with whom Mr. Justice Brennan, Mr. Justice White, and Mr. Justice Marshall concur, dissenting.
In this case the United States appeals from a District Court decision1 upholding the acquisition of stock in United Electric Coal Companies by Material Service Corp. and its successor, General Dynamics Corp., against a challenge that the acquisition violated § 7 of the Clayton Act, 15 U. S. C. § 18.2 The United States instituted *512this civil antitrust action on the claim that the acquisition may substantially lessen competition in the Illinois and Eastern Interior Coal Province (EICP) sales area coal markets. After trial on the merits the District Court rejected the Government’s proposed product and geographic markets and dismissed the action, concluding that the Government had failed to show a substantial lessening of competition in the markets the court deemed relevant.
I
The combination here challenged is the union of two major Illinois coal producers — Freeman Coal Mining Corp. and United Electric Coal Companies — under the ultimate corporate control of General Dynamics Corp. Material Service Corp. acquired all the stock of Freeman Coal in 1942 and began 'to acquire United Electric stock in 1954. By 1959, holdings in United reached 34%, and Material Service requested and received representation on United’s board of directors. As a result, Freeman’s president was elected chairman of United’s executive committee. “With the affiliation of Freeman and United Electric thus formalized in 1959, common control of the two coal companies was achieved.” 341 F. Supp. 534, 537 (1972).
General Dynamics acquired Material Service Corp. in 1959 and moved to solidify the union of Freeman and United by engaging in continued purchases of United’s stock throughout the early 1960’s. By 1966 it held nearly two-thirds of United’s outstanding shares and a succesful tender offer increased the holdings to over 90%. In early 1967 United became a wholly owned subsidiary of *513General Dynamics. With the 1959 union of Freeman and United Electric thus completed, the Government filed this action challenging the legality of the combination which produced in General Dynamics the Nation’s fifth largest coal producer with total annual production of over 14 million tons.
II
Section 7 of the Clayton Act, the standard against which this combination must be tested, proscribes such combinations “where in any line of commerce in any section of the country, the effect of such acquisition may be substantially to lessen competition . ...”3 “Determination of the relevant market is a necessary predicate to a finding of a violation of the Clayton Act . . . .” United States v. E. I. du Pont de Nemours & Co., 353 U. S. 586, 593 (1957). The court below concluded that “the energy market is the appropriate line of commerce for testing the competitive effect of the United Electric-Freeman combination.” 341 F. Supp., at 555. The court rejected the Government’s hypothesis of coal as a submarket for antitrust purposes as “untenable,” finding that United States v. Continental Can Co., 378 U. S. 441 (1964), “compel[s] this court to conclude that since coal competes with gas, oil, uranium and other forms of energy, the relevant line of commerce must encompass interfuel competition.” 341 F. Supp., at 556.
I read Continental Can to import no such compulsion. That case involved the acquisition of the Nation’s third largest producer of glass containers, Hazel-Atlas Glass Co., by Continental Can, the country’s second largest producer of metal containers. The District Court found interindustry competition an insufficient predicate for finding a § 7 line of commerce embracing both cans and *514bottles. We reversed, finding that interindustry competition mandated “treating as a relevant product market the combined glass and metal container industries and all end uses for which they compete.” 378 U. S., at 457 (emphasis added). But that interindustry market was only one of several lines of commerce in that case. Both parties conceded that “the can industry and the glass container industry were relevant lines of commerce.” Id., at 447. Since § 7 proscribes acquisitions which may involve a substantial lessening of competition in any line of commerce, the absence of anticompetitive effects in either the bottle or can markets could not sustain the acquisition since there existed a market — the glass/metal container market given recognition in this Court — in which the prohibited effect was present.
The District Court here found an energy market in which the combination did not work the prohibited effect. Whatever the correctness of that finding, Continental Can teaches us that it is of no help to appellees if there exist other lines of commerce in which the effect is present. Any combination may involve myriad lines of commerce; the existence of an energy market is not inconsistent with and does not negate the existence of a narrower coal market for “within this broad market, well-defined submarkets may exist which, in themselves, constitute product markets for antitrust purposes.” Brown Shoe Co. v. United States, 370 U. S. 294, 325 (1962).
