(dissenting) ;
I respectfully dissent for the reason that in my view the majority, which adopts a new ground (the horizontal impact of the Amerock-Stanley merger on actual competition) for upholding the Commission’s decision (which was squarely bottomed on elimination of potential competition), has automatically applied an oversimplified quantitative “rule of thumb” instead of recognizing economic realities of the market. It ignores the fact that the actual competitive overlap between the two companies is less than % of 1% of the market, that there has been a complete absence (1) of any trend toward concentration in the industry, (2) of any parallel action by leading producers, or (3) of any appreciable enhancement in the market power of the merged enterprise as compared with that of its two components.
The court’s decision, therefore, amounts to a holding that a horizontal merger is per se unlawful even where one party controls but an infinitesimally small percentage of the market and there is a complete absence of indicia indicating a possible anti-competitive impact. The effect is to repeal the principle that “foreclosure of a de minimis share of the market will not tend ‘substantially to lessen competition’ ”, Brown Shoe Co. v. United States, 370 U.S. 294, 329, 82 S.Ct. 1502, 8 L.Ed.2d 510 (1962). Since the record further establishes, as the majority implicitly recognizes, that the merger cannot be upset on the ground adopted by the Commission, i. e., elimination of potential competition, I would reverse the Commission's decision.
I further believe that the majority decision is unfortunate for the reason that by adopting a new ground for affirmance it has indulged in post hoc rationalization of agency action, a course condemned by the Supreme Court as incompatible with the functioning and integrity of judicial review. SEC v. Chenery Corp., 332 U.S. 194, 196-197, 67 S.Ct. 1575, 1760, 91 L.Ed. 1995 (1947); Burlington Truck Lines, Inc. v. United States, 371 U.S. 156, 168-169, 83 S.Ct. 239, 9 L.Ed.2d 207 (1962). The Hearing Examiner limited himself to a finding that the merger might lessen competition in the cabinet hardware market because of the elimination of Stanley as a potential competitior.1 Upon review his decision was summarized by the Commission as being based on this potential competition ground.2 After a lengthy discussion *510of the competitive effects of the Amerock acquisition the Commission concluded:
“Accordingly, we conclude that the fact of its entry by acquisition eliminated the substantial possibility that Stanley itself would become a significant competitor in the cabinet hardware market, with all of the increase in competitive vitality which Stanley’s presence would have implied.” (Appendix at 128).
Apparently recognizing the weakness of the case based upon elimination of potential competition the majority now turns to elimination of actual competition as a ground for affirmance. This substitution of a horizontal basis, not invoked by the Commission, constitutes a departure from well established principles of review.
It is true that in the course of its discussion the Commission alluded to the degree of concentration in the cabinet hardware market, citing some authorities which refer to horizontal acquisitions, and stating that “Stanley’s acquisition did or could contribute to the degree of competition prohibited by the statute.” However, it is apparent that these passing references were merely intended to buttress a decision focused entirely on potential competition. If the Hearing Examiner or the Commission had considered the merger unlawful because of the horizontal overlap between Amerock and Stanley, they would have so stated in plain language; only one sentence would have been required. They did not do so, undoubtedly because (as is shown below) the overlap was too miniscule and insignificant to warrant such a finding. Indeed the Commission said, “Thus, it is not the 1% increase in market share brought about by the Stanley-Amerock merger by itself which posits the anti-competitive aspect of this merger.” Although the Commission referred to Stanley’s actual competition as a fact to be considered, the real basis of its decision was its finding of foreclosure of potential competition.
Actual Competition
Even assuming that the Commission’s decision permits our consideration of the horizontal impact of the merger, im-disputed market facts demonstrate that the merger’s effect on actual competition was entirely too diminutive to violate § 7 of the Clayton Act. The majority concludes that since the cabinet hardware industry is concentrated, with four firms controlling 49-51% of the market, a merger of the first ranking firm, Amerock (23%), and the tenth ranking firm, Stanley (1%), though bordering on de minimis, may have a “tipping effect” in the industry, which might tend to lessen competition. In support of this conclusion it relies upon an isolated report by subordinate marketing personnel employed by Stanley, prepared on June 10,1964, and invokes Supreme Court decisions outlawing mergers by giants of industry in the fields of beer, aluminum, containers, shoes and banking, where there was either a higher market concentration or a trend toward such concentration in the industry under review.3
At first blush this traditional approach seems reasonable enough, particularly since antitrust law, at least with respect to the legality of mergers, has been characterized by decisions based on differ-*511enees in degree rather than upon sharp, crystallized lines. See United States v. Philadelphia National Bank, 374 U.S. 321, 362, 83 S.Ct. 1715, 10 L.Ed.2d 915 (1963). However, we are not here dealing with industrial giants threatening to gobble up lesser competitive fry but with modest participants in a relatively small industry. Unlike the cases relied upon by the majority, there is here a complete absence of any trend toward concentration. Although the industry is somewhat concentrated, other characteristics of the market, including the existence of a large number of strong and thriving competitors, demonstrate that the instant merger cannot have the possible anti-competitive impact attributed to it. As for the evidence relied on by the majority, it is undisputed that the Stanley task force report referred to by it was neither seen nor approved by Stanley’s management. On the contrary, the possibility of entry by either acquisition or by internal expansion was thoroughly rejected. On this record, therefore, I fail to find substantial evidence that would adequately sustain the view that the merger posed actual anti-competitive consequences which might tend to lessen competition. On the contrary, undisputed evidence with respect to competitive realities of the relevant market, including an overlap of less than 1% between the merging parties, compels me to conclude that neither the actual nor the potential competitive effects of the merger are unlawful and it should therefore be upheld.
