would obtain in an unregulated market”).
Nor did FERC envision above-market incentive ceilings when, in Order No. 99, it authorized producers to charge 200% of the otherwise applicable statutory ceiling rate for tight formation gas. Issued “against a backdrop of rising oil prices ... and relatively little experience with the effects of the NGPA on gas supplies and pricing,”4 Order No. 99 was the agency’s attempt to free high-cost gas from some, but not all, of the price restraints of regulation, in response to commenters’ claims that the regulated price for tight formation gas needed to be “at least ... comparable to the free market value of the gas” in order to encourage high-cost gas production.5
FERC recognized in Order No. 99, however, that “consumers should not be required to pay a price higher than the otherwise applicable NGPA price unless there is a reasonable basis for assuming that a higher price is necessary,” and emphasized that its incentive price “must not create perverse incentives.” 6 Because Congress had intended the NGPA to provide “a pricing scheme that related incentive prices to the price of competing fuels,”7 FERC looked to the cost of fuel oil in establishing the upper range for its tight formation incentive price.
When the bottom unexpectedly dropped out of the oil market two years later, FERC realized the need to reexamine its incentive price. Although the demand for natural gas had plummeted along with oil prices, prices for gas had only reluctantly responded to this decreased demand, due, FERC noted in its NOPR, “both to the pricing mechanisms built into the NGPA and the contracting practices of pipelines and producers.” 8 FERC recognized that, as the agency setting the ceiling price for tight formation gas, it bore some responsibility for the NGPA’s pricing mechanisms, and it therefore set out to consider “whether its own decisions in setting incentive ceilings under NGPA section 107(c)(5) ... [were] exacerbating the current gas pricing problems.”9 Tentatively concluding that a ceiling price for tight formation gas in excess of the commodity value of such gas was “contrary to the public interest,” FERC proposed to cap the tight formation incentive price at an imputed commodity value as measured by the price of competing fuels.10
The NOPR was therefore a reasonable proposal intended to remedy a situation that neither Congress nor FERC had ever contemplated — FERC-sanctioned incentive prices for high-cost gas significantly in excess of the prices that such gas could command in an unregulated market.
II.
The principal concern of the NOPR was that above-market incentive prices were “exacerbating ... pricing problems” in the gas industry. Although FERC now purports to repudiate the NOPR’s tentative conclusions, its most recent orders contain nothing to suggest that this basic concern was ill-founded.
FERC does not deny that its incentive price for tight formation gas has remained substantially higher than the market-clearing rate since 1983. Nor does FERC make *398any effort to explain how an above-market ceiling price could provide “reasonable incentives” for the production of high-cost gas in a period during which the demand for such gas has been in sharp decline.
Instead, FERC essentially sidesteps the question of the reasonableness of its above-market incentive prices by downplaying the harm that such prices have caused. Thus, FERC claims that pricing problems experienced in the gas industry in the early 1980s, rather than being attributable to the tight formation ceiling price, were an “inevitable part” of the NGPA’s transition from a regulated to a largely deregulated pricing environment.11 This observation, however, is simply beside the point. The NOPR did not suggest that all, or even most, natural gas pricing problems were caused by the tight formation incentive price; indeed, given that high-cost gas represents but a small fraction of the total supply of natural gas, such a claim would be patently absurd. FERC merely reasoned that its tight formation incentive price may have been exacerbating those pricing problems that already existed. This was no doubt true, given that the presence of an incentive price 200% above the otherwise applicable price ceiling maintained strong incentives for the production of high-cost gas when many of the reasons for those incentives had ceased to exist. In short, the fact that many pricing problems under the NGPA may have been “inevitable” hardly justifies the FERC’s refusal to remedy those problems within its regulatory control.
FERC further implies that any effects of above-market incentive prices on pipelines and consumers were de minimis. As an initial matter, it is doubtful whether FERC could justify retention of otherwise unlawful prices on this basis alone, bearing in mind our admonition in Williams I that an agency’s refusal to rectify a statutory violation on the grounds that its import is de minimis “must surely be narrowly circumscribed.” 872 F.2d at 447.
Even on its own terms, however, FERC’s suggestion of de minimis effects is utterly unconvincing. FERC notes that, despite the above-market ceiling rate, the prices actually obtained for tight formation gas had declined by 1988 to near-market levels “due to price-adjustment provisions, market-outs, and renegotiation.”12 In Williams I, however, we found that the ability of pipelines and producers to adjust and renegotiate their contracts did not justify FERC’s retention of an unjustified incentive price. We noted that
when producers and pipelines renegotiate contracts for tight formation gas, the fact that these contracts reference the current high incentive price will give producers additional leverage, allowing them to extract concessions that would otherwise be unavailable. To put it another way: ... [the availability of renegotiation] may enable pipelines to buy their way out of uneconomical contracts, but the existence of the high incentive price will force them to pay more for the privilege.