This principle found recognition in Continental Can where we recognized glass and metal containers “to be two separate lines of commerce,” despite finding that competition between the lines “necessarily implied one or more other lines of commerce embracing both industries.” 378 U. S., at 456-457 (emphasis added). It was also recognized in United States v. Aluminum Co. of America, 377 U. S. 271 (1964), which involved the com*515bination of an aluminum conductor manufacturer and a producer of both aluminum and copper conductor. The District Court there refused to treat aluminum conductor as a separate § 7 line of commerce because of the competition between aluminum and copper conductor. Though we found that competition sufficient to justify finding a single aluminum/copper conductor market, we reversed the District Court, holding that the interindustry competition did not preclude “division [of that market] for purposes of § 7 into separate submarkets.” Id., at 275.4
Coal has both price advantages and operational disadvantages which combine to delineate within the energy market “economically significant submarket [s].” 5 The consumers for whom price is determinative mark out a submarket in which coal is the overwhelming choice; the boundaries of this submarket are strengthened by coal's virtual inability to compete in other significant sectors of the energy market. Energy-use technology in highway and air transportation necessitates the use of liquid fuels. The relative operational ease of diesel-ized power plants has worked to virtually foreclose coal from the rail transportation market.6 Despite their higher cost, gas and oil enjoy a competitive edge in the space-heating market because of simple consumer preference for these sources of energy over coal.7
The market for coal is therefore effectively limited to large industrial energy consumers such as electric utilities and certain manufacturers with the ability and economic *516incentive to consider coal as an energy source.8 The court below noted that the “utility market has become the mainstay of coal production/’ 341 F. Supp., at 539. Within this sector coal’s economic advantage yields it an overwhelming share of the market. In each year from 1960 to 1967 (the period during which the Freeman-United Electric union solidified) coal accounted for over 90% of the B.t.u.’s consumed by steam electric utility plants in the EICP sales area; it also provided 74% of the B.t.u.’s consumed by cement plants in the same area and 94% of the B.t.u.’s consumed by such plants in Illinois.9
The coal market is therefore viewed by energy consumers as a separate economic entity confined to those users with the technological capability to allow the use of coal and the incentive for economy to mandate it. Within that market coal experiences little competition from other fuels since coal’s delivered price per B.t.u. in the areas served by Freeman and United Electric is significantly lower than that for any other combustible fuel except interruptible natural gas which is available only on a seasonal basis.10 Central Illinois Light Co., for example, purchases coal at 27 cents per million B.t.u.’s, *517firm natural gas at 45 cents, and oil (for ignition purposes) at 70 cents.11 Since coal consumption facilities are unique and not readily adaptable to alternative energy sources, there is little interfuel price sensitivity. As the court in Kennecott Copper Corp. v. FTC, 467 F. 2d 67, 79 (CA10 1972), stated in finding that “[t]he coal industry is a distinct submarket which has characteristics which are not shared by the other fuel industries,” coal prices “are now, and promise to be in the future, subject to the peculiarities of the coal business [since] other fuels appear to have a limited effect.”
The competitive position of coal is thus not unlike that of aluminum conductor in United States v. Aluminum Co. of America, supra. Like coal, aluminum conductor had “little consumer acceptance” for many purposes, but its substantial price advantage over other conductors gave it “decisive advantages” in those areas of the market where price was “the single, most important, practical factor.” 377 U. S., at 275-276. Despite the existence of some competition from other forms of conductor, those factors were sufficient to set aluminum conductor apart as an economically significant § 7 sub-market. That precedent seems to be indistinguishable; and thus whatever the existence of a § 7 energy market, coal constitutes an economically significant submarket for § 7 purposes.12
III
In rejecting the Government’s proposed geographic markets the court below adopted much narrower mar*518kets which, for the most part, followed ICC freight rate districts (FRD’s).13 The justification was that, since ordinary rail rates are the same for all mines in any particular FRD and since transportation costs are the principal competitive factor in coal marketing, mines in one FRD cannot effectively compete for the same customers with mines in other FRD’s. Since United Electric’s mines are located in the Belleville and Fulton-Peoria FRD’s and Freeman’s mines are located in the Springfield and Southern Illinois FRD’s, the combination of the two companies was found to present no risk of anti-competitive effects.
