SWYGERT, Circuit Judge.

Dean Milk Company (Dean Illinois) and Dean Milk Company, Incorporated (Dean Kentucky) petition to set aside an order of the Federal Trade Commission. The Commission, affirming the initial decision of the examiner in all respects save one, found that Dean Kentucky and Dean Illinois violated Section 2(a) of the Clayton Act, as amended by the Robinson-Patman Act, 15 U.S.C. § 13(a), by discriminating in the prices charged for milk in parts of Indiana and Kentucky between 1952 and 1960. Specifically, Dean Illinois was found to have engaged in proscribed price discrimination at the secondary (buyer) level by virtue of a quantity discount system implemented in the Terre Haute, Indiana market. Dean Kentucky and Dean Illinois were found to have engaged in proscribed territorial price discrimination at the primary (seller) level by virtue both of charging a lower price in the Evansville, Indiana-Henderson, Kentucky market than that charged by them in the Falls Cities (Louisville) market and of implementing a quantity discount system. In addition, both companies were found to have engaged in proscribed price discrimination at the primary and secondary level by virtue of a quantity discount system implemented in the Louisville market. The central issue posed by the Dean companies in this review is whether substantial evidence in the record supports the Commission's findings and conclusions.

I.

Henderson, Kentucky and Evansville, Indiana, themselves separated only by the Ohio River, are approximately 130 miles from Louisville, Kentucky. Before Dean Kentucky commenced operations in September 1952, there existed in Henderson, Evansville, and Louisville an "historical" one-cent per quart differential between the higher priced homogenized and lower priced creamline milk. As between the three cities the price for a quart of milk, whether homogenized or creamline, was one cent less in Evansville than in Louisville while in Henderson it was one cent more than in Louisville. The first action Dean Kentucky undertook upon entering the Louisville market in September 1952 and the Evansville-Henderson market in November 1952 was to uniformly eliminate in each of those markets the preexisting differential between the quart prices of the two varieties of milk. Thereafter, homogenized was sold at the lower creamline price not only by Dean, but also by all the other dairies in those markets. In addition, Dean Kentucky lowered the quart price of homogenized milk in Henderson to that of the new price it had set in Evansville, thereby eliminating the two-cent per quart price differential which had previously existed between those two cities.

Beginning in July 1954 in Evansville and in July 1955 in Henderson, Dean Kentucky introduced a quantity discount system similar to the one introduced at about the same time in the Louisville market. According to the testimony of a Dean Kentucky officer, approximately 95 per cent of Dean's customers in Evansville qualified for the maximum discount whereas approximately 50 per cent of Dean's customers in Henderson so qualified. In both cities, these customers were large chain stores.

On the basis of the foregoing activities, the Commission concluded that Dean Kentucky had engaged in unlawful territorial price discrimination to the detriment of its competitors and competition in the Evansville-Henderson market. The evidence supporting the Commission's conclusion was provided by witnesses from only four dairies, although there were at least nine dairies operating in that market at the time of Dean's entry.

The Commission's first witness was a former officer of Dairy Service, Inc., an Evansville Dairy, which ceased operating and merged into American Dairy, another Evansville dairy, in May 1954. He testified that Dairy Service, which sold approximately 65 to 70 per cent of its volume to retail home delivery customers, "had to reduce our price in order to hold onto our retail business." He attributed Dairy Service's declining profits prior to its cessation of business to "reduced sales and reduced gross profit by reason of receiving one cent less per quart on homogenized milk." He testified that the decision to cease operations was due to a "trend definitely showing where smaller operations were being forced either to sell or quit business" and to declining profits. On cross-examination, he could recall only one wholesale store account that Dairy Service lost to Dean. Moreover, the witness could not testify that Dairy Service lost any home delivery business "directly" as a result of Dean's activities.

Blue Ribbon Dairy, another local Evansville dairy selling at wholesale and retail, ceased operating in September 1953. The president testified that his company served the A P and Kroger stores in the area, losing the latter's business upon Dean's entry. The A P stores continued to carry his company's products along with those of Dean and another dairy. Blue Ribbon also served a local chain accounting for approximately one-third of its total business, to the exclusion of Dean. With respect to Dean's uniform one-cent reduction in the price of homogenized milk, he testified that "it could be the difference between profit and loss."

American Dairy Company was the second largest Evansville dairy, distributing dairy products at wholesale and retail primarily in the Evansville-Henderson market area. According to the testimony of an officer of American, Dean's entry, attended by its elimination of the price difference between homogenized and creamline milk, created "chaos in the market." With respect to the importance of this one-cent reduction in the price of homogenized, he testified, "That's the profit we are making in our business." Upon Dean's introduction of quantity discounts, American, according to the witness, offered its customers similar discounts, seeking thereby to retain business even though charging a lower price. He admitted however, that American's wholesale business increased since Dean entered the market, that Dean did not sell any milk to American's largest wholesale purchaser, and that he could recall only two American accounts which Dean "got in" and served in conjunction with American.

Evidence specifically touching on the Henderson segment of the market came from a single witness, the president of the Henderson Creamery Company, a dairy selling milk at wholesale and retail exclusively in Henderson and surrounding Kentucky towns. In explaining the higher milk price prevailing in Henderson prior to Dean's entry, he testified that Henderson Creamery traditionally purchased milk of higher quality than that purchased by the Evansville dairies. The effect of Dean's entry, according to his testimony, centered around: (1) the consequence of even a one-cent price cut which "just about removes all of the profit and throws you in a loss;" (2) the unfavorable image, "bad eye," and loss of wholesale business resulting from Dean's drop in prices; (3) the increased spread between the home-delivered price and the lower chainstore price causing a loss of home delivery business. Apparently, the loss of wholesale business was only temporary as he later testified that his company's wholesale volume had increased during the period covered by this action. He testified that no wholesale accounts were lost in their entirety by Henderson Creamery to Dean, although in some locations his business was split with Dean and as many as two other dairies. As to Dean's introduction of a quantity discount system, he testified that Henderson Creamery did likewise, "to hold our business, * * * to try to gain back our wholesale business that we had been losing."