872 F.2d at 449-50 (emphasis in original). Indeed, FERC’s contention that renegotiation and other market mechanisms have mooted the incentive price problem is rather startling given FERC’s frank admission elsewhere in Order No. 519 that its tight formation incentive price has “served (and will continue to serve) as a basis for settlements between high-cost gas producers and purchasers seeking to adjust their contractual arrangements to changing market conditions.”13 Thus, like the panel in Williams I, I would reject FERC’s suggestion that the problem of an unreasonable ceiling price is “no longer of any consequence simply because the pipelines, by renegotiating their long-term contracts, may [have] be[en] able to avoid some of the costs which the incentive price would otherwise [have] impose[d].” 872 F.2d at 450 (emphasis in original).
Moreover, the fact that, by 1988, many of the negative effects of FERC’s above-mar*399ket ceiling prices may have been ameliorated by industry adjustments to changed market conditions is irrelevant to the separate question of whether FERC itself had a duty to award pipelines retroactive relief for above-market payments they had been making since 1983. Because of the gas industry’s long-term “take-or-pay” contracts, pipelines and producers were able to rectify the anomaly of above-market incentives only over the course of several years; the NOPR proposal, on the other hand, would have eliminated the problem virtually at its inception. As we noted in Williams I, “pipelines were surely entitled to assume that the Commission would continue to monitor the ceiling price to ensure that it remained reasonable in light of subsequent developments.” Id. at 448. Instead, by its own admission, FERC abdicated this duty after issuing the NOPR and relied upon the gas industry, in time, to take care of the problem itself. We pointed out in Williams I, however, that such deference to the private sector is fundamentally inconsistent with the NGPA, which “can hardly be seen as reflecting unequivocal enthusiasm for market forces as price-setting mechanisms.” Id.
Perhaps the clearest refutation of FERC’s implication that the adverse effects of above-market incentive prices are de minimis is FERC’s own reference to substantial producer reliance interests as a justification for not applying its incentive price repeal to all post-NOPR wells. FERC notes that retroactive application of its price repeal to such wells would require gas producers to refund some of the profits they have earned from tight formation gas since 1983, while repealing the price prospectively for such wells would limit these producers’ future profits. But the agency simply cannot have it both ways. If the incentive price has created substantial reliance interests for sellers of gas in the form of sizable above-market-level profits (both past and future), then FERC cannot reasonably contend that the effects of that price on buyers of gas have been de minimis.
As we noted in Williams I, while pipelines may properly be held to the consequences of their own bad bargains, “it is quite another matter if the contracts are unfavorable because they incorporate an incentive price which is no longer consonant with the statute.” Id. at 450. After itself calling its incentive price into serious question, FERC has failed to make any showing that the retention of the incentive price after 1983 was “reasonable” as required by the NGPA. Accordingly, I discern no legitimate basis for FERC’s disavowal of its conclusion in the 1983 NOPR that the ceiling for tight formation gas prices had ceased to be in the public interest as of that time.
III.
Because I believe that FERC has never effectively rebutted its tentative conclusion in the NOPR that its above-market incentive price was no longer “reasonable” as of 1983, I next consider whether the agency had some other justification for not repealing its incentive price as of the publication date of the NOPR. In Order No. 519, FERC stated that application of its price repeal either retroactively or prospectively to post-NOPR wells would have unfairly frustrated the reliance interests of producers. FERC indicated that it did not “wish to take any action to undo existing settlements [based upon the former ceiling rate], to jeopardize future settlements, or to frustrate federal tax incentives adopted by Congress” and claimed that “[i]t would not be fair or in the public interest to expect producers to halt on-going gas development projects on the basis of only a proposal” because producers reasonably could “have concluded that the specific application dates of the proposed rule might never become law.” 14
On the record at hand, there is no reasonable way to justify FERC’s action on the basis of producer reliance interests. As FERC itself admits, an “incentive ceiling price should be removed at whatever point *400in time it ceases to be justified.” 15 FERC has never persuasively shown that its above-market incentive price was reasonable after 1983, which, as we noted in Williams I, is tantamount to a concession that the incentive price was unlawful from that time onward. While justifiable reliance interests are an appropriate consideration when weighing the proper scope of agency action, parties “cannot claim justifiable reliance or protectable expectations based on ... [agency] action which was illegal.” Tennessee Valley Mun. Gas Ass’n v. FPC, 470 F.2d 446, 453 (D.C.Cir. 1972). Absent any showing that the incentive price met the criteria of NGPA section 107(b) after 1983, the agency’s decision nonetheless to retain that price must be viewed as violative of the governing statute. Producers should have recognized that the incentive price was at all times subject to the Act’s “reasonableness” requirement, and that FERC was statutorily obligated “to monitor the ceiling price to ensure that it remained reasonable in light of subsequent developments.” Williams I, 872 F.2d at 448.