The error of the District Court in drawing the § 7 sections of the country “so narrowly as to place appellees in different markets” 14 is amply demonstrated by the overlapping distribution patterns of Freeman and United Electric. Though located in different FRD’s and thus supposedly not competitive, they sold one-half their output to the same customers at the same facilities. Lack of competition between FRD’s is further refuted by the existence of reciprocal selling patterns. For example, *519while United's Belleville FRD mine was selling 25% of its output to customers in the Southern Illinois FRD sales area, Freeman was selling 20% of its Southern Illinois FRD coal to Belleville sales area customers.
The inability of the lower court’s narrow markets to “ ‘correspond to the commercial realities’ ” 15 of the distribution patterns displayed in the record is explained by the undue weight given ordinary rail rates. While transportation costs are significant, ordinary rail rates are not the single controlling element of transportation costs. First, not all rail shipments are governed by FRD rates; many of the most significant shipments are transported via “unit trains” carrying only coal from a particular producer to a particular customer pursuant to a negotiated rate. Thus Freeman ships Southern Illinois FRD coal by unit train to a Belleville FRD sales area customer at a cost loioer than any Belleville FRD rate to that location. Second, not all coal transportation proceeds by rail. United transports most of its coal by barge, and in 1967 only one-half of all the coa-1 sold in the five States which receive coal from Illinois was transported by all-rail shipments.
Normal rail rates are thus not so limiting as to eliminate substantial competition between FRD sales areas. Coal producers may constitute strong competitive factors in areas up to 500 miles from the mine. Thus in 1967 Freeman’s Southern Illinois FRD Orient Mine shipped over 1.5 million tons of coal to customers 300 to 500 miles away. At the same time, United’s Fidelity Mine, only 40 miles from the Orient, shipped more than one million tons, over half its total production, to equally distant locations. Both Freeman and United Electric have mines which are capable of supplying any point in the EICP sales area.
*520Further, even assuming the existence of FRD markets, I think the court below erred in rejecting the Government’s proposed markets. As with product markets, § 7 does not necessitate an anticompetitive effect in any particular geographic market; its proscription reaches combinations which may substantially lessen competition in any section of the country. Thus, whatever the correctness of the District Court in finding FRD markets, the lack of anticompetitive effect in those markets is of no help to General Dynamics if competition may be lessened substantially in other geographic markets. And, as with product markets, the existence of FRD markets is not inconsistent with the existence of a myriad of other sometimes overlapping markets. Thus, in United States v. Pabst Brewing Co., 384 U. S. 546 (1966), we found Wisconsin, the Wisconsin-Michigan-Illinois tristate area, and the entire United States all to be relevant § 7 sections of the country in which to assess anticompetitive impact.
While existing sales patterns show that transportation costs are not as restrictive as the District Court found, long-range transportation costs and the national distribution of coal deposits serve to divide the country into regionally significant coal markets. Both Freeman and United Electric are located in the EICP, consisting of Central and Southern Illinois, Southwestern Indiana, and Western Kentucky, and parts of other nearby areas. The region overlies a geologically united coal-bearing rock sequence which is estimated to contain 36% of the Nation’s total coal resources. Because of the separation of the region from other major producing regions,16 *521EICP producers enjoy a substantial competitive edge with respect to sales in an area composed of Illinois, Indiana, Western Kentucky, parts of Tennessee, Eastern Iowa, Southeastern Minnesota, Southern Wisconsin, and extreme Eastern Missouri. In 1967, 82% of EICP coal was sold in this area. Freeman sold over 93% of its coal and United Electric sold over 97% of its coal in this area.
Within the EICP sales area, Illinois stands as an economically significant submarket. In 1967, 82% of the coal consumed in Illinois came from Illinois mines and 58% of the coal mined in the State was used there. Freeman sold 42% of its coal and United Electric sold 62% of its coal to Illinois consumers, more than either company sold in any other State. Since Illinois sales are dominated by Illinois producers and since all relevant Freeman and United Electric Mines are located in Illinois,17 the State constitutes a relevant and significant market for § 7 purposes. Although economic lines do not fall precisely along political boundaries, the Government is not required to delineate § 7 markets by “metes and bounds.” United States v. Pabst Brewing, supra, at 549. In holding a four-county group a relevant geographic market in United States v. Philadelphia National Bank, 374 U. S. 321 (1963), we noted the artificiality of such political boundaries but held- that “such fuzziness would seem inherent in any attempt to delineate the relevant geographic market.” Id., at 360 n. 37. The State of Wisconsin was held a relevant market in Pabst Brewing, supra, and in United States v. El Paso Gas Co., 376 U. S. 651, 657 (1964), we held that there could be “no *522doubt that California is a 'section of the country' as that phrase is used in § 7.”