Two important elements must be established if the Commission is to prevail in finding a violation of section 2(a) of the Robinson-Patman Act. First, the Commission must demonstrate a discrimination in price "between different purchasers of commodities of like grade and quality." Second, the Commission must demonstrate that "the effect of such discrimination may be substantially to lessen competition or tend to create a monopoly in any line of commerce, or to injure, destroy, or prevent competition * * *." Both questionable analysis and insubstantial evidence detract from the persuasiveness of the Commission's attempt to demonstrate the existence of these two elements.

Initially, we question the rationale by which the Commission concluded that Dean discriminated by charging prices in the Evansville-Henderson market different from those it charged in the Louisville market for like products. Although there were periods both before and after Dean Kentucky commenced operations when the Evansville-Henderson prices were lower than the Louisville prices, the Commission stated:

At all times between 1952 and 1960, respondents' [Dean] prices to their favored customers in Evansville-Henderson were lower than their prices for goods of like grade and quality in Louisville, even though the products sold in Evansville-Henderson were processed in Louisville and thereafter physically transported from the Louisville plant to purchasers in Evansville-Henderson.

With respect to this conclusion, we are at a loss to understand why the Commission glossed over the import of the preentry price differentials existing between Louisville and Evansville and Louisville and Henderson. Before Dean Kentucky's entry, the price for a quart of homogenized milk in Louisville was one cent higher than the price in Evansville and one cent lower than the price in Henderson. Upon entry, Dean uniformly lowered the Louisville and Evansville prices for a quart of homogenized milk one cent to the then prevailing price of creamline. Thus, the price in Evansville remained one cent lower than the price in Louisville, just as it had before Dean's entry.

Dean also uniformly lowered the Henderson price three cents to the new Evansville price. With respect to this reduction, the Commission argues before this court:

Petitioners [Dean] * * * misstate the record when they assert there is no "economic explanation" for the higher prices in Henderson. * * * The Henderson price of one cent a quart more than the Louisville price was, in short, as much a "historical" differential between the two markets as the difference in price between Evansville and Louisville. Petitioners label the Evansville price difference as "historical" and the Henderson price difference as "artificial," since the Evansville price provided petitioners with a convenient pretext for undercutting their rivals' prices in both cities.

Yet the Commission overruled the findings of the examiner and held that the cities of Evansville and Henderson constituted a single market. It ruled, "The Commission will treat these two cities as a single market because of their geographical proximity and the fact that local dairies sold in both cities." These statements demonstrate a contradiction in the Commission's position. For if Henderson and Evansville constituted a single market, as the Commission found, the most logical expectation would be that a single price for a like quantity and quality of milk would prevail throughout that market.

The evidence relied on by the Commission does not demonstrate that this expectation is unwarranted. An officer of Henderson Creamery was the only witness that testified to paying a higher price for raw milk than the other dairies in that segment of the market, attempting thereby to explain the preentry difference in price between Evansville and Henderson. Yet the evidence shows that American Dairy, which sold in both the Henderson and the Evansville segments of the market, was purchasing raw milk at a lower price than that paid by Henderson Creamery. Thus, there could be no justification for American or any other dairy, including Dean, able to purchase raw milk at a price lower than that paid by Henderson Creamery, to charge Henderson purchasers a higher price for a given quantity of milk than it would charge Evansville purchasers.

We do not believe, as apparently the Commission does, that the Robinson-Patman Act was intended to freeze prices at the level which would return a profit to a competitor in a market with the highest costs. For such an interpretation of the act would sacrifice the interests of the consumer and competition to the interests of a single competitor. Moreover, in view of the Commission's treatment of the two cities as a single market, any dairy charging a different price in one of the cities than it charged in the other for a like product would be engaging in price discrimination. Finally, the Commission's failure to consider the effects of the restraining order against Dean, requiring that prices charged in Henderson equal those charged in Louisville, raises a question as to the extent, if any, that Dean discriminated in price between those cities.

Although the Commission concedes that Dean's Louisville reduction in the price of homogenized milk was nondiscriminatory and economically justified, it argues that Dean's same uniform Evansville-Henderson reduction constituted a territorial price discrimination. Sound business and economic reasons underlay Dean's uniform reduction in the price of homogenized milk to the prevailing price of creamline milk in the Louisville market. Dean's explanation for the cut of one cent was that homogenized milk did not cost any more than creamline to manufacture. The evidence in the record supports Dean's explanation.

The Commission characterizes Dean's showing that there was no detectable cost difference between homogenized and creamline as "totally irrelevant" to the propriety of its uniform Evansville-Henderson reduction in the price of homogenized milk. Yet the milk Dean sold in both the Louisville and the Evansville-Henderson markets was processed in the same plant. In light of this fact, the Commission's argument that Dean's uniform elimination of the price differential between creamline and homogenized milk in the Evansville-Henderson market was discriminatory is anomalous. In response to Dean's showing the Commission emphasizes that the Evansville-Henderson dairies' profits were not in excess of normal competitive returns and that the testimony showed that one cent is the "profit on a quart (of milk) in normal operations." The Commission's response is both inadequate and contradicted by the evidence, for the operations of Dean's competitors in that market showed increased profits between 1952 and 1953, despite their meeting of Dean's uniform one-cent reduction.