Any doubts producers may have had on this score should have evaporated upon publication of the NOPR, which expressly questioned the continued legitimacy of the ceiling price and proposed to reduce the incentive price to market levels for all new tight formation wells. Producers who invested in tight formation gas after the issuance of the NOPR thus took a calculated risk — although the NOPR contained only FERC’s tentative conclusions, it expressly notified such producers that any rule ultimately adopted by FERC might rectify, as of February 22, 1983, the incongruity of above-market incentive prices. Any producers who forged ahead with tight formation investments on the chance that FERC would retain an incentive price that bore no rational relationship to market realities simply cannot lay claim to “reasonable” reliance interests.
The tenuous status of FERC’s incentive price was trumpeted even more loudly by our decision in Williams I, which delineated in some detail the dubious legality of above-market NGPA price ceilings. Yet, FERC has left its above-market incentive price intact even for those producers who developed tight formation wells after the issuance of Williams I. The practical effect of the agency’s decision is to allow private parties, despite receiving unequivocal notice that an existing rule may be unlawful, to exploit the rule with complete impunity simply by relying upon it before the agency has had an opportunity to take final corrective action. Such a definition of “reasonable reliance” interests, in my view, cannot help but create perverse and mischievous incentives for private party conduct.16
In Williams I, we held that FERC is obligated to adjust the incentive price periodically to reflect changes in the gas market. Apparently due solely to its own regulatory inertia following the issuance of the NOPR, the agency failed to carry out that obligation in this case. As I do not believe that FERC has ever provided an adequate explanation for its decision to accord neither retroactive nor prospective relief to pipelines and consumers charged above-market incentive prices for gas from *401post-NOPR wells, I would grant the petition for review and remand the case to FERC.
. Limitation on Incentive Prices for High-Cost Gas to Commodity Values, 48 Fed.Reg. 7469 (1983) [hereinafter NOPR].
2. Regulations Covering High-Cost Natural Gas Produced from Tight Formations, 45 Fed.Reg. 56,034 (1980) [hereinafter Order No. 99].
.Limitation on Incentive Prices for High-Cost Gas to Commodity Values, 55 Fed.Reg. 6367 (1990) [hereinafter, Order No. 519]; Limitation on Incentive Prices for High-Cost Gas to Commodity Values; Order Denying Rehearing, 55 Fed.Reg. 18,100 (1990) [hereinafter, Order No. 519-A].
. NOPR, supra note 1, at 7470.
. Order No. 99, supra note 2, at 56,037.
. Id. (emphasis added).
. Id. at 56,038.
. NOPR, supra note 1, at 7470. As we noted in Williams I, gas contracts between producers and pipelines "typically span very long periods of time, and the price is often contractually tied to the price ceiling established by the statute or regulations.” 872 F.2d at 441. Gas prices were thus relatively unresponsive in the short term to/ the extreme fluctuations in the worldwide energy market in the early 1980s, as "take-or-pay" contracts obligated many pipelines to pay the prevailing regulatory ceiling price, which, despite the changed market, remained the 200% incentive price established in Order No. 99.
. NOPR, supra note 1, at 7470.
. Id.
. Order No. 519, supra note 3, at 6375.
. Order No. 519, supra note 3, at 6375.
.Order No. 519, supra note 3, at 6373 (emphasis added).
. Order No. 519, supra note 3, at 6373-74.
. Order No. 519-A, supra note 3, at 18,105.
. Even were I to credit the reliance interests cited by FERC, I would view its blanket refusal to broaden the effect of its price repeal as arbitrary and capricious. Petitioner Williams Natural Gas Company ("Williams") had suggested less drastic means of protecting reliance interests, proposing, e.g., that the incentive price could be eliminated for future gas produced from wells funded by a pipeline or its customers, rather than by a producer, since no producer reliance interests would be implicated in such price reductions. For those wells which had been funded by producers, Williams suggested that retention of the incentive price should be permitted only until the costs of the well had been recovered. FERC did not specifically address these alternatives, merely stating in conclusory fashion that producers “generally" had legitimate reliance interests and that it would be "difficult and fruitless ... to determine on a case-by-case basis the adverse financial impact on a specific producer of changing the incentive ceiling price for all flowing gas.” Order No. 519, supra note 3, at 6374 n. 62. Given the apparent illegality of the incentive ceiling price at issue, this is hardly a satisfactory response.