This private antitrust action is before the Court for the second time. When the matter was first here, we held that the appellees had engaged in a broad conspiracy in violation of the antitrust laws and that their unlawful acts had injured National Farmers' Organization, Inc. (NFO). We remanded the matter to the district court with directions to determine the amount of NFO's damage.

On remand, NFO advanced two alternative damage measurement theories. Under a "market structure theory", NFO claimed damages of $18,490,694, trebled to $55,472,082. Under a "test market theory", it claimed damages of $7,403,374, trebled to $22,210,122. The district court refused to award any damages, and NFO again appeals.

We reaffirm our holding that NFO incurred substantial damages as a result of appellees' unlawful acts. We provide more specific directions to the district court with respect to the computation of damages and remand to it for that computation. We also direct the district court to determine and award appropriate attorney's fees and costs. We affirm the district court's denial of certain injunctive relief and its denial of class certification.

This action returns to the Court more than five years after we remanded it to the district court and more than seventeen years after it was commenced. The district court conducted the original trial in three phases. Phase I involved claims brought by Mid-America Dairymen, Inc. (Mid-Am) against NFO for violation of state and federal antitrust laws. Phase II involved NFO's counterclaims against Mid-Am, Associated Milk Producers, Inc. (AMPI), Central Milk Producers Cooperative (CMPC) and Associated Reserve Standby Pool Cooperative (ARSPC) for violations of the federal antitrust laws. Phase III involved AMPI's claims against NFO and nineteen individual NFO members for violation of the federal antitrust laws and certain provisions of the Agricultural Fair Practices Act of 1967, 7 U.S.C. Sec. 2303(c), (e) and (f). In an exhaustive 147-page opinion, the district court found in favor of NFO on the phase I and III claims and against NFO on the phase II claims. In Re Midwest Milk Monopolization Litigation, 510 F. Supp. 381 (W.D.Mo.1981) (NFO I).

On appeal, this Court affirmed the district court with respect to the phase I and III claims. Alexander v. National Farmer's Organization, 687 F.2d 1173, 1183-90 (8th Cir.1982) (NFO II). With respect to the phase II claims, we affirmed the district court's findings that NFO failed to present sufficient statistical evidence to support its claims of actual and attempted monopolization, id. at 1191-92, and that NFO had not produced sufficient evidence to support a finding of liability on the part of ARSPC arising out of its operation of standby milk pools, id. at 1207. We reversed the district court's rejection of the remainder of NFO's claims, however, finding that "AMPI, Mid-Am and CMPC did conspire to monopolize and eliminate competition in the marketing of Grade A raw milk produced in the Midwest, through the use of discriminatory pricing, coercive supply disruptions and threats of similar conduct, as well as bad faith harassment and threats of litigation against independent buyers of NFO milk." Id. at 1193.

We found that the conspiracy proved by NFO covered a ten-state midwestern region and had as one of its goals elimination of NFO as a competitor in the marketing of raw Grade A milk.1 Id. at 1193. With respect to the conspiracy element, we found that AMPI and Mid-Am each belonged to the other's organization and effectively divided marketing territories in the Midwest, South and Southwest; that AMPI was a member of CMPC and served as CMPC's marketing agent in Chicago, which area served as a base pricing point for other areas served by Mid-Am; that Mid-Am and AMPI officials met from time to time to discuss their concerns regarding NFO; and that, at one point, senior officials of CMPC, Mid-Am and AMPI held a joint meeting to discuss problems created by NFO's lower prices and successful solicitation of their members. Id. at 1194-95.

In examining specific overt acts of AMPI, Mid-Am and CMPC in furtherance of the conspiracy, we set forth examples of conduct which, viewed in the context of the entire record, "demonstrate[d] concerted, unlawful tactics." Id. at 1207. The examples included the following.

A. The Gandy Dairy Matter (Supply Cut-offs and Limitations)

Prior to May of 1971, Gandy Dairy, of Central West Texas, purchased all of its milk from AMPI. In March and April of 1971, however, Gandy agreed to buy some milk from independent suppliers. On the day the shipments from the independents began, AMPI began "short shipping" and making late deliveries of milk to Gandy. At the same time, AMPI began soliciting Gandy's customers and offering them direct shipments of competing products at prices near Gandy's cost of production. The avowed purpose of this conduct was to force Gandy to return to AMPI for its full requirements of milk. This Court found AMPI's conduct "blatantly predatory." Id. at 1196.