Product markets and market percentages, whether or not stipulated, are of significance in determining the probable effect of a merger on competition where they reflect actual competition between the parties. But to determine whether a merger forecloses actual competition, one looks to the actual competitive facts (the competitive overlap between the parties) within the given product market rather than halt at overall market percentages and look no further. This process does not, as the majority suggests, change the parties’ stipulation as to the overall market, or create distinct lines of commerce or sub-markets. It merely judges actual competitive effects according to economic realities within the stipulated market rather than ignore undisputed competitive facts of record.
Applying this standard the record is clear that in this case the stipulated overall market percentages, while useful for some purposes, do not reflect the actual competition, i. e., the competitive overlap, between the parties within the stipulated market. Undisputed facts with respect to the actual competition between Stanley and Amerock reveal, on the contrary, that the competitive overlap between them is far less than the 1% accepted by the majority and that the merger is thus governed by the accepted principle that “foreclosure of a de minimis share of the market will not tend ‘substantially to lessen competition,’ ” Brown Shoe Co. v. United States, 370 U.S. 294, 329, 82 S.Ct. 1502, 1526, 8 L.Ed.2d 510 (1962). The de minimis impact of the Stanley-Amerock merger is due mainly to three factors: (1) the differing types of business conducted by each of the two participants, which for the most part do not compete, (2) the differing channels of distribution used by each, which likewise do' not substantially compete, and (3) the fact that size does not carry any competitive advantages in this industry, where small companies thrive and have increased in number.
To appreciate the diminutive effect of the merger under review, certain facts about the industry, in addition to those described by the majority, should be known. Cabinet hardware comes in two kinds, (1) Residential, which is used in both houses and apartments, and (2) Architectural (Institutional) which is used in commercial or institutional structures. Residential hardware is highly stylized. Its production requires die-easting machinery, which is used to fashion intricate ornamentation. Architectural cabinet hardware, on the other hand, is primarily functional in design. Since sturdiness and durability are *512paramount, it is generally stamped out of metal.
Residential cabinet hardware is the vastly more significant part of the cabinet hardware industry, accounting for approximately 90-95% of the $76-$80 million total market — or about $72 million. The architectural market constitutes the remaining 5-10%, or approximately $8 million.
Distribution of cabinet hardware moves through six channels:
(1) Full Line Wholesalers, which carry predominately residential cabinet hardware for sale to retail hardware and department stores and, to a lesser extent to local builders and contractors ;
(2) Specialty Wholesalers, which specialize in either architectural or residential cabinet hardware and sell to lumberyards, retail hardware stores, cabinet shops, and to small companies which build and sell finished cabinets, including the attached hardware;
(3) National Accounts (e. g., Sears, Montgomery Ward, J. C. Penney), which purchase residential cabinet hardware directly from the manufacturer and distribute it through their own retail outlets;
(4) Original Equipment Manufacturers of Residential Cabinets (Residential OEM’s), which sell finished kitchen and bathroom cabinets. These companies are larger and more likely to be automated than residential cabinet shops; they avoid the wholesaler by purchasing directly from the manufacturer;
(5) Original Equipment Manufacturers of Architectural Cabinets (Architectural OEM’s) the analogue to Residential OEM’s in the architectural market. They purchase the hardware directly from the manufacturer and in turn sell finished institutional cabinets primarily to contractors for institutional structures;
(6) Contract Hardware Distributors, which sell the whole range of architectural hardware to institutional builders, generally on a bid basis.
Residential cabinet hardware is distributed through Full Line Wholesalers, National Accounts, Specialty Wholesalers and Residential OEM’s. The latter two are used to distribute % of all residential cabinet hardware. Architectural hardware, on the other hand, is distributed through Specialty Wholesalers, Architectural OEM’s and Contract Hardware Distributors. Each channel is aimed at particular kinds of ultimate buyers. There is little inter-channel competition. Thus, the method of distribution is crucial and represents an important gauge by which to measure competitive impact.
Ameroek is the largest firm in the industry with approximately 23% of the market in 1965, the year immediately preceding the merger. Its 1965 sales totalled $18,218,474, with all but approximately $200,000 derived from the residential market. In contrast, Stanley’s 1965 sales in the cabinet hardware industry totalled $814,000 or only 1% of the cabinet hardware market. More crucial, however, is the fact that only $200,000 of this figure, was derived from the residential market; the remaining $614,000 stemmed from architectural cabinet hardware sales.
Even these low figures do not accurately reflect Stanley’s minimal competitive impact. Stanley owned no die-casting machinery- — a sine qua non for full scale activity in the residential market. Several of its attempts to expand its share of the residential market had failed, giving it a poor reputation with distributors of residential hardware (i. e., Full Line Wholesalers, National Accounts, Specialty Wholesalers, and Residential OEM’s). In fact, Stanley’s sales in the cabinet hardware market were on decline. Thus, within the boundaries of the cabinet hardware market, which mark the outer limits of the “area of effective competition” under review, Brown Shoe Co. v. United States, 370 U.S. 294, 324, 82 S.Ct. 1502, 8 L.Ed.2d 510 (1962) (quoting from United States v. E. I. Du Pont De Nemours & Co., 353 U.S. 586, 593, 77 S.Ct. 872, 1 *513L.Ed.2d 1057 (1957), Stanley was definitely not. a significant competitive force.