Because of the legal and economic propriety of Dean's uniform reduction in the price of homogenized milk in Louisville, we fail to comprehend why the same reduction in Evansville-Henderson was not equally proper. The Commission contends that prior to the advent of Dean, no Henderson-Evansville dairies sold in the Louisville market which was "a completely separate market from the Falls Cities [Louisville] market." Underlying its contention must be the view that a multimarket seller cannot enter a new market (Evansville-Henderson) at a lower price than it is charging in any other geographically proximate market (Louisville), irrespective of either the economic justification of such a lower price or the tradition of a differential in the prices charged in the two markets. Cf. Balian Ice Cream Co. v. Arden Farms Co., 231 F.2d 356, 367 (9th Cir. 1955), cert. denied, 350 U.S. 991, 76 S.Ct. 545, 100 L.Ed. 856 (1956). Such a view would discourage the entry into new markets. We do not believe that the Robinson-Patman Act was intended to achieve such anticompetitive results. Likewise, a view of this kind, which tends to discourage new entrants, is difficult to reconcile with the Commission's concern, expressed in this case, over the "concentration of business in the hands of even fewer sellers".

We are aware of the Supreme Court's holding in FTC v. Anheuser-Busch, Inc., 363 U.S. 536, 549, 80 S.Ct. 1267, 1274, 4 L.Ed.2d 1385 (1960), that "[A] price discrimination within the meaning of that provision [Section 2(a)] is merely a price difference." Although there is no question that at times Dean's Evansville-Henderson price was lower than its Louisville price, we cannot accept the Commission's overbroad attack on this difference. Dean did not create the difference by uniformly lowering both the Evansville-Henderson and Louisville prices for homogenized milk. Rather this act, dictated by economic realties, observed without change a preexisting differential between the prices charged in the two markets.

Irrespective of our belief in regard to the existence of a territorial price discrimination, there is insufficient evidence in the record to support the Commission's conclusion with respect to the second element essential to finding a violation of section 2(a). Dean's uniform reduction in the price of homogenized milk in Louisville and Evansville-Henderson, its observance of the preexisting differential in prices between those markets, or its quantity discount system did not cause injury, or the "reasonable possibility" thereof, either to competition or to competitors of Dean in the Evansville-Henderson market. With respect to competitive injury, the Commission said:

[B]etween 1952 and 1960, the Evansville-Henderson market was a market in which profit margins were small and in which the number of competitors was limited. In such a market, the crippling or elimination of even one competitor takes on an added significance.

* * * * * *

The undercutting of prevailing prices by a large, multi-market competitor may well result in the elimination of several competitors, where, as here, the evidence indicates that the smaller competitors were constantly struggling to stay in business. The evidence showed that the larger dairies sustained losses of business and profits and some smaller dairies ceased operation after respondents [Dean] lowered the prevailing prices. We think that the evidence justifies the conclusion that respondents' pricing activities have already contributed to the loss of business and decline in profits sustained by several of the dairies, and the demise of others. Moreover, we think it clear that the conditions extant in the market — the low profit margins and the high mortality rates of the smaller dairies — support a conclusion that there is a reasonable possibility that continued sales by respondents at below cost prices or at prices which even they cannot realize a reasonable profit will cause the demise of other small dairies, with a consequential concentration of business in the hands of even fewer sellers. Moreover, there is a reasonable possibility that such pricing will permanently impair the ability of the remaining dairies to continue competing with respondents through expansion of facilities or intensive advertising or promotional programs.

Several general observations are pertinent before considering in detail the Commission's conclusions respecting competitive injury. The Commission's opinion omits any mention of Beatrice Foods Co., 1965-1967. Transfer Binder ¶ 17,244, at 22,326-22,328 (1965). That case, although brought under section 7 of the Clayton Act, contains a thorough examination of recent developments in the dairy industry. Technological advances, inefficient operations, and increasing costs were emphasized by the Commission in that case as major developments contributing to the decline in the number of small dairies throughout the nation. By failing to consider these developments as they relate to the discontinuance of operations of some of the local Evansville-Henderson dairies after Dean's entry, the Commission's conclusion that Dean's pricing activities in that market were a major contributing factor becomes questionable.

After making a number of preliminary findings, the Commission stated, "[T]he acts and transactions of Dean Kentucky will, * * * be deemed to be those of Dean Illinois;" moreover, Dean Illinois was viewed as the "perpetrator of the alleged offenses" due to its dominance over Dean Kentucky. Even assuming the validity of the Commission's identification of the two companies, we are unpersuaded by the Commission's argument premised on that identification:

Respondents' [Dean] pricing activities in the Evansville-Henderson market closely resemble the classic example of a large, multi-state seller entering a market in competition with smaller, local sellers and using its overall superior size, financial reserves, and higher prices in other markets as a crutch to enable it to undercut the prices of local competitors.

In an effort to bolster this argument, the Commission observed, "Patently, respondent's operating losses during 1958 and 1960 were recouped either from cash reserves or from some other market, such as Louisville, where their operations were showing profits." There is no evidence, however, that Dean's higher prices in Louisville (when in fact they were higher) served as a "crutch to enable it to undercut the prices of local competitors." Likewise (with the exception of Dean Illinois' practices in Terre Haute), the record is barren of any evidence as to prices charged by Dean Illinois in the many markets it served, let alone that such prices supported the "lower" prices in the Evansville-Henderson market. Because of these evidentiary deficiencies, the Commission's "classic example" argument is substantially weakened.