B. The Wanzer Dairy Matter (Discriminatory Pricing)

Since the Chicago area served as a base pricing point for other areas to the south and west, appellees sought to maintain an elevated price in it. In 1970, however, CMPC's price was being undermined by Wanzer Dairy's purchases of non-CMPC milk at lower prices. In response to the purchases, AMDI, a major trade association of proprietary milk buyers in Chicago, met with CMPC and demanded that it drop its price so long as Wanzer was able to obtain milk at a lower price. CMPC complied with the request. At the same time, however, CMPC announced a new price structure, imposing additional charges on any buyer taking less than all of its requirements from CMPC. Later in 1970, after several independent marketers who had previously supplied Wanzer joined CMPC, CMPC again raised prices to buyers taking less than all of their requirements from it. The price increase was especially harmful to Wanzer because it could not fulfill its milk requirements without purchasing some milk from CMPC.

Notwithstanding the price pressure, Wanzer continued to purchase milk from independents, aided to some degree by NFO's offer to reimburse Wanzer for some of the additional charges. In response, AMDI dealers met with CMPC and renewed their objections to paying the CMPC price while Wanzer was obtaining cheaper supplies of milk. After this meeting, CMPC sent a letter to Wanzer threatening to discontinue all milk shipments in five days because "we have become aware that you have been engaged with others in unlawful attempts to interfere with producers whose milk is subject to effective marketing agreements with CMPC and its members." A short time after the letter, Wanzer entered into a one-year committed supply contract with CMPC. CMPC then met with the AMDI dealers, informed them of the Wanzer contract, and reinstated its elevated price. See id. at 1198.

[T]he conduct so established is predatory on its face. Our examination of the underlying evidence only confirms that an inference of unlawful purpose must be drawn. It was clear error for the district court not to find that the pricing scheme was unlawfully discriminatory and that the threatened supply cutoff was unlawfully coercive.

Id. at 1199.

C. Beatrice Dairy (Sham Litigation or Threats of Sham Litigation)

In July of 1970, Beatrice Dairy of Fort Worth, Texas, began buying milk from NFO. By February of 1971, it was buying several million pounds of NFO milk per month. By early 1971, Mid-Am had advised Beatrice that it thought NFO's marketing was illegal and that it had the exclusive right to market some of the milk NFO was selling to Beatrice. Mid-Am later met with Beatrice and threatened legal action unless it was paid for the disputed milk. In March of 1971, Mid-Am sued Beatrice and NFO. Thereafter, Beatrice notified NFO that, due to the dispute and NFO's refusal to indemnify it for losses as a result of Mid-Am's action, it was immediately terminating purchases from NFO.

In summarizing the evidence with respect to the Beatrice sham litigation claim and rejecting the appellees' theory that NFO's injuries were self-inflicted, we stated:

Both Mid-Am and AMPI contend that Beatrice's cutoff of NFO was a unilateral, uncoerced business judgment based upon price and quality considerations.... The termination letter itself and contemporaneous deposition testimony of the Beatrice officials involved in the decision leave no doubt that the acute factor behind the cutoff was Mid-Am's legal claim.... Whatever price or other problems may have accompanied NFO milk, the undisputed facts are that Beatrice purchased an ever-increasing volume of such milk continuing right up to the sudden cutoff, which came twelve days after Mid-Am filed suit against Beatrice. In view of the timing and other contemporaneous evidence and the fact that the cutoff was made in consultation with Beatrice's legal department, there is simply no doubt that the Mid-Am litigation was a material cause of the cutoff.

Id. at 1202 n. 34.

Our prior opinion listed examples of other incidents and actions which could, in the totality of the circumstances, only be viewed as unlawfully aimed at eliminating NFO as a competitor. Among these actions were: Mid-Am's refusal to deal with NFO members who wished to terminate their relationship with Mid-Am regardless of the members' right to terminate, and Mid-Am's firing of a contract hauler for engaging in activities helpful to NFO. Id. at 1204-05.