To determine whether the joinder of the market shares held by Stanley and Amerock posed an anticompetitive threat one must look beyond wooden percentage figures to live economic realities in the market itself, including such indicia as “reasonable interchangeability of use or the cross-elasticity o.f demand between the product itself and substitutes for it . the product’s peculiar characteristics and uses, unique production facilities, distinct customers, distinct prices, sensitivity to price changes, and specialized vendors.” Brown Shoe Co. v. United States, 370 U.S. 294, 325, 82 S.Ct. 1502, 1523, 8 L.Ed.2d 510 (1962).
Applying these practical criteria, the record is clear that the percentage of actual competitive overlap between Stanley and Amerock prior to the merger was far less than Stanley’s 1% of the overall cabinet hardware market. Stanley’s sales of residential cabinet hardware, $200,000 out of total residential cabinet hardware sales of approximately $70,000,000, amounted to about % of 1% of the residential market, which is de minimis by any standard. Moreover, when one takes into consideration the channels of distribution used by the two parties to the merger, as required by Brown Shoe’s mandate to consider “distinct customers” and “specialized vendors,” the level of competition between them further decreases. Of Stanley’s $200,000 derived from residential cabinet hardware sales, $8,041 was distributed via a single National Account, with the remainder directed to hardware and lumber outlets, mostly through Full Line Wholesalers. In contrast, %’s of Amerock’s residential sales were through OEM’s and Specialty Wholesalers, with % going to Full Line Wholesalers and National Accounts. Thus, while there was a $20,000 competitive overlap, Stanley afforded no competition to Amerock’s major channels of distribution.
As for the vastly less important area of institutional sales, which comprise only 5% to 10% of the aggregate cabinet hardware market, Stanley’s percentage share is coneededly higher.4 5However, its larger share is misleading. Examination of competition within the relevant channels of distribution reveals that while Amerock’s sales of institutional hardware totalled $200,000 and Stanley’s were $614,000 each sold through noncompeting channels of distribution except for an overlap of $84,000.®
In summary, the total actual competition between Stanley and Amerock amounted to an overlap o.f only $284,000 ($200,000 of residential sales and $84,-000 of institutional sales), or approximately .35% of the market. Even if we disregard other relevant factors (discussed below) which militate against the likelihood of competition being diminished, the percentage of the market affected by the merger is simply too small. See, e. g., Tampa Electric Co. v. Nashville Coal Co., 365 U.S. 320, 329-334, 81 S.Ct. 623, 5 L.Ed.2d 580 (1961) (under analogous language of § 3 of the Clayton Act); American Smelting and Refining Co. v. Pennzoil United, Inc., 295 F.Supp. 149, 157 (D.Del.1969) (1% increase in concentrated industry reduced in significance because of the nature of industry); United States v. First National Bancorporation, Inc., 329 F.Supp. 1003, 1019 (D.Colo.1971), juris. noted, 407 U.S. 902, 92 S.Ct. 2431, 32 L.Ed.2d 679 (1972) (even 1.2% insubstantial).
Other indicia traditionally considered in determining the legality of a merger allegedly violative of § 7 of the Clayton Act confirm that the impact here was too minuscule to be of antitrust significance. *514For example, the instant merger completely lacks a factor found to be of controlling significance in most leading horizontal cases,6 i. e., an industry-wide trend toward concentration either before or after the merger. It is uncontested that Amerock’s share of the market remained constant through the 1965-1968 period. Moreover, there is no evidence of any other mergers between manufacturers in the industry at any time prior to or since the Amerock-Stanley merger.
Competition in the cabinet hardware industry has been vigorous, due in no small measure to the fact that it is a business in which numerous small companies, which can cater effectively to frequently changing stylistic demands of a specialized market, thrive. Thus size offers no effective competitive advantages. The market structure, regardless whether it be characterized as a “loose” oligopoly, satisfies “Congress’ desire to promote competition through the protection of viable, small, locally owned businesses”, Brown Shoe, supra, 370 U.S. at 344, 82 S.Ct. at 1534. In 1965 the four (4) largest firms comprised approximately 50% of the market, and the nine (9) largest totalled 65%. However, smaller firms, operating within limited geographical areas, have continued to offer vigorous competition, often enjoying a lion’s share of the business within their respective locales. In fact, during the period 1963-1968, the number of firms in the market actually increased to a total of 97.7 At the same time prices and profits remain low. Such modest price increases as have occurred have been attributable largely to inflation, and have been at a slower pace and lower rate than prices of other building products. There is no evidence of price leadership or price uniformity. Nor is there any showing that the merger might tend to maximize profits for the members of the so-called oligopoly.
The healthy competitive climate in the industry may also be attributed in part to other factors. Unlike other markets (e. g., Clorox, beer, containers, banking) advertising plays but a minor role in the sale of cabinet hardware. Hence, a large advertising budget that might be provided by a financially powerful entrant is superfluous. This unusual characteristic is due in part to the highly specialized, style-oriented nature of cabinet hardware and to the fact that before it reaches the consumer it has been incorporated into a cabinet which is usually sold to the consumer without reference to the brand identity of the hardware forming part of it. From the standpoint of the hardware manufacturer, the significant buyers are the different types of distributors described above, which are relatively few in number. In the dominant residential hardware market, where style is a vital competitive factor, the small manufacturer, offering aesthetically pleasing designs to distributors within a limited geographical area, can successfully compete without an enormous expenditure on consumer advertising that would be required to sell a homogeneous product purchased directly by the consuming public, such as Clorox.
