Permian Basin Area Rate Cases

*829 MR. Justice Douglas,

dissenting.

I.

What the Court does today cannot be reconciled with the construction given the Natural Gas Act by FPC v. Hope Natural Gas Co., 320 U. S. 591, 602 . In that case we said, in determining whether a rate had been properly found to be “just and reasonable” under the Act, that

(1) “it is the result reached not the method employed which is controlling”;

(2) it is “not theory but the impact of the rate order which counts”;

(3) “If the total effect of the rate order cannot be said to be unjust and unreasonable, judicial inquiry under the Act is at an end.”

The area rate orders challenged here are based on averages. 1 No single producer’s actual costs, actual risks, actual returns, are known.

*830 The “result reached” as to any producer is not known.

The “impact of the rate order” on any producer is not known.

The “total effect” of the rate order on a single producer is not known.

It is said, however, that if any producer is aggrieved, it may apply for relief and if it fails to obtain relief it can resort to the courts. But unless we know the standards which will govern in case it applies for relief, we are, with all respect, mouthing mere words when we say the *831 rate is “just and reasonable.” In absence of knowledge, we cannot possibly perform our function of judicial review, limited though it be.

It was urged in the separate opinion of Mr. Justice Jackson in Hope that a system of regulation be authorized which would center not on the producer but on the product “which would be regulated with an eye to average or typical producing conditions in the field.” 320 U. S., at 652 . But the Court rejected that approach, saying that §§ 4 (a) and 5 (a) of the Natural Gas Act contained “only the conventional standards of rate-making for natural gas companies.” Id., at 616.

Group regulation of rates is not, of course, novel. It has at times been authorized. The Federal Aviation Act of 1958, § 1002 (e), 72 Stat. 789 , 49 U. S. C. § 1482 (e), permits it. And see General Passenger-Fare Investigation, 32 C. A. B. 291. Under the War Power, extensive price regulation on a group basis was sustained. Bowles v. Willingham, 321 U. S. 503, 517-519 . The Interstate Commerce Commission has undertaken it, as revealed by the Divisions of Revenue cases. New England Divisions Case, 261 U. S. 184 ; United States v. Abilene & S. R. Co., 265 U. S. 274 ; Chicago & N. W. R. Co. v. A., T. & S. F. R. Co., 387 U. S. 326 . See also § 15 of the Interstate Commerce Act, as amended, 24 Stat. 384 , 49 U. S. C. § 15 (3). The requirement in the Divisions of Revenue cases is that the group evidence be “typical in character, and ample in quantity, to justify the finding made in respect to each division of each rate of every carrier.” 261 U. S., at 196-197 . In other words, where the rates fixed will recover the typical group cost of service, the individual producer’s right to a minimum of its operating expenses and capital charges is protected. Cost of service includes operating expenses and capital charges. FPC v. Natural Gas Pipeline Co., 315 U. S. 575, 607 (concurring opinion). With *832 that protection I can see no reason why group rates may not be sanctioned here. But more is required than the Commission undertook to do in these cases.

In the present cases the Commission found averages; but there are no findings as to the typicality and representative nature of those averages. 2 We certainly cannot *833 take judicial notice that the averages are typical. Mr. Justice Brandéis in the leading Divisions of Revenue case said that “averages are apt to be misleading” and they cannot be accepted “as a substitute for typical evidence.” 265 U. S., at 291 . Cf. American Motors Corp. v. FTC, 384 F. 2d 247, 251-259, 260-262 (C. A. 6th Cir. 1967).

The Commission found no m,edian. Moreover, as we observed in another context, it did not find what was “the average cost” of groups made up of individual members who have “a close resemblance” when it comes to the “essential point or points which determine the *834 costs considered.” United States v. Borden Co., 370 U. S. 460, 469 .

With respect to the cost of new gas-well gas, the Commission did not determine whether the average costs compiled from the questionnaires or derived from industry-wide data were typical or representative.

In finding the cost of flowing gas, the Commission noted that the 1960 level of costs compiled by the staff in large part from the questionnaire responses was “fairly representative of the costs during the three year period ending in 1960” (34 F. P. C. 159, 213) and that “[t]he 1960 test year is . . . typical of current and future costs of the flowing gas . . . .” Ibid. This reference to “representative” and “typical” costs, however, dealt only with the question of time — i. e., the staff’s use of 1960 data in developing its composite cost presentation was deemed permissible since 1960 was found to be a typical and representative year.