Viewing the record as a whole, we found that the appellees engaged in a broad conspiracy to monopolize the marketing of Grade A milk in a ten-state region by engaging in:

[a] series of unlawful acts ... aimed at eliminating independent sales of milk in general and NFO in particular, including discriminatory pricing and actual or threatened supply cutoffs, litigation and similar harassment. Id. at 1207-08.

In addition, we found that the appellees' unlawful conspiracy was a material cause of substantial injury to NFO. Id. at 1210. Accordingly, we remanded to the district court with directions to it to determine the amount of damages to which NFO was entitled and set forth guidelines to be used in that determination. Id. at 1208. The guidelines included the following:

(1) Because NFO is a nonprofit entity which distributes sales proceeds to members after deducting marketing expenses and its net revenues were tied to sales volume and not to price, NFO could not recover the difference between the price it would have obtained but for the conspiracy and the price it actually obtained. Such a price differential, we held, belonged to its members--not to NFO.

(2) Because the appellees' unlawful conduct caused NFO to lose both members and marketing volume, NFO could recover for losses of membership dues and checkoff fees.

(3) Because NFO had clearly established the fact of injury,2 the district court had broad latitude in assessing the amount of damages. Id. at 1210.

On remand, the district court afforded NFO the opportunity to present additional evidence in support of its damage claim. NFO expressly declined to do so, choosing instead to rely upon the damage theories and supporting evidence it presented at the original trial. Alexander v. National Farmers' Organization, 614 F. Supp. 745, 758, 775 (W.D.Mo.1985) (NFO III). After extensive briefing and argument, the district court rejected NFO's damage claims in their entirety. Id. at 792-93, 795, 799. In our view, it erred in so doing. Thus, although this matter has already consumed months of the district court's time, we have no alternative but to remand for an award of damages consistent with the detailed instructions set forth herein.

NFO introduced two alternative damage theories at trial, primarily through the testimony of its damage expert, Robert R. Nathan. The theories were both used to estimate NFO's lost dues and checkoff fees. The theories were described as the "test market" and "market structure" approaches. The district court rejected both theories, finding that "Nathan's testimony and NFO's damage exhibits adduced in connection therewith cannot be accepted as credible under all the circumstances of this case." NFO III, at 792. Accordingly, the court concluded that "a just and reasonable estimate of NFO's alleged lost market penetration damages cannot be made on any credible relevant data adduced by NFO." Id. at 793. We affirm the district court's rejection of the market structure theory.3 With respect to the test market theory, however, we hold that it is a proper basis for a substantial damage award.

Preliminarily, we note that the district court appears to have used an inappropriate legal standard in rejecting Nathan's testimony and NFO's damage exhibits. In order to recover damages, NFO had to show: (1) that the appellees violated the antitrust laws; (2) that the violation caused it injury; (3) that the injury was of a type sought to be prevented by the antitrust laws;4 and (4) the amount of its injury. See Zenith Radio Corp. v. Hazeltine Research, Inc., 395 U.S. 100, 123-25, 89 S. Ct. 1562, 1576-77, 23 L. Ed. 2d 129 (1969); McDonald v. Johnson & Johnson, 722 F.2d 1370, 1378 (8th Cir.1983), cert. denied, 469 U.S. 870, 105 S. Ct. 219, 83 L. Ed. 2d 149 (1984); Von Kalinowski, 10 Antitrust Laws and Trade Regulation Secs. 115.03, 115.04.

While the first three of these elements must be established by a preponderance of the evidence, a lesser standard of proof is applicable to proof of the amount of damage. See, e.g., J. Truett Payne Co. v. Chrysler Motors Corp., 451 U.S. 557, 101 S. Ct. 1923, 68 L. Ed. 2d 442 (1981) (distinguishing higher standard of proof for establishing fact of injury from lower standard of proof for establishing amount of damage); Story Parchment Co. v. Paterson Parchment Paper Co., 282 U.S. 555, 562, 51 S. Ct. 248, 250, 75 L. Ed. 544 (1931) (same).