Of further competitive significance is the fact that among the many successful smaller manufacturers in the cabinet hardware market, several possessed the resources to expand in size if they considered such a move to be economically advantageous. At least four were subsidiaries or divisions of large diversified companies with substantial capital.8 In *515addition, several leaders in the manufacture of decorative furniture hardware, with die-casting facilities capable of being used to make residential cabinet hardware virtually identical with that made by those already in the field, had entered the residential hardware market, and others stood at its edge.
In such a commercial milieu, marked by robust and energetic competition, it is not surprising to find that Stanley’s prior attempts at full-scale entry into the residential hardware market, where over 90% of all cabinet hardware is sold, had proved a failure. Indeed its sales in the hardware industry generally were on the decline.
The decisions cited by the majority in support of its conclusion that the horizontal impact of the merger might substantially lessen competition are all clearly distinguishable in legally significant respects. To the extent that “substantial” lessening of competition is measurable in terms of gross dollar volume Stanley’s $814,000 annual sales of cabinet hardware, which involved a competitive overlap of $284,000, stands like a tiny pigmy in sharp contrast to the huge overlaps of $750,000,000 in United States v. Philadelphia National Bank, 374 U.S. 321, 331, 83 S.Ct. 1715, 10 L.Ed.2d 915 (1963); $79,000,000 in United States v. Continental Can Co., 378 U.S. 441, 450, 84 S.Ct. 1738, 12 L.Ed.2d 953 (1964); $84,000,000 in United States v. Von’s Grocery Co., 384 U.S. 270, 297, 86 S.Ct. 1478, 16 L.Ed.2d 555 (1966); $2,200,000 in United States v. Aluminum Co. of America, 377 U.S. 271, 84 S.Ct. 1283, 12 L.Ed.2d 314 (1964); and many millions of dollars in Brown Shoe Co. v. United States, 370 U.S. 294, 82 S.Ct. 1502, 8 L.Ed.2d 510 (1962); and in United States v. Pabst Brewing Co., 384 U.S. 546, 86 S.Ct. 1665, 16 L.Ed.2d 765 (1966), where the precise dollar volumes are not reported. Taking into account that the relevant market in the present case is the entire United States a holding that a $284,000 competitive overlap constitutes substantial lessening of competition in violation of § 7 of the Clayton Act amounts to an unwarranted overkill, hardly the type of target Congress intended to reach.
A comparison of other relevant circumstances confirms the inapplicability to the present case of the “horizontal” decisions relied on by the majority, in each of which the market was tainted by anti-competitive forces not found here. In Brown Shoe, which involved a merger of Brown Shoe Co., the third largest seller of shoes in the United States, and Kinney, the eighth largest company by dollar volume, there had been a steady and long-continued trend toward concentration in the relevant markets, characterized by vertical acquisition by manufacturers of retail outlets, which then sold a greater share of the new parent’s shoes than formerly, and by horizontal mergers of these integrated behemoths. Hand in hand there went an equally relentless decrease in the number of shoe-manufacturing plants. “Brown Shoe was found not only to have been a participant, but also a moving factor, in these industry trends.” 370 U.S. at 302, 82 S.Ct. at 1511. In six cities the combined share of children’s shoe sales by the merged companies, Brown Shoe Co. and Kinney, exceeded 40%, and in one city their combined share of women’s shoe sales exceeded 57%, 370 U.S. at 343.
In Philadelphia National Bank, supra, the merger, which was between the second and third largest competitors in the market area, would have resulted in one bank controlling 34-36 % of a market in which the four largest banks would control 77-78%. Here also there had been a market trend toward concentration, with the number of competitors plummeting and the merger participants repeatedly playing a major role in that trend. The Court observed that the number of commercial banks had declined from 108 in 1947 to 42 in 1962, with *516Philadelphia National Bank having acquired nine formerly independent banks and Girard, the other party to the merger, six.
In Continental Can Co., supra, the top two firms, of which Continental Can was one, controlled nearly 49% of the $3 billion market (combined metal-glass containers) and the top six firms controlled 70.1%. Continental had a history of growth by horizontal acquisitions, having acquired no less than 35 firms in various interrelated container markets since 1913. 378 U.S. at 445, 84 S.Ct. 1738. Lastly, the company acquired under the challenged merger, Hazel-Atlas, ranked sixth largest with 3.1% of the market and had competed vigorously against Continental Can, competition which would be eliminated by the merger.
Von’s Grocery, supra, which involved the third and sixth ranking firms in the market, is another example of one more merger in an industry marked by a precipitous trend toward increasing concentration of ownership, in which the larger firms played an active role in the continuing acquisition process. Of these firms the present participants were two of the most aggressive, each having doubled in size during the ten years immediately preceding the merger under attack.