IV
While finding no violation of § 7 in the Freeman-United Electric combination, the District Court did not make clear the standard used in reaching that ultimate conclusion. The court did not mention what it thought to be the relevant market shares nor did it discuss the effect of the combination on industry concentration. The court merely found that Freeman and United Electric do not compete because they are located in different FRD geographic markets, and because they sell different types of coal. As already discussed, nearly all the mines of both companies are located in Southern Illinois, and as demonstrated by past distribution patterns, with an ability to compete effectively at distances up to 500 miles, their presence in different minute FRD’s within Southern Illinois has simply not rendered them noncompetitive. The differences in the types of coal sold, moreover, are irrelevant. It is true, as the court below notes, that United Electric sells strip-mined coal while Freeman extracts deep reserves, but the fact that the companies sold half their output to common customers demonstrates that at least a significant portion of the consuming public is understandably unconcerned with the details of extraction. While it is also true that only Freeman sells metallurgical coal and a byproduct known as dust, this says nothing more than that the companies do not compete in metallurgical coal or dust; it does not relieve the court of the responsibility for evaluating the anticompetitive effects in nonmetallurgieal coal production — production which accounts for 100% of United’s and 92% of Freeman’s business.18
*523The court further found that United Electric, standing alone, would not contribute meaningfully to further competition since virtually all its economically mineable strip reserves were committed under long-term contracts and it possessed neither the capability to obtain more strip reserves nor the expertise to develop its deep reserves. Although the doctrine was not invoked by name, this appears to be an application of the “failing company” defense. See Citizen Publishing Co. v. United States, 394 U. S. 131 (1969). If it is, the court proceeded on an analysis made at the wrong time and failed to discuss the legal standards employed in finding the defense to be established. The finding that 48 of United’s 52 million tons of strip reserves were committed related to the time of trial. But, since the rationale of the failing-company defense is the lack of anticompetitive consequence if one of the combining companies was about to disappear from the market at any rate, the viability of the “failing company” must be assessed as of the time of the merger. United States v. Greater Buffalo Press, 402 U. S. 549, 555 (1971); Citizen Publishing Co. v. United States, supra, at 138.
The Court urges that United’s weak reserve position, rather than establishing a failing-company defense, “went to the heart of the Government’s statistical prima facie case based on production figures.” Under this view United’s weak reserve position at the time of trial constitutes postacquisition evidence which diminishes the possibility of anticompetitive impact and thus directly affects the strength of time-of-acquisition findings. The problem *524with this analysis is that the District Court made no time-of-acquisition findings which such postacquisition evidence could affect. The majority concedes the obvious need for a limitation on the weight given postacquisition evidence and notes that we have reversed cases where “too much weight” has been given. Here the postacquisition events were given all the weight because all the District Court’s findings were made as of the time of the trial. While findings made as of the time of the merger could concededly be tempered to a limited degree by post-acquisition events, no such findings were ever made.
Many of the commitments here which reduced United’s available reserves occurred after the acquisition; 21 million tons for example were committed in 1968. Similarly, though the District Court found further mineable strip reserves unavailable at the time of trial, there is no finding that they were unavailable in 1959 or 1967. To the contrary, the record demonstrates that other coal producers did acquire new strip reserves during the 1960’s.19 United’s 1959 viability is further supported by the fact that it possessed 27 million tons of deep reserves. While we do not know if all these reserves were economically mineable at the time of the acquisition, there was no finding that they would not become so in the near future with advances in technology or changes in the price structure of the coal market.20 Further there was no contention or finding that further deep reserves were not available for acquisition.21 The District Court *525merely concluded that United had no “ability to develop deep coal reserves.” 22
While it is true that United is a strip-mining company which has not extracted deep reserves since 1954, this does not mean that United would not develop deep-mining expertise if deep reserves were all it had left or that it could not. sell the reserves to some company which poses less of a threat to increased concentration in the coal market than does Freeman. United Electric was not, as the Court suggests, merely one of many companies with the possible “inclination and the corporate treasury” to allow expansion into “an essentially new line of business.” United was a coal company with a thriving coal-marketing structure. At the time of the merger it had access to at least 27 million tons of deep reserves and it had operated a deep mine only five years previously. While deep-coal mining may have been an essentially new line of business for many, it was for United merely a matter of regaining the expertise it once had to extract reserves it already owned for sale in a market where it already had a good name.