The Court professes to find that the Commission adequately determined that the averages it employed were “typical” and “representative.” Ante, at 802-803, n. 79. But the statements plucked from the Commission’s opinion do not support that interpretation.

The Commission also observed, with respect to the questionnaire data, that 42 of the major producers (representing all but one of the major producers in the Permian area) responded on the Appendix B questionnaires. The Commission agreed with the Examiner that “the data provided by the major producers with respect to their Permian production was fully representative of area costs,” and that exclusion of the Appendix C returns from small producers would have only a de minimis effect. 34 F. P. C., at 214. But although the data submitted by the major producers were found to be typical data for the area, and I assume also for the major producers in the area, there are no findings whether the averages *835 compiled from the data were typical or representative of the costs of those major producers or of other producers in the area.

The Commission’s statement that the sources used “in combination provide an adequate basis for the costs we have found” certainly cannot be read as a finding that those sources were “typical and representative.” Nor does the fact that the sources were “recognized, published statistical data sources,” or “well-recognized and authoritative,” mean they also contained typical and representative averages.

An average cost is not only apt to be “misleading”; it may indeed not be representative of any producer.

The Commission allowed a 12% rate of return, the return being “on capital invested in finding new gas well gas.” 34 F. P. C., at 306, 343. “Production investment costs” constituted this “capital invested” and were the bases to which the Commission applied the 12% rate to arrive at a return of 5.21$ per Mcf to be included in the rate base for new gas-well gas. 34 F. P. C., at 197, 204. These “production investment costs” included successful well costs, lease acquisition costs, and the cost of other production facilities. But they were likewise determined on the basis of averages. See 34 F. P. C., at 197-198, 295, 377-382.

The average per capita income of a Middle East kingdom is said to be $1,800 a year. But since one man— or family — gets most of the money, $1,800 a year describes only a mythical resident of that country.

The 12% return allowed by the Commission and computed on an average-cost basis may likewise have no relation whatever to the reality of the actual costs of any producer.

One producer’s cost, though varying from year to year, may average out at $1 per Mcf. Another’s may average out at 5(4 per Mcf. Does that make 52.5$ per Mcf repre *836 sentative of either producer or typical of all producers, or, indeed, typical of any producer, even if the 52.5$ per Mcf is stable over the entire period of years?

The Commission could follow the lead of the Interstate Commerce Commission and produce rates on a group basis. But it simply has not done so in any rational way.

Averages are apt to take us with Alice into Wonderland. That is one reason why the case should be remanded to the Commission for further findings.

The Commission will allow individual application for relief from these new rates. But it has not prescribed the terms and conditions on which relief will be granted. It has said, however, that an individual producer must show more than that its cost of service is greater than the averages on which the rate is based. 34 F. P. C., at 180.

In a regulated industry there is no constitutional guarantee that the most inefficient will survive. Hegeman Farms Corp. v. Baldwin, 293 U. S. 163, 170-171 .

That assumes, however, an ability to withdraw from the business. But a producer of natural gas may not abandon its existing facilities that supply the interstate market without Commission approval. United Gas Pipe Line Co. v. FPC, 385 U. S. 83 .

The Commission says that a producer will be able to obtain relief to cover its out-of-pocket expenses. 34 F. P. C., at 226. Do they include return, depreciation, depletion, exploration, development, and overhead? The Court of Appeals did not know (375 F. 2d, at 30); and we certainly do not. The remand by the Court of Appeals for further definition was therefore clearly necessary. For even if we need not know the precise impact of the new group rate on each producer at the time of the group rate order, we certainly must know the conditions on which a producer can get relief before we can say that a rate as to it is “just and reasonable.”

*837 Although we assume that the Act authorizes group rate-making, we cannot disregard the basic structure of the Act, patterned on the “conventional standards of rate-making” (FPC v. Hope Natural Gas Co., supra, at 616) and providing in §§ 4 (a) and 5 (a) that all rates of “any” natural gas company be “just and reasonable.” Beyond the group is the single producer; beyond the community of producers is the individual. The ultimate thrust of the Act reaches the individual producer; and unless we know what the group rate in final analysis does to it or disables it from doing we cannot perform our duty of judicial review.