Another decision referred to by the majority, United States v. Pabst Brewing Co., 384 U.S. 546, 86 S.Ct. 166, 16 L.Ed.2d 765 (1966), also cannot fairly be relied upon as governing the present case. It involved the merger of two huge brewers, which competed against each other in 40 states, in “an industry marked by a steady trend toward economic concentration”. 384 U.S. at 550, 86 S.Ct. at 1668. Furthermore, there had been a steady increase in the size of the industry leaders, with a corresponding decrease in the overall number of competitors. The total number of competitors selling beer had declined from 206 to 162 in the five-year period from 1957 to 1962. Once again a strong anti-competitive trend had to be curtailed in its incipiency.
Lastly, the majority’s heavy reliance upon United States v. Aluminum Co. of America, 377 U.S. 271, 84 S.Ct. 1283, 12 L.Ed.2d 314 (1964), amounts to a wooden-like acceptance of substantially similar market percentage figures without making an analysis of the market structure to determine the true significance of the figures. Although the acquisition by Alcoa, with 27.8% of the aluminum conductor market, of Rome, with 1.3%, appears at first flush to be superficially similar as the merger here under review, a glance at the market there reveals how vastly different it is from that here. Unlike the present case, the top two competitors, including Alcoa, the industry leader, controlled 50% of the market and nine firms controlled 95.5%, foreclosing all but a tiny fraction from other competitors. More important, there had been a rapid trend toward concentration. “The absorption of Rome by Alcoa was one of the five acquisitions by producers of primary aluminum since 1957. . . . These moves, and the threat they were thought to pose, were specifically identified as factors influencing Alcoa’s 1959 decision to acquire Rome. . . . As a result of this series of mergers, there now remain only four nonintegrated fabricators of aluminum conductor whose individual shares of total industry production amounted to more than 1 %.” 377 U.S. at 279, n. 6, 84 S.Ct. at 1288. Since Rome had been an aggressive competitor, it is not surprising to find a judicial willingness to curtail the predatory conduct of the industry’s leader, a giant by almost any standard, in attempting to absorb it. No case has gone beyond Alcoa. In view of the disparity in market concentration and the insubstantiality of the competitive overlap between Stanley and Amerock, which amounts to less than % of 1% of the stipulated market, extension of Alcoa to the instant facts is unwarranted.
Thus, even assuming that more stringent rules may be applied to horizontal mergers, because of the greater probability of their having an anti-corn-*517petitive effect, than that to vertical or conglomerate mergers, in this case, unlike those relied upon by the majority, we have undisputed “evidence clearly showing that the merger is not likely to have such anticompetitive effects.” United States v. Philadelphia National Bank, supra, 374 U.S. at 321, 83 S.Ct. at 1718.
Potential Competition
Having concluded that the Commission’s decision should not be affirmed on the ground adopted by the majority, i. e., that the horizontal aspects of the merger may tend substantially to lessen competition, there remains the question of whether it should nevertheless be sustained on the ground advanced by the Commission, i. e., that the merger might substantially lessen competition by eliminating Stanley as a potential significant competitor in the manufacture and sale of residential cabinet hardware in that Stanley might have expanded either internally or by “toehold” acquisition of a smaller manufacturer, and that the merger therefore had the effect of increasing barriers to entry, and of destroying Stanley’s disciplining effect at the edge of the market.
Although the majority opinion purports not to reach the potential competition issue, it does not hesitate to denigrate the Commission’s position in several respects.9 In these views I concur. It further appears that none of the grounds advanced by the Commission in support of its potential competition theory is supported by substantial evidence. Indeed, in reviewing the record we are confronted with a rather extraordinary infirmity, bordering upon a denial of due process. The Commission’s case before the Hearing Examiner rested heavily upon the testimony of officials of two of Stanley’s competitors, supported by documents from the files of Stanley and Ameroek. The two witnesses testified over Stanley’s objection that it had not been allowed the opportunity to complete relevant discovery that would be material in its cross-examination of them. Nevertheless their testimony was taken and Stanley was later refused the right to recall the witnesses for the purpose of cross-examining them with evidence that had been adduced through intervening discovery. The Examiner’s decision relied heavily upon the testimony of the two witnesses, which he quoted at length. When the unfairness of this procedure was raised upon appeal to the Commission, it decided to treat the testimony of the Commission’s witnesses as stricken rather than permit cross-examination, with the result that it relied solely upon documentary evidence. Undisputed facts of record demonstrate that this reliance on the documentary evidence was improper, especially in view of the inconclusiveness of its contents.
The linch-pin of the Commission’s opinion is its conclusion that but for the merger it was reasonably probable that Stanley would have entered the residential cabinet hardware market by internal expansion. This finding was bottomed entirely upon a series of reports and studies prepared by subordinate staff marketing personnel in the Stanley organization, which considered the possibility of internal expansion substantially prior to Stanley’s acquisition of Ameroek. As is often the case in large organiza- ' tions, including governmental agencies such as the Federal Trade Commission and the Department of Justice, subordinates who are not invested with any decision-making responsibility will sometimes make colorful statements that do not reflect the views of the management. Ever since the famous “charnel house” letters used in United States v. Hartford-Empire Co., 46 F.Supp. 541 (N.D.Ohio 1942), modified, 323 U.S. 386, 65 S.Ct. 373, 89 L.Ed. 322 (1945), and the “Treanor” letter in Aetna Portland *518Cement Institute v. FTC, 157 F.2d 533, 577 (7th Cir. 1946), reversed, 333 U.S. 683, 68 S.Ct. 793, 92 L.Ed. 1009 (1948), such gems have been seized upon as the darlings of the government investigator. Even in cases where the contents have been unauthorized and indeed are directly contrary to management decisions, they are paraded as if they had been approved. That is the case here.