This lower standard derives from the principle that a wrongdoer should not profit from the harm occasioned by its act. See, e.g., Bigelow v. RKO Radio Pictures, Inc., 327 U.S. 251, 264-66, 66 S. Ct. 574, 579-80, 90 L. Ed. 652 (1946) ("The most elementary conceptions of justice and public policy require that the wrongdoer shall bear the risk of the uncertainty which his own wrong has created."), quoted in J. Truett Payne Co., 451 U.S. at 566, 101 S.Ct. at 1929; see also Story Parchment Co., 282 U.S. at 563, 51 S.Ct. at 250 ("[I]t would be a perversion of fundamental principles of justice to deny all relief to the injured person, and thereby relieve the wrongdoer from making any amend for his acts.") Therefore, in proving the amount of damage, the antitrust plaintiff need only present evidence from which the fact finder may make a just and reasonable estimate of the damage that is not based on speculation or guesswork. J. Truett Payne Co., 451 U.S. at 566-68, 101 S.Ct. at 1929-30; Zenith Radio Corp., 395 U.S. at 124, 89 S.Ct. at 1577; Bigelow, 327 U.S. at 264, 66 S.Ct. at 579; Story Parchment Co., 282 U.S. at 563, 51 S.Ct. at 250.5

Applying the standard, the Supreme Court has recognized that "the vagaries of the marketplace usually deny us sure knowledge of what plaintiff's situation would have been in the absence of the defendant's antitrust violation," J. Truett Payne Co., 451 U.S. at 566, 101 S.Ct. at 1929, and has admonished the courts to "observe the practical limits of the burden of proof which may be demanded of a treble-damage plaintiff who seeks recovery for injuries from a partial or total exclusion from a market." Zenith Radio Corp., 395 U.S. at 123, 89 S.Ct. at 1576. The Court has acknowledged that "[a]ny other rule ... would be an inducement to make wrongdoing so effective and complete in every case as to preclude any recovery, by rendering the measure of damages uncertain." Zenith Radio Corp., 395 U.S. at 124, 89 S.Ct. at 1577.

Appellees contend that this Court's prior opinion established only that they engaged in unlawful conduct, leaving issues of causation to be determined on remand. We disagree. The opinion is clear that NFO established the fact of injury, that is, antitrust violations by the appellees which were a material cause of some injury to its business or property. See NFO II, 687 F.2d at 1210 ("Because the fact of injury is unmistakable on this record, the district court has rather broad latitude in assessing the amount of damages which NFO shall recover.")

Specifically, we cited examples of the appellees' unlawful conduct which caused injury to NFO. Among these were: CMPC's conduct with respect to Wanzer dairy,6 Mid-Am's unlawful threats of litigation against Beatrice dairy,7 and Mid-Am's and AMPI's conduct with respect to Foremost Dairy.8

Since our prior opinion clearly established the fact of injury to NFO, the district court on remand should have separately and independently examined the damage evidence put forward at trial by NFO in light of the lesser standard of proof applicable to determining the amount of damages. Instead, the district court rejected the evidence in its entirety because certain portions of it rested upon proposed findings the court rejected in the liability phase of the case. Such a practice, in effect, ignores the lesser standard of proof applicable to the amount of damages.

Moreover, in light of our prior opinion and the damage evidence adduced by the parties, even if the district court's review had not been so tainted, its finding that NFO offered insufficient proof to recover any damages would be clearly erroneous. As we point out below, NFO's damage evidence, while not perfect, clearly establishes the basis for a significant award. It is to this basis that we now turn.

Under the test market theory, NFO sought to use its performance in an identifiable portion of the relevant market as a "yardstick" to measure its performance in the remainder of the market.9 NFO chose as its yardstick measure its performance in federal milk market order number 68, a market order encompassing Minneapolis and St. Paul, Minnesota, and the surrounding area.

After selecting the order, NFO obtained from the United States Department of Agriculture (USDA) the total volume of milk pooled on it in each month for the years 1971 to 1975 and the total volume of milk pooled by NFO on it during the same period.10 From these figures NFO derived its monthly percentage market share on the order. This computation yielded a 10.5 percent share for 1973. NFO used this 10.5 percent share as a basis for estimating its lost checkoff fees and membership dues on the damage orders.