*526V
Thus, from product and geographic markets to market-share and industry-concentration analysis to the failing-company defense, the findings below are based on legal standards which are either incorrect or not disclosed. While the court did gratuitously state that no § 7 violation would be found “even were this court to accept the Government’s unrealistic product and market definitions,” this conclusory statement is supported by no analysis sufficient to allow review in this Court. The majority notes that production figures are of limited significance because they include deliveries under long-term contracts entered into in prior years. It is true that uncommitted reserves or sales of previously uncommitted coal would be preferable indicia of competitive strength, but the District Court made no findings as to United’s or Freeman’s respective market shares at the time of the acquisition under either of these standards.23
*527On the basis of a record so devoid of findings based on correct legal standards, the judgment may not be affirmed except on a deep-seated judicial bias against § 7 of the Clayton Act. We should remand the case to the District Court with directions to assess the impact of the Freeman-United Electric combination on the Illinois and EICP sales area coal markets as of 1959.24 We should direct the court to make findings of respective market shares, and further to evaluate United Electric's viability as an independent producer or as the possible “acquiree” of a company other than General Dynamics as of 1959, in light of the strict standards applicable to the failing-company defense. Since we abdicate our duty for responsible review and accept the mere conclusion that no § 7 violation is established on the basis of a record with none of these necessary 'findings, I dissent from the affirmance of the District Court’s judgment.
341 F. Supp. 534 (1972).
Title 15 U. S. C. § 18 provides:
“No corporation engaged in commerce shall acquire, directly or indirectly, the whole or any part of the stock or other share capital and no corporation subject to the jurisdiction of the Federal Trade Commission shall acquire the whole or any part of the assets of *512another corporation engaged also in commerce, where in any line of commerce in any section of the country, the effect of such acquisition may be substantially to lessen competition, or to tend to create a monopoly.”
Supra, n. 2.
Similarly, in United States v. Philadelphia National Bank, 374 U. S. 321 (1963), we held commercial banking a §7 line of commerce even though banks compete with other institutions with respect to some services such as the making of small loans.
See Brown Shoe Co. v. United States, 370 U. S. 294, 325 (1962).
341 F. Supp., at 539.
Ibid.
The only other significant use for coal is metallurgical in nature. Metallurgical coal is used as a product in the manufacture of steel. The use of such coal as a product sets it off in a separate market from nonmetallurgical coal which is used as an energ}' source.
Although nuclear -and geothermal power may draw some utility consumers from the coal market in the future, nuclear fuel is not consumable in existing fossil-fuel plants nor is nuclear fuel presently an alternative for nonutility coal consumers. Thus, whatever the future inroads of alternative fuels, there remains a significant class of energj' consumers which looks only to coal.
Interruptible gas is sold at a lower rate and is available only when it is not required by firm-rate customers which are supplied according to their needs and which always have priority.
Oil is used by some coal consumers for purposes to which coal is not suited such as starting up boilers or kilns.
Even the court below gave some recognition to coal as a separate market in its discussion of the relevant geographic markets. The geographic markets were delineated along “the distributive patterns of . . . coal,” separating out those “mines to which coal consumers can practicably turn for supplies.” 341 F. Supp., at 556 (emphasis added).