II.

If we move to the regulation of the group as such and consider the impact of these rate orders on it, we are likewise not able on the present record to perform our function of judicial review.

It is impossible to say whether the proper revenue requirements of the group can be satisfied under this rate order. For the costs represent averages; and there is no way for us to find from the record whether these averages are typical and what the impact of the rates on the group will be.

The error is compounded when the costs used are the purported costs of gas-well gas and do not include the costs of casinghead gas, residue gas derived therefrom, and gas-well gas from combination leases. The Commission concluded that the costs of casinghead gas and residue gas produced therefrom did not exceed the costs for gas-well gas. Yet at the same time it rejected proffered evidence of higher costs of processing gas to remove liquid hydrocarbons. Commission expertise should not be allowed to make its own “facts” to justify the desired result.

*838 Beyond that are the quality adjustments. Upward price adjustments are permitted for Btu content above 1,050 per cubic foot and downward adjustment for Btu content below 1,000. The Commission was concerned with the value of the “energy content of the gas, which in reality is what the consumer is purchasing.” 34 F. P. C., at 223.

With that standard in mind it allowed price reductions

(1) where the gas contains more than 10 grains of hydrogen sulphide or 200 grains of total sulphur per Mcf;

(2) where it contains more than .009 pound per Mcf of water;

(3) where it contains more than 3% by volume of carbon dioxide;

(4) where the gas pressure is less than 500 pounds per square inch.

When any of these standards are not met, the applicable ceiling price is adjusted downward by the net cost of processing the gas to bring it up to standard.

Under the Commission’s standards about 90% of the flowing gas moving interstate from the Permian Basin is not of the pipeline quality that the Commission has prescribed. 375 F. 2d, at 30. What the costs will be to1 convert the gas to these new standards is not found in this record. Perhaps this deficiency is due to the fact that the Commission, almost as an afterthought and not with clear, advance notice, decided to deal with detailed quality standards. But without knowing these costs through competent evidence, neither we nor the Commission has any way even to guess at whether the new rates will satisfy the criteria of Hope.

III.

The Court approves the Commission’s treatment of the quality adjustments as a risk of production. But *839 whether they be labeled a risk of production or a cost would seem to be irrelevant. That is a matter of semantics as far as the standards of Hope are concerned. For the question is whether we can reasonably determine the end result from the computations of the Commission, including both risk and cost factors.

Any unknown cost is a risk. But the Commission should not be permitted to excuse its failure to solicit or proffer appropriate evidence concerning the cost of converting gas into pipeline quality by labeling that cost a “risk.” The Court of Appeals recognized this point. See 375 F. 2d, at 31-32, 35. Commissioner O’Connor noted in his opinion concurring in the denial of rehearing that: “To bury the quality impact in our rate of return determination is to overlook the basis for the 12 per cent allowance: comparable return on equity of 10-12 per cent by the far less risky operations of transmission companies.” 34 F. P. C., at 1081. And, as one commentator recently observed:

“The Commission stated that the rate of return also reflected the risk of finding gas of less than pipeline quality — a clever way of avoiding the quality discount problem. Since there was no evidence in the record as to what those discounts would be, one can only say that 'risks’ were involved. It is a novel doctrine, indeed, that the rate of return should be adjusted to reflect the risk that the regulatory cost computations are incorrect.” 3

The Court concedes that the lack of specific findings concerning the effect of the quality adjustments upon the rate of return was “an unfortunate omission.” Ante, at 812. But it proceeds to scratch about for evidence *840 to support the Commission. With all respect, there is no competent evidence in the record to permit a meaningful determination of the impact of the quality deductions. 4 The Court of Appeals was clearly correct in *841 remanding to the Commission for proper findings on this point.

Behind the veneer of the Court’s opinion may be an unstated premise that the complexity of the task of regulating the wellhead price of gas sold by producers is both so great and so novel that the Commission must be given great leeway. But the permissible bounds, so far as judicial review is concerned, are passed when guesswork is substituted for reasoned findings, when the Commission can avoid finding “costs” by the convenience of calling them “risks,” when rates of return are computed for those mythical producers who happen to meet the “average” specifications.