Appellees contend that the test market theory does not provide a sufficient basis to support NFO's damage claim. The validity of the theory turns primarily on whether the relevant conditions in the other damage orders sufficiently resemble the conditions in order number 68--the proposed test order (or "yardstick measure")--that extrapolation of NFO's performance to the other damage orders will provide a just and reasonable estimate of NFO's damage.

This Court has previously accepted the use of a yardstick measurement in determining the amount of antitrust damages. In Agrashell, Inc. v. Hammons Products Co., 479 F.2d 269 (8th Cir.1973), this Court approved a yardstick measurement of the plaintiff's (Hammons') damages for its alleged "failure to grow." The plaintiff's damage expert testified that he chose as the yardstick measure a portion of the relevant market where the defendant's antitrust violations were not prevalent.

This Court acknowledged that the expert's yardstick measurement had many defects, including that the test area had only one customer in it, that other customers almost always purchased smaller volumes, and that the test area customer may not have viewed the plaintiff's and defendant's products as interchangeable. In addition, the Court noted that the plaintiff's damage expert acknowledged he had not considered whether companies other than the defendant sold similar products or the impact of such sales. Finally, the Court noted the expert offered no support for the position that the test area was relatively unaffected by the defendant's unlawful conduct.11 Despite these defects, the Court found, as a matter of law, that the expert's testimony sufficiently proved the amount of damage, citing the principles set down by the Supreme Court in such cases as Zenith Radio Corp. and Bigelow. Id. at 283.

Similarly, NFO's yardstick measurement is not perfect. The record contains evidence that the order was unique in several respects, including its location in a surplus production area and the fact that a greater number of facilities existed in the area to handle that surplus. In addition, NFO enjoyed a relatively strong membership base on the order.12 Finally, as the district court highlighted in its decision, there is evidence in the record that dairy production in Minnesota grew at a faster rate than in the rest of the nation. NFO responds that milk pooled on order 68 comes from places other than Minnesota and that the growth of Minnesota production does not destroy the comparability of the order.13

Probably the most troublesome criticism of the use of order 68 as a yardstick, however, is the implication that the liberal pooling provisions applicable on the order allowed NFO to pool large quantities of milk on it in order to artificially inflate its performance.14 Yet, whatever the merit of this contention, the record does not disclose that the appellees ever presented any evidence from which the district court could have reduced the yardstick measurement. The only suggestion was that the court should entirely ignore NFO's performance in order 68. Such a course cannot be sustained.

For all of the shortcomings of the order as a yardstick measurement, the undeniable fact is that NFO actually managed to pool a 10.5 percent share on it over the entire year in 1973. In addition, review of this Court's prior opinion indicates that the order was not free from the appellees' unlawful acts and that NFO might have pooled an even larger share had it not been for these acts. Specifically, this Court cited Mid-Am's threat of litigation against Cloverleaf Creamery, a dairy located in order 68, as corroborating other evidence establishing a pattern of conduct involving threats of sham litigation. See NFO II, 687 F.2d at 1202 n. 33. Moreover, even if order 68 had been free of unlawful conduct by the appellees, it cannot be presumed that the appellees' unlawful conduct elsewhere would have had no "spillover" effect onto the order. Thus, to the extent that the appellees' unlawful acts restricted NFO's efforts in order 68, its use would tend to underestimate NFO's damages.

There are also factors indicating that order 68 was comparable to the other damage orders. Most significantly, as in Zenith Radio Corp., the relevant product and seller were the same in the test market area as in the damage area. In addition, with the exception of pooling provisions, it appears that the federal regulations applicable to the federal orders are substantially similar.

Finally, we note that the appellees offered no reasonable substitute yardstick measurement. Instead, the record reflects that the appellees' strategy at trial was to present only evidence which tended to show NFO had not been injured at all. To this end, they presented testimony of Dr. Babb and George Hansen, comparing, for the years 1970-77, NFO's performance on the damage orders with its performance on 19 non-damage orders. According to the testimony, because NFO performed as well or better on the damage orders as on the non-damage orders, it was not injured at all by the appellees' unlawful acts.