Freight rate districts are producing areas grouped for ICC rate-making purposes; all mines within each producing area are accorded the same rates to the same consuming destinations. See Ayrshire Collieries Corp. v. United States, 335 U. S. 573, 576 (1949). The other markets accepted by the District Court are Commonwealth Edison and the Metropolitan Chicago Interstate Air Quality Control Region. Commonwealth Edison was found to be unique in light of its massive coal requirements, its purchasing patterns which are “quite distinct from [those] followed by other consumers,” and its singularly extensive commitment to nuclear energy. The MCIAQC, consisting of six Northeastern Illinois counties and two Northwestern Indiana counties, was found unique because of its singular access, through water and rail arteries, to almost all FRD’s in the Midwest.
See United States v. Philadelphia National Bank, 374 U. S., at 361.
See Brown Shoe Co. v. United States, 370 U. S., at 336.
The Nation’s other major coal producing regions are: (1) the Eastern Coal Province of Western Pennsylvania, West Virginia, Eastern Kentucky, and parts of Ohio, Tennessee, and Alabama; (2) the Western Interior Coal Province comprised of Central Iowa, Northern and Western Missouri, and Eastern Oklahoma; and (3) scat*521tered deposits in Montana, Wyoming, Colorado, and Utah. Jurisdictional Statement 5.
United Electric also controls some coal deposits in Colorado and Oklahoma which are not in issue in this case. 341 F. Supp., at 538 n. 8.
The lack of competition from United for a mere 8% of Freeman's business is simply irrelevant. In United States v. Aluminum *523Co. of America, 377 U. S. 271 (1964), we struck down a combination which affected competition in the aluminum conductor market, and that result was not affected by the irrelevant fact that one of the companies, Rome Cable, also engaged in the production of copper conductor.
See Brief for United States 71.
Research into new methods of extraction or a rise in the price of coal could make reserves which are uneconomical to mine at any given time economically mineable in the future.
To the contrary, United Electric acquired substantial new deep reserves after the time of the acquisition since it now owns about 44 million tons of deep reserves and controls by location another 40 to 50 million tons. Reserves are controlled by location if, in *525order to be mined at all, they must be mined by those who control, by ownership, lease, or option, the contiguous reserves.
If that conclusion is to lend support to the combination on the ground that United “standing alone, cannot contribute meaningfully to competition,” it must be made in light of the stringent standards applicable to the failing-company defense. In Citizen Publishing Co. v. United States, 394 U. S. 131, 138-139 (1969), we said that the defense is one of “narrow scope” and that the burden of proving the defense is “on those who seek refuge under it.” We also stated that the prospects of continued independent existence must be “dim or nonexistent” and that it must be established that the acquiring company is the only available purchaser. See also United States v. Greater Buffalo Press, 402 U. S. 549, 555-556 (1971), and United States v. Third National Bank in Nashville, 390 U. S. 171, 189 (1968).
The District Court did find that, as of 1968, Freeman controlled 6.5% of the total coal reserves dedicated to existing mines in the EICP. At the same time, United Electric controlled 2.5% of that total,, but almost all of this was contractually committed. If market shares are to be determined by percentage of total reserves, what is necessary is a finding as to each company's 1959 share of uncommitted Illinois and EICP reserves — including reserves which were economically mineable or which might have become so in the reasonably near future and further including an estimate as to uncontrolled reserves which might have been acquired by either company in the reasonably near future.
The District Court also found that, as of 1968, the two companies together accounted for 10.9% of the EICP coal production, and that this figure represented more than a 10% decrease from the combined production for 1959. Combined 1959 production by the companies was thus at least 12.1% of the EICP total. If market shares are to be determined by percentage of industry sales, this figure is in excess of percentages found illegal in markets with a trend toward concentration (see, e. g., United, States v. Von’s Grocery Co., 384 U. S. 270 *527(1966) (7.5%), and United States v. Pabst Brewing Co., 384 U. S. 546 (1966) (4.49%)), and the court below recognized an increase in concentration in the coal market. It might be argued, however, that, if market share is to be determined by sales, the production figures found by the court below are not the relevant ones for they include production which goes to meet obligations incurred in long-term contracts entered into in prior years. In terms of competition, if sales are the relevant criteria, what is needed is a finding of “new” sales (sales of previously uncommitted coal) as a percentage of total industry new sales in Illinois and the EICP at the time of the acquisition.
Common control of the two companies was achieved in 1959 and the combination was completed in 1967; at oral argument both parties conceded that the merger “took place” in 1959.