In its discussion of both the territorial discrimination and the existence of competitive injury in the Evansville-Henderson market, the Commission failed to distinguish between the pre-quantity-discount period (1952-1954) and the post-quantity-discount period (1954-1960). We believe this treatment impedes a rational analysis of Dean's pricing activities when considered either separately or in combination. Apparently, the Commission viewed Dean's quantity discount system in the Evansville-Henderson market as a factor serving to aggravate further the anticompetitive situation resulting from Dean's alleged territorial price discrimination. To impart clarity to the testimony regarding competitive injury, we will first consider the testimony of the two witnesses whose companies ceased operations before the introduction of quantity discounts in 1954. We will then discuss the testimony of the remaining two witnesses whose companies operated in the market throughout the period covered by this action.

In concluding that the requisite competitive injury in the Evansville-Henderson market resulted from or was "reasonably possible" because of either Dean's territorial price discrimination or its quantity discount system, the Commission relied heavily on the fact that five local dairies ceased operating after Dean's entry into that market. This fact, standing alone, would seem to suggest not only increasing concentration in the market through actual present injury to competitors, but also "possible" future consequences of the same nature. Our examination of the record, however, does not disclose substantial proof establishing the existence of a causal connection between Dean's activities in that market and the disappearance of those competitors.

The record contains testimony concerning only two of the five dairies that ceased operation. Both of these, Blue Ribbon and Dairy Service, went out of business (the former in September 1953, the latter in May 1954) before Dean introduced quantity discounts in either Henderson or Evansville. Thus, their testimony is probative, if at all, only of the effect of Dean's uniform reduction in the price of homogenized milk and its observance of the preexisting differential in prices between the markets in Louisville and Evansville-Henderson.

With respect to Dean's reduction, the witness of Blue Ribbon testified that one cent could be the difference between a profit and a loss. Three other Evansville dairies for which profit figures were available, however, all showed significant increases in both their total annual sales and their net profits from 1952 to 1953. Because the Commission failed to introduce any operating figures for Blue Ribbon, we are left to speculate as to both the magnitude of its losses and the possible effect other factors, such as rising expenses, had on its operations. From these facts, coupled with the Commission's awareness that technological changes and inefficiency were causing many small dairies (of which Blue Ribbon was one) to cease operations, the conclusion that Blue Ribbon lost money and went out of business could more reasonably be attributed to its small size and inefficiency than to Dean's uniform price reduction in the Evansville-Henderson market. Moreover, there is no evidence that Dean's difference in price between Evansville-Henderson and Louisville caused Blue Ribbon's difficulties. Yet evidence of such a causal relation is essential to bring a price discrimination within the proscription of the Robinson-Patman Act. Apparently, the Commission relied on its "classic example" argument to fill this evidentiary void. As we previously indicated, however, there is also insufficient evidence to support that argument.

The evidence with respect to Dairy Service raises the same problem concerning the establishment of a causal relation between Dean's territorial price differences and that dairy's difficulties. The testimony of the Dairy Service witness concerning the adverse effect of matching Dean's one-cent price reduction cannot be reconciled with his company's increased sales and profits from 1952 to 1953. Since he could testify to the loss of only a single wholesale account to Dean (allegedly miniscule in size), and further since 70 per cent of his dairy's business was home delivery, it is reasonable to assume that the losses in sales and profits suffered in 1954 were attributable to a decline in home delivery business. Any loss of home delivery business, however, could not possibly have been caused by Dean's uniform price reduction in the Evansville-Henderson market since Dairy Service lowered both its home delivery and wholesale prices in response to Dean's action. Likewise Dairy Service's losses could not be attributable to Dean's observance of the preexisting differential in prices between Evansville-Henderson and Louisville.

Evidence of a more likely reason for Dairy Service's merger with American Dairy appears in the record. The Dairy Service witness testified to the "trend definitely showing where smaller operations were being forced either to sell or quit business. * * *" He further offered that "it was much more economical to operate the combined business, the milk operation, in the main plant" as the reason why American Dairy discontinued the bottling of milk at Dairy Service's plant after the merger. This testimony is entirely in accord with the Commission's findings in Beatrice Foods concerning the economic plight confronting small dairies.

Gold Medal Dairy, Keach Dairy of Henderson, and Purity Dairy were the other three dairies that ceased operating in the Evansville-Henderson market after Dean's entry. The first two merged with American Dairy, the last merged with Ideal Dairy, the largest company in the market. The only evidence in the record with respect to these three dairies is the profit and sales figures of Gold Medal. There is no evidence of the performance of these dairies, the effect Dean's practices had on them, or their reasons for discontinuing business. When the fact that several dairies entered the Evansville-Henderson market after Dean's entry is considered with the evidentiary deficiencies concerning the relation of Dean's activities to the cessation of operations of five dairies in that market, the Commission's conclusion that Dean's activities were responsible for "the demise" of these dairies and raised a reasonable possibility of causing "the demise of others" is without substantial evidentiary support.