The study, however, is fatally defective in a number of respects. Most importantly, it is clear that NFO's dairy marketing efforts were focused in the Midwest, the area where its membership base was the strongest.15 Thus, one would expect NFO to make its strongest showing in that area. The fact that it did so despite the appellees' unlawful practices does not mean that NFO suffered no damage. Indeed, NFO's relative success in the face of the numerous unlawful acts of the appellees is an indication of NFO's capability in a violation-free market.16 In addition, the appellees made no showing of comparability of the non-damage markets. Finally, the conclusion to be drawn from the appellees' damage theory--that NFO suffered no damages at all from the unlawful conduct--is fundamentally at odds with this Court's prior opinion which established that NFO had suffered at least some damage as a result of the unlawful conspiracy.

In sum, we recognize that NFO's test market theory is not perfect. Nonetheless, in light of the Supreme Court's repeated admonition that it does not " 'come with very good grace' for the wrongdoer to insist upon specific proof of the injury which it has itself inflicted," J. Truett Payne Co., 451 U.S. at 566-67, 101 S. Ct. 1929 at 1929 (quoting Hetzel v. Baltimore & Ohio R. Co., 169 U.S. 26, 18 S. Ct. 255, 42 L. Ed. 648 (1898)), we hold that the Federal Market Order 68 is sufficiently comparable to the other damage orders that extrapolation of NFO's performance on that order to the other damage orders will provide a just and reasonable estimate of NFO's damage.

Using the 10.5 percent share it achieved on order 68 as a yardstick against which to measure its performance on the remaining damage markets, NFO presented evidence relating to elements of its damage claim expressly approved of in our prior opinion---lost checkoff proceeds and lost membership dues.

NFO paid its member producers the proceeds from sales of their production, less a marketing fee (ordinarily measured in cents per hundredweight of milk handled). This marketing fee is referred to as a "checkoff."17 See NFO II, 687 F.2d at 1208. NFO's lost checkoff fee calculation involved estimating the additional volume of milk it would have marketed absent the restrictive practices, multiplying that volume by the checkoff fee it would have obtained, and deducting the extra expense it would have incurred in handling the extra volume.

In order to estimate the additional volume of milk it would have marketed absent the restrictive practices, NFO first obtained from the USDA a summary of the total milk pooled on each market order. Next, it multiplied the total volume of milk pooled on each order by the 10.5 percent yardstick measure to obtain the volume of milk it would have pooled absent the restrictive practices. Finally, NFO subtracted from that volume the volume of milk it actually pooled on each order to arrive at the total additional volume it would have marketed.18

Once NFO had computed this additional volume of milk, it could determine the amount of its damages by multiplying the volume by the three cents per hundredweight checkoff it received on all milk marketed during the damage period.19

The appellees object to NFO's checkoff computation on the ground that the three cent per hundredweight figure represents NFO's gross losses and does not include expenses associated with marketing the additional volume of milk. NFO trial exhibit numbers 1003 and 1004, however, summarize the additional expenses associated with NFO's achievement of the 10.5% percent share on the damage orders. While the expenses associated with the additional volumes of milk and with collecting the checkoff were not deducted from the basic checkoff figures, we note that they were deducted from a separate claim for bargaining expense checkoff fees.20 We can see no reason to require NFO to twice deduct such expenses from its damage claim.

In addition, the appellees argue that the checkoff computation overstates NFO's losses because it fails to take into account NFO member dairy producers who did not market their milk through NFO but who, nonetheless, voluntarily contributed the three cent checkoff to NFO. They contend that NFO member milk producers are not required to market through NFO and that a number of NFO producer members chose to market their milk by some other means (including through one of the appellees). Therefore, to the extent that some NFO members who did not market through NFO voluntarily paid NFO the three cent per hundredweight checkoff, the damage claim is inflated.

The record indicates that NFO received the three cent checkoff from a number of its members who did not market through NFO and that this number was more than de minimis.21 It does not, however, indicate whether the voluntary contributors pooled milk on any of the damage orders or the volumes they pooled. The question whether the voluntary contributors pooled milk on the damage orders is particularly important because, to the extent they pooled milk elsewhere, their voluntary contributions should not reduce the damage award. In other words, the fact that an NFO member pooled milk through another marketing organization yet voluntarily contributed a checkoff fee to NFO, is only relevant if the milk on which the contribution was made was pooled on one of the damage orders.

