Opinion for the Court filed by Circuit Judge STEPHEN F. WILLIAMS.
The Trans Alaska Pipeline System (“TAPS”) is a 48-inch diameter pipeline carrying crude oil from Alaska’s North Slope approximately 800 miles south to Valdez, Alaska. Each shipper delivers its own crude oil to the pipeline, in which the oils are commingled; at the terminus the shipper takes delivery of a proportional share of the common stream. The crude oils delivered initially differ from each other in various characteristics that affect market value. Beсause of the commingling, a shipper will not in all likelihood receive the same quality of oil at Valdez that it delivered to the pipeline. Without some adjustment, the ones delivering relatively higher-value crudes would unfairly lose, and the ones delivering lower-value crudes would unfairly gain. The parties here battle over the formula governing the adjustment, which the Federal Energy Regulatory Commission controls in the exercise of its authority to regulate interstate oil pipeline rates. 1
*1288
Exxon Company, U.S.A.
2
and Tesoro Alaska Petroleum Company filed complaints with the Federal Energy Regulatory Commission assailing aspects of the prevailing formula. Exxon challenges the formula itself, a so-called “distillation” methodology that the Commission adopted in 1993 and later modified in 1997; Tesoro contests the specific valuation of two “cuts” of petroleum,- West Coast naphtha and West Coast vacuum gas oil (“VGO”). A rate order must be modified where “new evidence warrants the change.”
Tagg Bros. & Moorhead v. United States,
In 1984 the Commission approved a settlement agreement establishing a “Quality Bank” to make the required adjustments between shippers. See
Trans Alaska Pipeline System,
Starting in 1987, the amount of natural gas liquids (“NGLs”) in the stream increased, changing the picture — or at least the perception. Two factors contributed to this increase. First, natural gas operations expanded in Prudhoe Bay, resulting in sharply increased deliveries of NGLs at the head of the pipeline.
OXY USA Inc. v. FERC,
Responding to the resulting complaints under § 13(2) of the Interstate Commerce Act, the Commission in 1989 started to
*1289
investigate the gravity method. It found that the method was no longer just and reasonable and, in approving a contested settlement in 1993, adopted the distillation method. See
For some cuts there were acceptable indicators of market value from the Oil Price Information Service (“OPIS”) or Platt’s Oilgram. No such markers were available, however, for distillate, VGO or resid, or for West Coast naphtha. For these cuts the settlement proposed to use prices for kindred products, adjusted for differences between them and the actual cuts. The Commission rejected this approach, saying that for a system to be nondiscriminatory it must use “market prices, uncomplicated by subjective adjustments.”
Id.
at 62,289. As part of this “No Adjustment Policy,” the Commission rejected the proposed use of adjusted West Coast prices to value the West Coast naphtha cut and instead set a Gulf Coast price for the cut. On rehearing, it also ordered the use of Gulf Coast prices for West Coast deliveries of VGO.
Tesoro Alaska Petroleum Co. v. Amerada Hess Pipeline Corp.,
In 1996, while the
OXY
remand was under way, Exxon filed a complaint against seven TAPS owners pursuant to §§ 9, 13(1) and 15(1) of the Interstate Commerce Act, 49 U.S.CApp. §§ 9,13(1), 15(1) (1988) — leading to the present case. Upholding an ALJ decision, the Commission dismissed the complaint, holding that Exxon had failed to produce evidence of changed circumstances to justify re-examination of the 1993 adoption of the distillation method.
Exxon Decision,
Tesoro participated in the proceedings before the ALJ on Exxon’s complaint, raising issues that the ALJ ultimately identified as different from Exxon’s. The ALJ’s order of dismissal mooted Tesoro’s arguments but noted that Tesoro was free to file its own complaint.
Exxon Co., U.S.A. v. Amerada Hess Pipeline Corp.,
Petitioners argue that because their complaints were disposed of by Motion for Summary Disposition, our review is de novo. That would be true if we were reviewing a district court’s equivalent action. But these dismissals implicate the Commission’s expertise and policy-making authority, compelling deference.