Officials of American Dairy and Henderson Creamery provided the only other testimony in support of the Commission's case against Dean in the Evansville-Henderson market. The major portion of the testimony of the witness for American Dairy concerning competitive injury centered around the effect of Dean's quantity discount system on American's operations. He attributed American's loss of 2,000 home delivery customers after Dean's entry to the widening spread between the home-delivered price and the lower in-store price of milk which he said was due to quantity discounts initiated by Dean. In addition, he testified that American responded to Dean's discounts by offering the same discounts to its wholesale customers. Since Dean did not engage in the home delivery of milk, and further, since it gained only 2 per cent of the wholesale business in the market from 1952 to 1960, there is no reason to believe that American's former home delivery customers purchased Dean's milk in the chain stores. In any event, whatever loss of home delivery business American suffered was due to the lower chain store prices and not to differences in prices either within the Evansville-Henderson market or between that market and the Louisville market. Finally, his testimony concerning the effect of Dean's activities on his dairy's wholesale business indicates that that business was not adversely affected but in fact increased.

With respect to Henderson Creamery, the Commission misstates the record when it observes, "Although its overall wholesale sales increased somewhat between 1952 and 1960, the increase seems primarily attributable to increased sales in rural areas." The comparison testified to was between 1950 and 1960 and only about one-half of the $90,000 increase was due to a new rural route. In addition, Henderson's sales and profits figures are entitled to little weight since they reflected a combination of form and dairy operations without segregation.

The testimony of Henderson Creamery's president on the question of business losses is of little probative value. For example, he stated, "At the time when Dean came in, we lost a lot of wholesale business." (Emphasis added.) In view of the restraining order entered against Dean within a month after it entered the Henderson market, any losses suffered "when Dean came in" could not have been due to any pricing activities by Dean in that segment of the market A more tenable explanation for these losses was a consumer preference for other brands of milk. Corroboration for this explanation is found in his other testimony that Henderson Creamery was "given a bad eye * * * to the public," that his company "lost no wholesale accounts" to Dean, and that those accounts in which he lost some business were split with Dean and several other dairies.

Heavy emphasis is placed by the Commission on Utah Pie Co. v. Continental Baking Co., 386 U.S. 685, 87 S.Ct. 1326, 18 L.Ed.2d 406 (1967), decided by the Supreme Court after the Commission's decision in this case. Like the Commission's case concerning the Evansville-Henderson market, Utah Pie dealt with territorial price discrimination. The issue before the Court in that case was whether plaintiff in a treble damage action had adduced sufficient evidence of probable injury to competition within the meaning of section 2(a) of the Robinson-Patman Act. In finding that there was sufficient proof, the Court mentioned "persistent sales below cost," "radical price cuts themselves discriminatory," and a "drastically declining price structure." Utah Pie Co. v. Continental Baking Co., supra at 702-703, 87 S.Ct. 1326. The first two factors are significant in that the Court indicated that proof of their existence both could give rise to an inference of predatory intent and apparently would preclude characterizing a company's activities as motivated by "only fierce competitive instincts."

The Commission asserts, "For most of 1958 and a large portion of 1960, respondents [Dean] sold to their favored customers in this market [Evansville-Henderson] at a price which was below their delivered cost." This assertion conveys the mistaken impression that such below-cost sales were continuously and intentionally made. The record indicates that during these periods, Dean's sales were often not below its "delivered cost." More importantly, Dean did not engage in the kind of aggressive, persistent, below-cost sales of the magnitude that occurred in Utah Pie. In contrast, Dean's below-cost sales in the Evansville-Henderson market were sporadic and often the extent to which costs exceeded prices was negligible. There were no "radical" price cuts in the Evansville-Henderson market such as those in Utah Pie. Finally, in the Evansville-Henderson market, there was not a "drastically declining price structure." In contrast to the approximate 33 per cent decline in the level of net wholesale prices over a forty-four month period in Utah Pie, net wholesale prices in Evansville-Henderson were about the same in 1960 as they were in 1952.

We hold that the record in this case does not support the Commission's conclusion that either actual or reasonably possible injury to competition in violation of Section 2(a) resulted from Dean's pricing activities in the Evansville-Henderson market.

II.

The Falls Cities (Louisville) market includes the City of Louisville, suburbs in Jefferson County, Kentucky, and the Indiana cities of New Albany, Jeffersonville, and Clarksville (across the Ohio River). Dean Kentucky entered this market in September 1952 through the acquisition of Fenley's Model Dairy. Several weeks after Dean commenced operations in the Louisville market, it eliminated the one-cent per quart differential between creamline and homogenized milk, selling homogenized at the lower creamline price.

A quantity discount system was introduced by Dean in November 1954 and continued until December 1960. On several occasions during this period, Dean also offered earned service and earned handling discounts. These discounts in combination ranged from a minimum of 2 per cent to a maximum of 10 per cent, although there were a number of changes both in the percentage allowances and the qualifying volumes. During most of 1959 and 1960, Dean offered all wholesale purchasers a uniform 7 per cent discount irrespective of the volume purchased.

The Commission determined that Dean's discount system resulted in injury to competition at the primary level in the Louisville market in violation of Section 2(a) of the Robinson-Patman Act. Dean does not dispute that it discriminated in price "between different purchasers of commodities of like grade and quality." Dean does, however, dispute that the "effect of such discrimination may be substantially to lessen competition * * * in any line of commerce, or to injure, destroy, or prevent competition * * *." Thus, the issue in this segment of the appeal is whether substantial evidence in the record supports the Commission's conclusion that Dean's discount system resulted in injury to competition proscribed by the Robinson-Patman Act.

After the Commission reviewed relevant court precedent on primary-level discrimination, it devised a number of criteria to determine competitive injury. In summary, the Commission observed:

[A] finding of possible substantial competitive injury on the seller level is warranted in the absence of predation where the evidence shows significant diversion of business from the discriminator's competitors to the discriminator or diminishing profits to competitors resulting either from the diversion of business or from the necessity of meeting the discriminator's lower prices, provided that these immediate actual effects portend either a financial crippling of those competitors, a possibility of an anticompetitive concentration of business in larger sellers, or a significant reduction in the number of sellers in the market.