If the task of regulating producer sales within the framework of the Natural Gas Act is as difficult as the present cases illustrate, perhaps the problem should be returned to Congress. But certainly we do little today to advance the cause of responsible administrative action. With all respect, we promote administrative irresponsibility by making an agency’s fiat an adequate substitute for supported findings.

IV.

New Mexico and Texas, in which the Permian Basin is located, have comprehensive oil and gas conservation codes. 5 A substantial portion of their taxes on the pro *842 duction of natural gas within their boundaries goes into school funds. They say that the “public interest” entrusted to the Commission by 15 U. S. C. § 717 (a) includes the interest of the States where the gas is found. They claim that pricing can be disastrous to the producing States and urge the need for threefold findings by the Commission to ensure an adequate supply of natural gas for future use:

“First, the Commission must determine the quantity of gas needed to constitute an adequate future supply. Secondly, it must make a conclusion as to the level of exploration and development which will produce the needed gas supply. Finally, it must prescribe a rate which will elicit that level of exploration and development.”

They argue that where Commission rates are lower than existing contract rates, continued operation is uneconomical in many so-called “stripper fields”:

“Although daily per well production from these fields is relatively low, their combined remaining recoverable reserves nevertheless constitute a considerable percentage of the total reserves for the area which will be forever lost if it becomes necessary to plug and abandon these fields for economic reasons.”

The Court of Appeals did not entertain these objections (375 F. 2d, at 18) because it read the Hope case as foreclosing them.

Hope, however, did not involve regulation of producers of natural gas, only interstate pipelines. At that *843 time, Phillips Petroleum Co. v. Wisconsin, 347 U. S. 672 , giving the Commission authority over these producers, had not been decided. In Hope we assumed that the Act meant what it said in § 1 (b) when it did not extend federal control to the “production or gathering of natural gas.” We were not then reviewing a federal order fixing wellhead gas prices for producers. Wellhead gas was not even involved in the Hope case. We were concerned there with abuses and overreaching by pipeline companies. We said:

“If the Commission is to be compelled to let the stockholders of natural gas companies have a feast so that the producing states may receive crumbs from that table, the present Act must be redesigned. Such a project raises questions of policy which go beyond our province.” 320 U. S., at 614 .

Now that Phillips has put the prices of producers under federal control, the interests of the producing States must be considered, appraised, and weighed as an important ingredient of the “public interest.” Regulation of wellhead prices by the Commission directly influences the level and feasibility of production, and can significantly affect the producing States’ regulation of production. See Phillips Petroleum Co. v. Wisconsin, supra, at 689-690 (dissenting opinion). 6

As the Court today says in another context, price in functional terms can be “a tool to encourage” the production of gas. Ante, at 760. The effect of price on the regulatory responsibilities of the several States must therefore be weighed, unless contrary to the mandate of the Act regulation of production is to pass into federal hands.

What the merits may be on this issue we do not know. The matter is complicated. For example, it seems *844 that the revenues of the processing plants are dérived primarily (about 80%) from the liquids which they extract from the casinghead gas, rather than from the sale of the residue gas. We do not know how to appraise the chances that this gas would be flared rather than processed if the price were too low. For example, it might be that the processing plants would continue to purchase and process casinghead gas as long as the revenues from the liquids extracted plus those from the residue gas processed exceeded the cost of gathering, processing, and marketing the gas. As long as there is a market for the residue gas remaining after extraction of the liquids, it might be that the processor would sell it at almost any price rather than flare it, in order to recover at least part of his costs. This assumes, of course, that the processor has already made the investment in equipment necessary to purify the residue gas to make it salable, and that the operating costs of this process are not prohibitive. Conceivably, the price of the residue gas could influence the processing plants in deciding whether to maintain or install the equipment and procedures necessary to make salable quality residue gas as the liquids are being extracted. We do not know how many processors do not now have that necessary equipment or the cost of operating and maintaining that equipment.

If the processor is willing to gather and process the gas because of the value of the liquids extracted, it might be that a producer would be willing to sell its casinghead gas rather than flare it, in order to obtain some payment for the gas. On the other hand, the price of the casing-head gas might well be critical for marginal producers, whose revenues from the sale of casinghead gas justify keeping their oil wells in production. But we have no *845 evidence concerning how many oil producers in the Permian Basin area could be termed “marginal.”