Accordingly, NFO's failure to include voluntary contributions in its damage calculation is not a basis for totally rejecting it. Instead, NFO should be given an opportunity on remand to point out to the district court, from the existing record, evidence which will permit the court to make a reasonable approximation of the amount voluntarily contributed to NFO via the three cent checkoff from its members pooling milk on the damage orders who did not market milk through it.22

In addition to lost checkoff fees, our prior opinion recognized that NFO is entitled to recover membership dues for those members it would have added absent the restrictive practices. NFO II, 687 F.2d at 1209. Computing these dues, NFO first obtained from the USDA the yearly total of the number of producers pooling milk on each damage order. It then multiplied the total by the 10.5 percent yardstick measurement. This yielded an estimate of the number of producers NFO would have pooled in each year. See NFO Trial Exhibit No. 1024. NFO then subtracted from this number the greater of the actual number of producers it pooled on each order or the estimated number of NFO producer-members on each order23 to arrive at a conservative estimate of the number of members it would have added absent the restrictive practices.24 See NFO Trial Exhibit Nos. 1025-1034. Lost membership dues were then calculated by multiplying the number of members that would have been added by the net dues collected per member in each year.25

Appellees challenge NFO's calculation of lost membership dues on a number of grounds. First, as with the checkoff fee, they contend the calculation fails to adequately take into account the costs associated with the additional members. We disagree. Review of the record indicates that NFO deducted from its gross dues figures amounts representing the costs of additional personnel required to process the dues payments, the costs of sending the member an NFO newsletter, and the costs of providing each member with a membership card. See NFO Trial Exhibit Number 1006.

The appellees also contend that NFO's lost dues computation is fatally flawed because it assumes that the members NFO lost as a result of the unlawful conspiracy would have paid their dues. The appellees cite the fact that in 1975 NFO was owed nearly $48,000,000 in membership dues as supporting their contention. Thus, they argue it would be erroneous to assume that any members NFO would have added but for the unlawful conspiracy would have met their dues obligations.

Although there is evidence that NFO wrote off a substantial amount of dues receivable as uncollectible, the dues cannot be collected dues unless and until they are owed. Moreover, whatever merit there may be to the proposition that NFO has difficulty collecting dues from its general membership, the appellees made no showing that dairy producers who marketed their milk through NFO similarly failed to pay their membership dues.26 Finally, even if the appellees had made such a showing, it would not properly lead to a complete denial of NFO's dues claim. Instead, the appropriate course would be to reduce the dues claim to reflect the probability that they would not be paid.

The appellees also challenge NFO's calculation of its membership base. First, they challenge the accuracy of the computer tapes from which NFO derived its membership estimate. Second, they contend that the assumption used by NFO to calculate the number of dairy producers on each order who were members of NFO was unreasonable. The assumption was that the proportion of NFO member dairy farmers to the total number of dairy farmers in a state was similar to the proportion of NFO member farmers in a state to the total number of farmers in that state. Third, they challenge NFO's use of a single time period in estimating its membership. They contend that, by limiting the membership calculation to those who were members in 1969, NFO overstated its damages to the extent that producers who joined NFO after that time and did not market their milk through NFO were included in the damage calculation. Fourth, they argue that the lost dues figure is inflated because it assumes that a producer will market milk on only one order.27

While each of the appellees' objections in this regard might have merit in theory insofar as they point out potential sources of error in NFO's damage calculation, the record is entirely lacking evidence showing actual error or evidence from which the award could be reduced to correct any errors. Rather, the appellees contend that the potential for error renders the damage evidence unreasonable and speculative. This Court, however, must review the record in light of the Supreme Court's admonition to observe the practical limits of proof in calculation of the amount of damages. Zenith Radio Corp., 395 U.S. at 123, 51 S.Ct. at 1576; supra at 1293. Having done so, it is clear that, in the context of this case, the process by which NFO calculated its lost dues was just and reasonable. Accordingly, we hold that NFO's lost membership dues calculation is a proper element of its damage claim.

In addition to evidence concerning checkoff fees and membership dues, NFO presented evidence relating to a number of assumptions necessary to its damage calculation. Among these were assumptions concerning the time NFO would enter the various damage orders, the time it would take NFO to achieve its mature market share on the damage orders, the conditions on certain orders after their mergers, and the conditions on all of the damage orders in the intervening period from completion of the damage calculation to final judgment.