Applying the foregoing criteria, the Commission emphasized the fact that the percentage rate of Dean's growth was considerably greater than that of either the market as a whole or Dean's competitors individually. On the basis of this fact, the Commission concluded that the "program of quantity and other discounts was the prime factor in their [Dean's] unusual growth." What made this relation between Dean's discount system and its growth invidious, in the Commission's view, was the effect of that discount system on Dean's competitors. First, the Commission found that the discounts, initiated by Dean and met by its competitors, lowered wholesale prices. This decline, according to the Commission, apparently resulted in several dairies losing specific wholesale accounts and suffering reduced volume in others. Second, because of the "inherent `tying' effect" of quantity discounts, "in that they encourage purchaser [sic] to confine his purchases to a single seller or a small number of sellers so that he will obtain the maximum discount," the Commission concluded "that the discounts were instrumental in preventing competing dairies from increasing their sales to stores." It was the Commission's view that not only were competing dairies prevented from increasing their sales through loss of existing accounts and inability to acquire new ones, but that these dairies were also prevented "from breaking into the wholesale market." Third, "quantity discounts were a contributing factor in initiating and sustaining the trend away from the home delivery of milk and milk products in the Falls Cities [Louisville] market." Because the quantity discounts initiated by Dean and met by its competitors widened the preexisting one-cent differential between the in-store and home-delivered price of a quart of milk, the chain stores were able to attract significant amounts of the home delivery business of Dean's competitors.

The Commission observed that the effect of Dean's quantity discounts on the local dairies was "both permanent and profound," due to their loss of wholesale accounts, their disproportionately small share of the expanding market, and their decline in home delivery sales. According to the Commission, the "continued operating losses" of Dean's competitors, "forced to grant substantial discounts during a period of rising operating costs" and declining volume, foreclosed the prospect of meaningful competition. In the Commission's view, the "severely weakened financial condition" of Dean's competitors portends not only their inability to expand facilities and engage in "an aggressive marketing effort," but also an increased concentration in the market.

Most of Dean's competitors in the Louisville market testified in behalf of the Commission. In substance, their testimony concerning the impact of Dean's practices in the Louisville market focused on several factors. First, the uniform elimination of the preexisting differential between the price of creamline and homogenized milk might be the difference between making a profit and sustaining a loss. Second, although few of the dairies lost wholesale accounts to Dean after it introduced quantity discounts, the effect of those discounts was that Dean's competitors experienced a smaller wholesale volume. Most of this volume loss occurred in stores carrying more than one brand of milk. Once Dean obtained a place for its milk in such stores, the grocery did not purchase as much of the competitor's milk, although it usually continued to carry more than one brand. Third, because the system of discounts initiated by Dean led to an increase in the differential between the home-delivered and the chainstore price of milk, Dean's competitors suffered losses of varying degrees in their volume of home delivery business. Fourth, the other Louisville dairies adopted similar discount schedules in response to those of Dean to retain the wholesale business they had, even though to do so meant reduced prices and profits. Fifth, after Dean's entry into the Louisville market, the number of dairies decreased due to losses sustained as a result of Dean's marketing practices.

After considering the record as it relates to the Commission's conclusion that Dean's quantity discount system in the Louisville market violated Section 2(a) on the primary level, we find that its conclusion is not supported by substantial evidence. Initially, we note that the record discloses no evidence of predation or "buccaneering" on Dean's part; the Commission concedes as much. According to the Commission, a case of proscribed competitive injury can be established either from evidence of significant diversions of business from the discriminator's competitor to the discriminator or diminishing profits to competitors due to diversion or meeting the discriminator's lower prices. In addition, these "immediate actual effects [must] portend" a financial crippling of competitors, a possibility of concentration of business in larger sellers, or a significant reduction in the number of sellers. It is true that not only did some of Dean's competitors lose business but also that Dean gained business. Further, it is true that some of Dean's competitors suffered declines in profits and even sustained losses. However, the record does not show that these diversions and losses were the "immediate actual effect" of Dean's discrimination in prices (quantity discounts) in the Louisville market.

As part of the Commission's opinion, there is a chart depicting the class I milk utilization of Louisville dairies in pounds from 1953 through 1960. Of the fifteen dairies listed on that chart, excluding Dean, eleven registered gains in class I utilization while only four registered losses. What these figures suggest, and what the Commission seems to concede, is not that Dean diverted any significant volume of business from its competitors, but rather that Dean grew at a faster rate than its competitors.

The Commission focuses on Dean's discount system as "the prime factor in their [its] unusual growth." Underlying this position is the mistaken belief that all competitors in a market have a vested right to share in its growth in the same proportion as their existing shares of the market. The practicalities of competition, however, necessarily result in some competitors growing at the expense of their smaller or less efficient rivals. A more tenable explanation for Dean's growth can be found by analyzing a number of other factors, all of which have substantial evidentiary support, but some of which the Commission did not consider fully.