Motor Vehicle Manufacturers Ass’n of the United States v. State Farm Mutual Auto. Ins. Co.,
* * *
In
Tagg Bros. & Moorhead v. United States,
The Commission acknowledges the authority of
Tagg Bros.,
but reframes Justice Brandeis’s formula — allowing re-opening for “new evidence” — into one requiring evidence of “changed circumstances.” It is unclear if any such limit may be imposed. In
OXY
itself we observed, “[t]hе fact that a rate was once found reasonable does not preclude a finding of unreasonableness in a subsequent proceeding.”
[A] subsequent modification of the significant facts or a change or development in the controlling legal principles may mаke that [judicial] determination obsolete or erroneous, at least for future purposes. If such a determination is then perpetuated each succeeding year as to the taxpayer involved in the original litigation, he is accorded a tax treatment different from that given to other taxpayers of the same class.
Commissioner of Internal Revenue v. Sunnen,
Exxon provided the testimony of Dr. Pavlovic, an economic consultant, who tested the accuracy of the modified distillation methodology for 34 crude oils in the California crude oil market from 1993 to 1996. Pavlovic used regression analysis to *1291 compare the relative values of the cuts produced by the distillation method with actual market prices. He claimed his tests showed that the distillation method “substantially over-values low-value, heavier petroleum and substantially undervalues high-value, lighter petroleum.” Joint Appendix (“J.A.”) at 430. He also testified that this bias “increase[d] dramatically in 1994 and remain[ed] large thereafter.” Id. at 451. A perfect pricing method would produce a coefficient of 1.0 in a regression of the method’s relative values on those of the benchmark market. Whereas the coefficient — also called a bias measure— was indeed just over 1.0 for 1993 (1.07), it jumped in 1994 to 1.65 and remained around 1.6 for the next two years. Id. at 495. Pavlovic argued that he could reject, at a statistically significant level, the hypothesis that the bias measure was 1.0 in 1994-1996. 4 Id. at 453.
Pavlovic’s testimony appears to constitute not only new evidence but changed circumstances as well. It shows that, for reasons not yet conclusively determined, the degree of bias resulting from the use of the distillation method rose from imperceptible to severe after 1993. The Commission’s answer — that it “consistently has refused to base its decisions on how the TAPS Quality Bank should operate based on regression analyses of West Coast or world crude values,”
Exxon Decision,
The Intervenors offer an explanation that may be part of what the Commission in fact had in mind: “Dr. Pavlovic’s analysis was like testing a methodology designed to value Alaskan apples by applying that methodology to a crate of California oranges.” Intervenors’ (BP Exploration (Alaska), Inc. et al.) Br. at 13 (“Interve-nors’ Br.”). This glib use of the old apples-oranges metaphor overlooks the problem confronting the Commission:. There simply are no market prices for the Alaskan crude oils delivered into TAPS. If there were, there would be little or no issue about inferring their relative values. To the extent that the California crudes are similar to the Alaskan crudes, Pavlo-vic’s technique seems to test the accuracy of the distillation method. Compare J.A. at 440-43.
Exxon also provided the testimony of Mr. Moore, an engineer, and Dr. Haus-man, an applied economist. Moore stressed that the gravity of ANS crude oil (consisting of the streams that enter at the start of the pipeline at Pumping Station #1 and the return stream from refineries along its path) had increased from about 28° API in 1992 to about 30° API in 1996. Id. at 323. In the face of the theory that gravity increases due to NGL blending had undermined the gravity formula, Moore tracks the seemingly close positive relationship between the fraction of NGLs and gravity оf this crude oil, and, critically, between the gravity and price of ANS crude (measured against West Texas Intermediate, a “marker crude” used as a reference for valuing other crudes). Id. at 323, 419, 421.