It may be that the posture of Hope was the reason why this phase of the case was not developed. Whatever the reason, it must be developed if the interest of the producing States is not by judicial fiat to be subjected entirely to complete federal supremacy, contrary to the promise in the Natural Gas Act.

In its effort to determine costs of production, the Commission sent out questionnaires (Appendices A, B, and C), to 458 producers in the Permian Basin area, 361 of which were named respondents in these proceedings. Appendices B and C inquired as to production costs; Appendix A covered drilling costs. Appendix B was a comprehensive questionnaire designed for major producers, while Appendix C was a simplified form for small producers (those with under 10,000,000 Mcf in nationwide jurisdictional sales in 1960). Small producers, however, could answer either Appendix B or C.

The Commission received complete responses on Appendix B from 67 producers, of which 25 were small producers. Responses to Appendix C were filed by 105 small producers. (Some of the responses represented composite data for more than one company.) The Commission excluded the Appendix C replies from consideration. 34 F.x P. C. 159, 213-214.

The Commission’s staff used these responses to develop a composite cost of service study. The staff arranged the Appendix B replies on various charts, arraying the data from high to low in respect to various categories (e. g., total unit costs and allow- *830 anees, cash expense unit costs). Then, weighted cost averages were computed — i. e., the replies on Appendix B were given a weight proportional to the volume Mcf covered by the responses.

In establishing the rate for new gas-well gas, the Commission elected to proceed by determining costs on a national, rather than an area, basis. 34 F. P. C., at 191. It used the Permian questionnaire responses, however, as “a vital source of information,” ibid., employing them in determining various components of the final national average cost. See id., at 191-200. The Commission also turned to various “well-recognized and authoritative industry data sources [which] were utilized by various witnesses in the proceeding.” Id., at 191. These included such sources as the United States Census Bureau’s Census of Mineral Industries for 1958 (wherever this source was used, the figures were trended to 1960 on the basis of the Permian questionnaire data), the 1961 Chase Manhattan Bank’s Annual Analysis of the Petroleum Industry, and the 1958 Joint Association Survey (a survey made by three industry trade groups based on questionnaires mailed to all member companies).

Various adjustments were made because of factors such as atypical years or the Permian questionnaire data being disproportionate to the national figures. See 34 F. P. C., at 194^196.

The Commission’s rate for flowing gas was based primarily on the questionnaire data which had been compiled by the staff into a composite cost of service study. The Commission in this instance based the ceiling price on Permian Basin area costs, although it used nationwide data in determining exploration and development costs. See 34 F. P. C., at 212-218. And, although the term “flowing gas” was defined to include casinghead gas, residue gas derived therefrom, and old gas-well gas, the Commission used only the costs of the old gas-well gas in determining the area rate. Id., at 208-212.

Nor did the Commission discuss the distribution of the data within the grouping being considered — that is, matters of the concentration, symmetry, and uniformity of the data.

The Commission asserts in this Court that “while producer costs vary widely from year to year on an individual-company basis, in the long run the costs of most producers tend to approximate the industry average.” In support of this assertion, it cites record testimony and refers to the existence of fairly stable industry averages for drilling costs of successful wells as compared with erratic figures for individual companies. Apart from the fact that not all of the testimony cited stands for the proposition stated by the Commission, but indicates at most only that there is less instability in individual producers’ costs over time rather than that they tend to average out, there was conflicting testimony on the point of representativeness offered by a witness for the Sun Oil Company, who showed that certain averages were not representative of the basic data because the distribution of the data was so widely spread and skewed from the mean. The fact that there were no comprehensive cost data suitable for supplying all the necessary elements of a cost study (see 34 F. P. C., at 191) does not excuse the Commission from finding whether the data it chose to use were typical and representative. In fact, the necessity of making such a finding is accentuated, because of the number of different sources entering into the computation of virtually all of the individual cost components. See 34 F. P. C., at 191-207, 212-218.

The Commission stated that it would use national rather than area data in arriving at a cost for new gas-well gas, noting: “It may be that in some areas production costs may vary sufficiently from the national average to warrant a different treatment but on the record in this case we agree that cost of new gas-well gas should be determined on the basis of nationwide data.” 34 F. P. C., at 191. Since the Commission was discussing the use of area versus national costs, that statement at most suggests only that the Permian *833 Basin composite costs did not vary sufficiently from the national average costs to warrant not using the latter, rather than that the Commission was comparing the national average with individual producer costs in the Permian Basin.