Although the record contained some evidence that Dean replaced competitors in some wholesale accounts, the bulk of the testimony showed that Dean's products were carried in addition to those of its competitors in the grocery's dairy case. Moreover, most of Dean's competitors met Dean's quantity and other discounts at the time that they were implemented. Thus, in a given wholesale account, the price charged to a grocery by each dairy whose milk it carried and the price the grocery in turn charged the consumer for the milk of each dairy were usually the same. An analogous situation existed when all the dairies followed Dean in uniformly reducing the price of homogenized milk one cent. Any diversion of business to Dean resulting from such a uniform price reduction could only be premised on factors other than price which influence a consumer. By the same token, when the milk of Dean and one or more other dairies was selling at the same price in a particular store, the fact that the consumer chose to purchase Dean's milk would in most instances be due to factors other than Dean's quantity discounts. We believe the Commission overlooked the role "consumer preference" played in the relative growth rates of the dairies in the Louisville market.

The Commission also relied on the losses suffered by Dean's competitors, allegedly due to quantity discounts, as a factor contributing to Dean's disproportionately large rate of growth. That some of the Louisville dairies suffered declining profits or losses after Dean's entry into the market is not disputed. But mere mention of these losses, without further explanation, is both misleading and incomplete. Of the thirteen Louisville dairies that introduced figures of yearly sales and profits, only five included figures which revealed their expenses and costs of goods sold. In the absence of any indication of the yearly costs and expenses incurred by the other eight dairies, the bare statement of their sales and profits (or losses) is virtually meaningless in demonstrating a causal connection between the dairies' employment of quantity discounts and their performance. Moreover, a close examination of the more detailed figures of the five dairies fails to disclose any discernible relationship between the discounts they paid and the profits or losses they experienced. As the Commission observed, the costs of many of the smaller Louisville dairies were rising. And in Beatrice Foods, the Commission recognized that technological changes, small size, and inefficiency have impaired the ability of many dairies to remain in business or engage in meaningful competition.

Our analysis of the reasons for the declining profits of Dean's competitors is buttressed by considering the profit and sales performance of Dean. After the first two years, when losses were sustained, Dean's operations showed large profits thereafter. Since Dean was paying substantial sums as rebates under its quantity discount system, its profits must in some measure have resulted from efficient operations. Moreover, there is no indication that the higher prices received by Dean from those accounts where discounts were small supported the lower prices received from those accounts where discounts were high. Even further, if Dean had paid the highest discounts to every wholesale account purchasing its milk in Louisville, its operations would still have shown sizable profits over the years in question. Although we are not suggesting that these considerations concerning Dean's performance are determinative, we think the Commission did not give them the weight they deserved.

The Commission's finding of a causal relation between Dean's quantity discounts and its competitors' losses is further weakened by the substantial amount of evidence suggesting the existence of a price-fixing conspiracy among Dean's competitors in Louisville. In 1948 an indictment was returned charging price fixing among the local Louisville dairies. There was other evidence of price discussions among Dean's competitors and of meritless lawsuits brought against Dean which were in part financed by dairies not parties thereto. These were admittedly aimed at getting Dean to raise its prices. The Commission misconstrued the import of this evidence by characterizing, as "a defense of an affirmative nature," that "proof of a price conspiracy among their [Dean's] competitors prevents any finding of injury to competition on the seller level." By taking this view, the Commission foreclosed itself from considering these facts as they related to the issue of the cause of the losses sustained by Dean's competitors.

An additional element in the Commission's conclusion that Dean's quantity discounts resulted in injury to competition centers on what it termed the "inherent `tying' effect" of such discounts. According to the Commission, these discounts and the effect they had were "instrumental in preventing competing dairies from increasing their sales to stores." In its brief, the Commission illuminates some of the evidentiary support for its proposition,

[G]rocers in the Fall Cities [Louisville] market, testified that their business was "keenly competitive," that gross and net profit margins are very small, and that consequently they take advantage of cash discounts and any other discounts offered.

A "long step" was needed for the Commission to conclude that Dean's quantity discounts had an "inherent `tying' effect" on the basis of testimony that grocers take advantage of discounts. For several reasons, we believe that the evidence did not warrant the Commission in taking that step. First, in those stores purchasing a large volume of milk, their purchases could be and were in fact divided among more than one dairy without the sacrifice of any quantity discounts. Second, since nearly all of Dean's competitors offered the same quantity discounts as Dean did, if a grocery in fact decided to confine its purchases to only a few or even one dairy in order to obtain the highest discount, there is no "inherent" reason why it should have confined its purchases to Dean rather than to Dean's competitors. Third, the evidence indicates that Dean actually replaced competitors in relatively few wholesale accounts. Fourth, some of the testimony the Commission relied on to demonstrate this "inherent `tying' effect" disclosed that some groceries purchased from several dairies, both large and small, during the time the discounts were in effect. Fifth, there is no warrant, either evidential or logical, for the Commission's "patent" hypothesis that quantity discounts of themselves would be more likely to prevent a new chain store from splitting its accounts among several dairies or from offering to "dairies not selling in its other stores the opportunity of selling in the new store." In sum, we believe the Commission was incorrect in concluding that Dean's quantity discounts had an "inherent `tying' effect."

Finally, there is a lack of substantial evidence to support the Commission's conclusion that Dean's quantity discount system was a "contributing factor" in the decline of the home delivery business of Dean's competitors. Although there is evidence of such a decline while Dean was offering quantity discounts, it does not inevitably follow that the decline was caused by the discounts. The Commission has elsewhere observed that "supermarkets came largely to replace home delivery as the major channel of fluid milk distribution." Beatrice Foods Co., 1965-1967 Transfer Binder ¶ 17,244, at 22,327 (1965). In the instant case, there is evidence that chain store milk prices were dropping and that home delivery customers were shifting their purchases to chain stores because of lower prices. But the fact that this phenomenon was not limited to the Louisville market suggests that factors such as the growth of shopping centers, lighterweight and more convenient containers, and increasing efficiencies in chain store distribution also played a part in this shift away from the home delivery of milk. Consequently, we believe that attributing the home delivery decline to Dean's quantity discount system was both uncritical and unwarranted.