Exxon invokes the testimony of Moore not only to suggest that the premise of the 1993 abandonment of gravity was mistaken, but also to explain why one should expect occurrence of the sort of distortions that Pavlovic seemingly showed in the distillation method. Although there is an obvious link between crude oil values and petroleum prоduct prices, they do not move in lockstep. The same event may drive the prices of the two in different directions. If a refinery shuts down because of equipment problems, for instance, it is likely to raise the market price of the *1292 refined petroleum product, but may reduce demand for crude oil and lower its price. See id. at 342.
Hausman testified that regression results on price and gravity data from 1988 to 1997 showed that gravity was a significantly positive predictor of the price of ANS crude. Id. at 511-13. This evidеnce seems to show a positive correlation between (1) NGL blending, (2) gravity, and (3) value. Petitioners rely on the Moore and Hausman testimony to suggest that the distillation method seriously undervalues the NGL contribution to the TAPS stream. Aspects of Hausman’s testimony might pose questions about the line between evidence that is “new” and evidence of changed circumstances. After all, Hausman claims that his regression results apply to the entire period of his data — from 1988 to 1997 — indicating that the pre-1993 data also show the positive effect of gravity on ANS value. Overall, however, it certainly confirms the implication of the Pavlovic and Moore testimony that the distillation method is seriously flawed.
To dismiss the testimony of Moore and Hausman, the Commission’s decision invoked our deference in
OXY
to the ALJ’s conclusion that “the current straightline gravity basis for valuing crude oil does not assign an accurate value for NGLs,” and our statement, “It follows from this conclusion that shippers delivering oil with a high NGL content were either being overcompensated or undercompensated under the gravity methodology-”
Tesoro challenges the current pricing of the naphtha and VGO cuts based on Gulf Coast market prices under the modified distillation method. Tesoro wants the Quality Bank to value West Coast naphtha using a formula based on West Coast gasoline prices and West Coast VGO using OPIS’s quoted price for West Coast high sulfur VGO.
Tesoro argues that the Commission’s use of the published Gulf Coast price, rather than a formula based оn West Coast gasoline prices, significantly undervalues West Coast naphtha.
5
Its claim is based on three propositions. First, Tesoro presents evidence, and the Intervenors acknowledge, that Gulf Coast deliveries of ANS crude “have declined considerably from the somewhat less than 20% level that existed in 1993.” Intervenors’ Br. at 25; see also
Distillation Decision,
Second, the Commission has changed its outlook on the use of “adjusted” market prices. In
OXY
we held that the Commission’s “No Adjustment Policy,” which used market prices instead of formulas, lacked an adequate foundation with respect to the distillate and resid cuts.
Despite the Commission’s claims to the contrary, its decisiоn to use the Gulf Coast price for West Coast deliveries of naphtha appears to have been based at least in part on the now abandoned No Adjustment Policy. Indeed even its brief invokes the benefits of making no adjustments. See Respondents’ Br. at 13. And the Interve-nors admit that the Commission declined to adopt a formula based on West Coast gasoline prices for naphtha because it did not want to assume “the risk that the formulas the parties proposed would be manipulated in some fashion to favor one party over another.” Intervenors’ Br. at 26. Even on a narrow view of OXYs impact on the No Adjustment Policy, the changes in the Gulf Coast naphtha market improve the trade-off for using an adjusted price rather than one from the dwindling Gulf Coast market.
Third, Tesoro presents evidence that appears to show a large disparity between the Commission’s valuation and the true value of West Coast naphtha. According to Mr. Stancil, an engineer and energy consultant, the quoted Gulf Coast naphtha price in December 1996 ($26.38 per barrel) undervalued West Coast deliveries by $2.71 per barrel.
6
This alleged disparity dwarfs the ones that required remand in
OXY.
See
These three propositions may well be answerable by the Commission. But without an adequate Commission response, they at the least establish a prima facie case that new evidence warrants re-examination of how West Coast naphtha should be valued.
Tesorо also challenges the valuation of the VGO cut. In its original decision on the distillation method, the Commission found that illiquidity in the West Coast market required it to use Gulf Coast prices for West Coast VGO.