Perhaps for a group as large and diversified as that involved in this case, typical and representative averages cannot be computed. Hunt Oil Company presses this point strongly, contending that wide variations in unit costs are an inherent characteristic of gas and that a uniform ceiling rate fixed at the average composite cost level is unlawful per se because of the wide disparity in costs among different categories of gas as well as among different producers. The Commission itself noted this fact of wide variation in individual costs as part of its justification for basing costs on overall producer experience (see 34 F. P. C., at 179); but, as pointed out, it failed to go forward and determine whether the averages used to construct this overall producer experience were typical and representative. If they were not, then perhaps the Commission could have subdivided the group until it arrived at groupings whose members possessed essentially similar characteristics. Cf. United States v. Borden Co., 370 U. S. 460, 469 . This would not mean that the Commission would in effect be returning to an individual producer regulatory method; rather, the Commission could stop the subdivision at that point where group averages became typical and representative. But, as this ease now stands, the Commission has not made the necessary findings; and, of course, this Court, lacking the required expertise, cannot undertake to supply those findings for the Commission, nor is it our function to do so. See, e. g., United States v. Abilene & S. R. Co., 265 U. S. 274 .

Kitch, The Permian Basin Area Rate Cases and the Regulatory Determination of Price, 116 U. Pa. L. Rev. 191 , 201 (1967) (footnote omitted).

Counsel for the Commission observe in their brief to this Court that “[n]o more precise determination was possible in the state of the record” than the 0.70 to 1.5$ range for the average adjustment for quality predicted by the Commission in its opinion denying rehearing. See 34 F. P. C., at 1073. Counsel also cite to certain record testimony and exhibits to support the Commission’s determination of this 0.7(4 to 1.50 range.

It should be noted first that the 0.70 to 1.50 prediction is an average. I have already discussed the misleading nature of averages not found to be typical and representative, and those observations are equally pertinent here. Moreover, we have no idea whether the Commission relied in making its prediction on any of the sources cited by Commission counsel to this Court.

In computing the 0.70 to 1.50 range in its opinion denying rehearing, the Commission apparently relied on Commissioner O’Connor’s statement in his concurring opinion to the initial decision that the average adjustment would be between 1.00 and 1.70, and then adjusted those figures to allow for certain changes made with respect to quality standards in the decision denying rehearing. But at the time of the Commission’s initial decision, Commissioner O’Connor did not and could not know the costs incurred by the pipelines in bringing gas up to pipeline quality, for the pipelines’ processing costs were not in the record. Commissioner O’Connor based his estimate in large part on contract exhibits, as is evident from his opinion; and he noted that a precise adjustment for quality could not be ascertained from those exhibits. See 34 F. P. C., at 266. His view of the evidence on this point was clearly stated in his opinion concurring in the denial of rehearing, in which he observed that the record “does not permit a meaningful determination of the impact.” 34 F. P. C., at 1081.

Commission counsel also note the Examiner’s finding that 10 represented a reasonable estimate for bringing new gas-well gas up to pipeline quality and 10 to 1.50 for old gas-well gas. But, as counsel admit, this finding was not made in conjunction with defining pipeline quality standards on which the costs of conforming the quality of the gas would be based. In fact, the Examiner con- *841 eluded that: “This record does not permit the determination of a complete set of quality and value differentials.” 34 F. P. C., at 370.

The percentage calculations translating the 0.70 to 1.50 range into terms of rate of return, which are relied upon by the Court, were presented by Commission counsel to this Court and do not appear in the Commission’s opinion or in the record.

See N. M. Stat. Ann., c. 65 (1953); Tex. Stat. Ann., Art. 6004r-6066d (1962). In 1935, Texas, New Mexico, Kansas, Oklahoma, Illinois, and Colorado agreed upon an interstate compact for the conservation of oil and gas. Congress subsequently gave its consent *842 to the compact on August 27, 1935, for a period of two years. Pub. Res. No. 64, 49 Stat. 939 . The compact has been extended by the compacting States, with the consent of Congress, for successive periods without interruption, the latest extension being from September 1, 1967, to September 1, 1969. Pub. L. No. 90-185, 81 Stat. 560.

See also H. R. Doc. No. 342, 84th Cong., 2d Sess., 2 (1956).