Although the record in this case demonstrates that Dean's Louisville competitors underwent changes and that some suffered adversities and even demise, substantial evidentiary support is absent for the conclusion that Dean's pricing practices were the cause of these events. For the foregoing reasons, we are constrained to disagree with the Commission's determination that Dean's quantity discounts were responsible for either actual or potential injury to competition at the seller level in the Louisville market in violation of Section 2(a).

III.

The only issue raised by the Dean companies with respect to the findings of price discrimination at the secondary (buyer) level in Terre Haute and Louisville is whether "either or any of the purchases involved in such discrimination are in commerce," as required to give the Commission jurisdiction of violations of Section 2(a) of the Robinson-Patman Act. The evidence with respect to Terre Haute showed that the milk sold by Dean Illinois in that market was processed at its Rochester, Indiana plant. Although the movement of milk to Terre Haute from the Rochester plant was itself wholly intrastate, the Commission, agreeing with the examiner, found that "the record shows that a substantial portion of the milk processed at the Rochester plant was acquired from out-of-state producers." Relying on Foremost Dairies, Inc. v. FTC, 348 F.2d 674 (5th Cir.), cert. denied, 382 U.S. 959, 86 S.Ct. 435, 15 L.Ed.2d 362 (1965), the Commission concluded that "purchases by customers in Terre Haute are in commerce."

Substantial evidence in the record supports the Commission's conclusion. In Foremost, a multistate dairy processor supplied milk to customers in the Albuquerque, New Mexico market, 20 per cent of which was consistently produced in Colorado. Before the milk reached the ultimate consumer in Albuquerque, it was processed and bottled at Foremost's Santa Fe, New Mexico plant. The Fifth Circuit stated:

We think the sales in the instant case were "in commerce" as that phrase is used in section 2(a) * * *. [T]he milk passed in a steady flow from the farms in Colorado through the Sante Fe processing plant, where it underwent a rather negligible processing operation, which did not change its character appreciably, to the shelves of retail grocery establishments in Albuquerque. Foremost Dairies, Inc. v. FTC, supra, 348 F.2d at 677.

The record in the instant case discloses invoices showing that during at least four of the months between January 1958 and September 1960, raw milk supplied to the Rochester, Indiana plant was produced in and delivered from Wisconsin. Moreover, during the testimony of a Dean Illinois official concerning the source of the Rochester plant's milk, counsel for Dean said, "Well, we admit in the answer, if the Court please, that Dean purchases raw milk in the State of Wisconsin and Illinois." We believe that these invoices and counsel's "admission" entitled the Commission to conclude that a requisite amount of the milk purchased from Dean by wholesalers in Terre Haute during the period covered by this action was "in commerce."

With respect to the "in commerce" aspects of Dean Kentucky's and Dean Illinois' secondary-level price discriminations in Louisville, the record discloses that virtually all of its raw milk was purchased from the same local Louisville producer's cooperative. The cooperative in turn obtained the milk from its producer members whose farms were located within a radius of about seventy miles of Louisville. This area included parts of Indiana. A Dean Kentucky official, when asked what proportion of the milk handled by the cooperative was produced in Indiana, responded with a figure of 25 per cent. Irrespective of whether the cooperative served as a conduit for the sale of milk of its Indiana farmer members or as the agent of Dean in acquiring milk, we believe that 25 per cent of the milk processed in Dean's Louisville plant was passing in a continuous flow from Indiana farms. Consequently, we agree with the Commission that the "in commerce" requirement was satisfied with respect to the secondary-level case in Louisville. Foremost Dairies, Inc. v. FTC, supra.

IV.

Dean finally complains about the scope of the Commission's order, urging that it is unduly broad. According to Dean, although there was no evidence that it engaged in proscribed pricing practices in any other markets than Louisville, Evansville-Henderson, and Terre Haute, or with respect to any other processing plants than those located in Louisville and Rochester, "the cease and desist order covers both area and local price discrimination in every community in which Dean Illinois and its subsidiaries do business." In response, the Commission argues that a broad order is appropriate for two reasons: (1) since the discriminatory practices can readily be repeated in other markets served by Dean, the broad order will relieve the Commission of dealing piecemeal with such possible unlawful practices; (2) since Dean could merely shift sales from the Louisville and Rochester plants to other plants, limiting the order to those two plants alone would enable Dean to evade the order with ease.

We agree in part with the arguments of both Dean and the Commission. There was evidence in the record that Dean Illinois is now supplying the Terre Haute market fom its processing plant located in Chemung, Illinois. This fact alone suggests the inadequacy of limiting the order solely to the Rochester and Louisville plants as Dean suggests. On the other hand, there is no evidence in the record that Dean engaged in unlawful pricing practices at the secondary level in any markets other than Terre Haute and Louisville. Consequently, the order should be modified, with respect to secondary-level discrimination, to prohibit Dean Kentucky and Dean Illinois from selling the same milk products to competing purchasers at different prices in either the Terre Haute or the Louisville market, irrespective of the processing plant serving those markets.

Enforcement of the Commission's order with respect to the primary-level violations is denied. The proceeding is remanded to the Commission solely for the purpose of modifying its order with respect to the secondary-level violations in conformity with the views set forth in this opinion. As modified, the Commission's order with respect to the secondary-level violations will be enforced.