The Commission provides two responses. The first is plainly inadequate. The Commission says that it considered Tesо-ro’s evidence in 1998 in evaluating — and rejecting — a proposal to change the valuation of West Coast VGO because of a change in how Gulf Coast VGO prices were being reported.
Second, the Commission argues that in its decision to usе the West Coast NGL market another factor was in play — differences between the coasts’ petrochemical industries and consequently their demand for NGLs — that is apparently inapplicable to VGO. That much is true. In approving a switch to reliance on the West Coast market for West Coast NGLs in 1998, it reasoned that the higher demand for NGLs on the Gulf Coast made it a less reliable standard for West Coast NGLs than the West Coast market, despite the latter’s illiquidity.
To this the Commission’s reply is that it has no “generic standards for liquidity.” Respondents’ Br. at 67. Thus the 1998 decision on NGLs could not have reflected a change in such standards. But the answer is illogical. Clearly the Commission regards illiquidity as a factor weighing against the use of a possible benchmark. When it changed its position in 1998 to allow use of the illiquid West Coast market it invoked disparities between the Gulf and West Coast markets, but it never asserted that there has been a change in those disparities. Accordingly, Tesoro’s inference that the NGL change reflects a reduced Commission anxiety about illiquidity is hardly unfair. Of course it may be that for VGO the Commission might reasonably find the balance to tilt more strongly in favor of Gulf Coast prices, so that Tesoro’s claim will ultimately lose. But Tesoro has shown enough to get its foot in the door, entitling it to an attempt to persuade the Commission to make a different trade-off on VGO.
Whether or not Exxon’s ultimate goal is to resurrect the gravity method in some form (a matter the parties dispute), its evidence at least suggests changed circumstances regarding the reasonableness of the distillation methodology. Tesoro has also supported its prima facie case about the valuation of the naphtha and VGO cuts. The Commission’s failure to respond meaningfully to the evidence renders its decisions arbitrary and capricious. Unless an agency answers objections that on their face appear legitimate, its decision can hardly be said to be reasoned.
International Harvester Co. v. Ruckelshaus,
Exxon in a footnote arguably challenges the Commission’s observation that “changes in the quality bank would be on a prospective basis, and no damages awarded.”
Exxon Decision,
*1295 * * *
The absence of a reasoned Commission explanation requires us to reverse and remand the case for further proceedings.
So ordered.
Notes
. The authority was originally vested in the Interstate Commerce Commission, then trans *1288 ferred to the Federal Energy Regulatory Commission when it replaced the Federal Power Commission in 1977. See 49 U.S.C.App. §§ 1 et seq. (1988); see also 49 U.S.C. § 60502 (“The Federal Energy Regulatory Commission has the duties and powers related to the establishment of a rate or charge for the transportation of oil by pipeline or the valuation of that pipeline that were vested on October 1, 1977, in the Interstate Commerce Commission or an officer or component of the Interstate Commerce Commission.”) (emphasis added). The Commission's jurisdiction over the rates for oil going through to Valdez is uncontested. See Trans Alaska Pipeline System, 23 FERChl 61,352 at 61,762 (1983).
. Exxon Company, U.S.A. was a division of Exxon Corporation. Since filing its appeal, Exxon Corporation has merged with Mobil Corporation to become Exxon Mobil Corporation.
. Unless stated otherwise, all citations to FERC orders have the title "Trans Alaska Pipeline System”.
. The parties appear to agree that here the reference point is November 30, 1993, the day on which the Commission ruled that the dis-tillalion method was just and reasonable. See
Exxon Decision,
. Naphtha is used to make gasoline. Because there are considerable trades on the Gulf Coast, there is a quoted price for Gulf Coast naphtha. On the West Coast, however, refineries overwhelmingly use their own naphtha rather than buying it. See J.A. at 940-41.
. Using adjusted gasoline prices, Stancil claims that West Coast naphtha is worth $29.09 per barrel. See J.A. at 1001.