Out of scores of studies prepared by low-level members of the staff of Stanley’s Hardware Division, some of whom did not even have decision-making authority within that division and none of whom had the power to determine whether or not Stanley should expand its residential cabinet hardware production or acquire a competitor, the Commission has culled a few juicy comments such as the suggestions that acquisition of another manufacturer “Would knock out a competitor” or enable it “To obtain a dominant role in the hardware market.” The Commission has committed the error of repeatedly mischaracterizing these studies as “management documents,” “management reports,” and “decisions made by Stanley’s management,” ignoring the undisputed fact that the documents were, by and large, neither seen nor approved by Stanley’s management and that the management made unequivocal decisions to the contrary, which post-dated the studies and predated its acquisition of Ameroek.
Even if some weight is given to the studies of Stanley subordinates relied upon by the Commission, a careful review of all such reports (as distinguished from the few quoted from by the Commission) reveals beyond question that their overall thrust was to reject internal expansion into the residential market and to concentrate on Stanley’s exising product line, which was devoted mainly to architectural hardware. Furthermore, when action was finally taken by those corporate officers responsible for determining the company’s basic policies, the decision was to reject the views of subordinates who advocated expansion into the residential field and to adopt the view that the Hardware Division should concentrate on improving its existing architectural hardware line. More specifically, although the so-called “task force report,” dated June 10,1964, upon which the Commission relies, made by a group of relatively low-level staff personnel in the Hardware Division, suggested that acquisition of two cabinet hardware manufacturers, Ajax or Jaybee, be considered, it explicitly urged that internal expansion be rejected as unprofitable. Thereupon the Department Managers of the Hardware Division, albeit still a stratum below the decision-making level, rejected entry by either route (acquisition or internal expansion), taking the position that internal expansion in the residential hardware field was not justified after weighing the size of the investment that would be required against the losses or limited profit that could be anticipated.
July 10, 1964, was the first time when Stanley’s management became involved in the review of the question whether the company should expand its cabinet hardware division, acquire a competitor, or take some other course. On that date the Department Managers of the Hardware Division advised D. W. Davis, Executive Vice President, of their conclusions and recommended that the Hardware Division limit itself to concentrating on development of an improved line of architectural hardware, for which die-casting equipment would not be needed. In the meantime independent management consultants, Stewart, Douglas & Associates, were retained to identify new product areas in which the Hardware Division might expand. On December 23, 1964, it rendered its “think-tank” report which reinforced the conclusions of the Hardware Division by recommending against product expansion.
Armed with the foregoing reports and recommendations Davis held a meeting on January 8, 1965, with the various Managers of the Hardware Division, at which the final basic policy decision was reached that instead of expanding in the cabinet hardware field Stanley would *519concentrate on improving its existing product lines and maximizing sales and profits from those lines. Toward that objective responsibility for cabinet hardware was transferred to the Manager responsible for Contract (architectural) Products.
Despite overwhelming proof to the contrary the Commission, apparently proceeding on the erroneous assumption that the burden was on Stanley rather than on Commission counsel, see United States v. Penn-Olin Chemical Co., 246 F.Supp. 917, 919 (D.Del.1965), aff’d by an equally divided court, 389 U.S. 308, 88 S.Ct. 502, 19 L.Ed.2d 545 (1967), concluded, notwithstanding Stanley’s unequivocal rejection of acquisition or expansion, that “[tjhere is no persuasive indication in the record that respondent irrevocably abandoned either of these alternatives [acquisition of a competitor or internal expansion into resident hardware] in January of 1965.” (Appendix at 126). This crucial finding, unsupported by any substantial evidence, reflects administrative zeal which, while generally desirable, was misdirected in the present case. In a belated effort to support it with record proof, Commission counsel point to some rather euphemistic generalities in the Hardware Division’s Long Range Plan, dated June 17, 1965, and in accompanying working papers, to the effect that during the latter part of the next five years the Division would “give increasing attention to the residential market.” These statements by lesser personnel, however, are a far cry from reversal or modification of the final decision taken by Stanley’s management on January 8, 1965. It was against this background that executives of Amerock in October, 1965, approached Stanley’s Executive Vice President regarding the possibility of a merger. Faced with “a whole new ball game" Stanley for the first time reconsidered its earlier decision and finally decided in favor of the newly proposed merger.
Viewed in the light of Stanley’s January 8, 1965 decision, it is clear that there is no substantial evidence to support the Commission’s conclusion that but for the merger Stanley was a “reasonably probable” entrant into the residential hardware market. The Commission’s evidence is far less persuasive than that in Penn-Olin, supra,, where the district court concluded on remand that despite Olin’s staff reports recommending entry
“The government has failed to sustain its burden of establishing by a preponderance of the evidence that . there would have been a reasonable probability that Olin would have constructed a chlorate plant in the Southeast.” United States v. Penn-Olin Chemical Co., 246 F.Supp. 917, 928 (D.Del.1965), aff’d by an equally divided court, 389 U.S. 308, 88 S.Ct. 502, 19 L.Ed.2d 545 (1967).
The Commission next asserts that the elimination of Stanley as a potential competitor may have substantially reduced competition by removing it as a disciplining force on the edge of the market. To exercise such an effect a minimum requirement is that “the merging firm at the edge of the market must be recognized by those in the market as the most likely entrant or one of a very few likely entrants. . . . ” Turner, Conglomerate Mergers and Section 7 of the Clayton Act, 78 Harv.L.Rev. 1313, 1363 (1965). A careful search of the record discloses no evidence that firms in the residential market recognized Stanley as a likely entrant.10 On the contrary, the record indicates that Stanley’s position on the edge of the residential market had no competitive effect on those in the market. This is in dramatic contrast to potential competition cases in which the doctrine has been applied. See, e. g., United States v. El *520Paso Natural Gas Co., 376 U.S. 651, 659, 84 S.Ct. 1044, 12 L.Ed.2d 12 (1960); FTC v. Procter & Gamble Co., 386 U.S. 568, 87 S.Ct. 1224, 18 L.Ed.2d 303 (1967).11 Thus, another peg upon which the Commission relies must fall.
The Commission’s conclusion that the instant merger raised barriers to entry is likewise not supported by substantial evidence. The Commission found three barriers to entry: (1) the cost of the initial investment in die-casting equipment and design which would be required (approximately $600,000); (2) the need for a full line of up-to-date product styles; and (3) the cost of sales promotion and advertising in trade journals. At the same time the Commission rejected the Hearing Examiner’s reliance as a factor on the difficulty of securing channels of distribution.
There is no direct or circumstantial evidence in the record that the $600,000 initial start-up costs, $400,000 of which might be needed to acquire machinery, would be increased by the Stanley-Amerock merger. Nor is there any substantial evidentiary support for the conclusion that the need for product differentiation might be increased. Stanley was not a style oriented company; its history was grounded in the production of non-ornamental hardware and tools. Finally, there is no substantial evidence to support the conclusion that the merger would increase the need for sales promotion and advertising. The Commission conceded that “advertising at the consumer level may be small” but relied on the cost of advertisements in trade publications — somewhere in the vicinity of $50,000-$100,000. These figures are a far cry from the $80,000,000 per year found to be a decisive element in the landmark case of FTC v. Procter & Gamble Co., 386 U.S. 568, 87 S.Ct. 1224, 18 L.Ed.2d 303 (1967); see also General Foods Corp. v. FTC, 386 F.2d 936 (3d Cir. 1967), cert. denied, 391 U.S. 919, 88 S.Ct. 1805, 20 L.Ed.2d 657 (1968). Furthermore, the Commission fails to point to any evidence indicating that the instant merger would increase an entrant’s need for, or cost of, advertising. What is crucial is that unlike Procter & Gamble and General Foods, which dealt with consumer items where massive advertising was the crucial competitive factor, the role of advertising and sales promotion here is of minor significance and hence does not pose a barrier to entry.
The concept that a merger may operate to raise barriers of entry is based upon the assumption that price levels and profit margins in the industry would, but for the merger, be sufficiently high to overcome the cost of entry and thus attract new entrants. Here that is not the case. Prices were low, profits were modest at best, and there is no evidence that ease of entry, which existed before the merger, was in any way affected by it.
In support of its conclusion that the merger eliminated Stanley’s potential competition, the Commission advanced for the first time a new theory, namely, that but for the merger Stanley might have entered the residential market by a so-called “toe-hold” acquisition, which here means that it might have acquired control over one of the smaller, non-dominant firms in the cabinet hardware market instead of merging with Amerock. This theory was neither suggested in the complaint nor advanced in the hearings before the Examiner. Nor was any mention of it made by the Hearing Examiner. Indeed it was not suggested, either by the Commission or any other body, until a year after the record in the case had been closed and three months after the parties had submitted their briefs to the *521Commission. Then for the first time, the idea was advanced by the Commission in Bendix Corp. v. FTC (FTC Dkt. No. 8739), 3 Trade Reg. Rep. ¶ 19,288 (June 18, 1970), which was reversed by the Sixth Circuit on the ground that “Bendix had no notice that it was charged under the toe-hold theory . . . and . . no opportunity to present evidence in defense against this theory.” Bendix Corp. v. FTC, 450 F.2d 534, 537 (6th Cir. 1971).
Since Stanley was never given timely notice of the toe-hold theory or an opportunity to be heard and to introduce evidence with respect to it, I concur in Judge Kaufman’s view12 that the Commission was precluded from basing its decision on that ground. See Bendix Corporation v. FTC, 450 F.2d 534 (6th Cir. 1971); cf. Rodale Press, Inc. v. FTC, 132 U.S.App.D.C. 317, 407 F.2d 1252 (1968). To countenance such a procedure would be to deny Stanley the due process that is mandated by the Fifth Amendment and by § 5 of the Administrative Procedure Act.13
In any event the Commission’s newly espoused post-hearing “toe-hold” theory is not supported by substantial evidence. The record is clear that the June 10, 1964 recommendation of subordinate Hardware Division personnel that Ajax or Jaybee be acquired was rejected later that month by the Department Managers of that Division, who recommended to D. W. Davis, Executive Vice President, that no such acquisition be made. This recommendation was accepted and approved by the company’s management.
Conclusion
Vigorous enforcement of antitrust laws to bar mergers that may substantially lessen competition is essential to the economic lifeblood of this nation. But Congress has not outlawed all mergers. Nor has it, despite recommendations of some distinguished experts in the field, prescribed arbitrary standards for il-'ogality which would prohibit every merger falling within a precise definition, even those in which the market percentage of one of the parties may be de mini-mis. On the contrary, Congress has recognized that antitrust flexibility is essential, since not all mergers threaten harm to the public, some even have the effect of promoting competition, and the issues turn on an understanding of the complexities of any given market.
The Commission assumes the burden of proving that any merger attacked by it will have the probable effect of substantially lessening competition or tending to create a monopoly. This it has failed to do in the present case. After a Hearing Examiner erred in denying Stanley the right to cross-examine those witnesses relied upon by him as the basis of his decision, the Commission, rather than remand for cross-examination, sought to sustain his decision upon a different evidentiary basis. Apparently recognizing the weakness of the grounds chosen by the Commission, the Court has affirmed on still different grounds. In view of the undisputed evidence as to the extremely limited actual competitive overlap, the vigorous competition in the marketplace, the absence of any trend toward concentration and the small percentage of the market controlled by Stanley (whether it be labelled .35% or 1%), the effect is automatically to hold such a merger illegal, thus in substance declaring it to be a per se violation because of the percentages involved.
Since the Commission’s decision is not supported by substantial evidence, I would reverse.
. The Hearing Examiner concluded:
“The acquisition and merger of Am-erock the leading independent domestic manufacturer and seller by Stanley increased this high level of concentration and may result in a substantial lessening of competition in the said cabinet hardware market because of the elimination of Stanley as a reasonably probable, potentially capable and significant domestic manufacturer and seller in the residential cabinet hardware segment of the said market, and particularly in the decorative residential cabinet hardware sales area of the said market.”
. The Commission described the Hearing Examiner’s Initial Decision as follows:
“Thus, the examiner found that respondent’s acquisition of Amerock lessened competition because (1) absent the merger Stanley would probably have entered the cabinet hardware market on its own; (2) the elimination of Stanley as a potential competitor had the effect of increasing the barriers to entry; and (3) Stanley, as a potential *510competitor, had an influence on the performance of the cabinet hardware market.
“On the basis of these findings, the hearing examiner concluded that the effect of the merger may be to substantially lessen competition in the cabinet hardware market in violation of Section 7 of the Clayton Act and Section 5 of the Federal Trade Commission Act.”
. United States v. Pabst Brewing Co., 384 U.S. 546, 86 S.Ct. 1665, 16 L.Ed.2d 765 (1966); United States v. Aluminum Co. of America, 377 U.S. 271, 84 S.Ct. 1283, 12 L.Ed.2d 314 (1964); United States v. Continental Can Co., 378 U.S. 441, 84 S.Ct. 1738, 12 L.Ed.2d 953 (1964); Brown Shoe Co. v. United States, 370 U.S. 294, 82 S.Ct. 1502, 8 L.Ed.2d 510 (1962); United States v. Philadelphia National Bank, 374 U.S. 321, 83 S.Ct. 1715, 10 L.Ed.2d 915 (1963).
. Stanley’s $600,000 sales constituted 7%% of all institutional cabinet hardware sales; Amerock’s $200,00 equalled about 2%%.
. The $84,000 overlap is based on the fact that Stanley’s sales through Architectural OEM’s was only $44,000 (compared with Amerock’s sales of $160,000) and Amerock’s sales of institutional hardware to Contract Hardware Distributors only $40,000 (compared with Stanley’s sales of $570,000).
. E. g., United States v. Philadelphia National Bank, supra; United States v. Von’s Grocery Co., 384 U.S. 270, 86 S.Ct. 1478, 16 L.Ed.2d 555 (1966); United States v. Aluminum Co. of America, supra; Brown Shoe Co. v. United States, supra; United States v. Continental Can Co., supra; United States v. Pabst Brewing Co., supra.
. Based upon figures published in the trade magazine Kitchen Week.
. National Lock, Tassell, Faultless and Bossick-Sack, for example, are respectively subsidiaries of Keystone Consolidated Industries, Gulf & Western Industries, Bliss & Laughlin Industries, and Stewart-Warner Corporation. In addition Yale & *515Towne, Embart Corp., and Eaton Corp., all large companies produce their own lines of cabinet hardware either directly or through subsidiaries.
. See footnotes 8 and 23 (supra) of the majority opinion. Footnote 8 rejects the Commission’s contention that it proved that Stanley was the most likely potential entrant, that it served to discipline the market and that the merger raised barriers to entry. Footnote 23 similarly disparages the Commission’s toe-hold theory.
. Compare footnote 8 of the majority opinion, which reaches substantially the same conclusion. The Examiner made no finding that Stanley, as a potential competitor at the market’s edge, had any competitive effect upon those in the market, and witnesses representing competitors in the cabinet hardware market testified that it did not have such an influence.
. It may well be, as the majority opinion suggests, that FTC v. Procter & Gamble, 386 U.S. 568, 87 S.Ct. 1224, 18 L.Ed.2d 303 (1967), requires a finding that the acquiring firm was “the most likely entrant.” Id. at 581, 87 S.Ct. 1224. Since we conclude that under any standard there is no substantial evidence to indicate that Stanley exercised a disciplining effect, that issue is not reached. Similarly, we need not discuss the significance of other potential entrants on the market.
. Compare footnote 23 of the majority opinion.
. Section 5 of the APA provides in relevant part:
“Persons entitled to notice of an agency hearing shall be timely informed of . . . (3) the matters of fact and law asserted.” 5 U.S.C. § 554.