NFO's damage calculation reflects its intention to move as rapidly as possible in entering each of the damage orders, yet recognizes that simultaneous and successful entry on all ten damage orders was not likely. Thus, the calculation assumes that NFO either did or would have marketed milk on federal order numbers 30, 62, 64, 106 and 126 beginning in 1970, on federal order numbers 60, 61, 68 and 73 beginning in 1971, and on order number 65 beginning in 1972. NFO's claim includes no damages for any years prior to actual or intended entry on each federal order.

The appellees contend that the dates of NFO's entry on each of the damage orders are erroneous and unsupported by the record. We disagree. Many of the same portions of the record cited with respect to the focus of NFO's efforts also support its claim as to entry on the damage orders. See Trial Transcript at 2548 (testimony of Al Scott); 1697-1700 (testimony of Oren Staley); 6443 (testimony of Donald Berkhahn). In addition, the unlawful acts of the appellees cannot be viewed in isolation from the context in which they occurred. See NFO II, 687 F.2d at 1193-94, 1208 (citing Continental Ore Co. v. Union Carbide & Carbon Corp., 370 U.S. 690, 699, 82 S. Ct. 1404, 1410, 8 L. Ed. 2d 777 (1962)). It is indisputable that the unlawful conspiracy had a damaging effect on NFO's ability to expand as it had intended. Thus, we hold that the assumptions as to the dates NFO would have entered the damage markets are a reasonable element in NFO's damage calculation.28

NFO's damage estimate also reflects the fact that it would not likely achieve its mature market share on each damage order in the same year it began marketing milk on that order. Instead, based on its experience in the test market order (federal order number 68),29 NFO assumed that it would achieve its mature (10.5%) market share in three years with one-third of the mature share being achieved in each year.30

The appellees object to this assumption, arguing that NFO's performance on other market orders does not support it. In essence, the question of time to achieve mature market share is part and parcel of the larger question whether NFO's experience on order 68 may be used as a yardstick against which NFO's performance on the other damage orders may be measured. Therefore, consistent with our holding above that order 68 may reasonably be used as a yardstick, we hold that NFO's assumption, derived from its performance on that order is reasonable and was properly used in its damage computation.

3. Combination of Federal Orders and Post Computation Figures

The damage calculation also includes assumptions as to damage orders that were merged with other orders. In 1975, several Texas federal market orders were merged with North Texas order number 126 to create a new federal order number 126. In 1976, federal order numbers 60, 61 and 68 were merged into a new federal order number 68. Since the mergers made calculation of damages on the merged orders nearly impossible, NFO assumed that conditions on each of the merged orders would remain the same as the last year prior to the merger.31 See Trial Transcript at 9697-9700.

Similarly, the damage calculation also assumes that conditions on the market orders would remain essentially unchanged after 1977. This assumption provided a method by which NFO could predict the damages it would incur during the interval between completion of its damage calculation and final judgment in the case.32 Accordingly, because the most recent figures available at the time of the damage calculation were for August of 1977, NFO used 1977 figures in its damage calculations for the three years following 1977, essentially assuming static conditions for the three-year period.33 See Trial Transcript at 9658.

The district court held that the assumptions as to static conditions were unreasonable and not supported by the record. More specifically, the court rejected NFO's contention that the universe of producers from which it would derive its market share was static and found instead that the number of producers in the dairy industry was decreasing. NFO III, 614 F. Supp. at 790-91.

The flaw in the district court's approach is that, although the number of producers has decreased over time, the total volume of milk production has increased. In addition, the record reflects that the share of that production marketed by cooperatives has increased. See Trial Transcript at 14,421. Thus, to the extent that NFO's damages rest on the number of producers in the market, its assumption of a static situation may tend to overestimate its actual performance. However, to the extent that its damages rest on volume of milk marketed, its assumption of a static situation would tend to underestimate its actual performance. In any event, viewed as a whole, the record simply does not support the district court's rejection of NFO's damage evidence in its entirety on the basis of the decreasing number of producers. Accordingly, on remand, the district court shall afford NFO the opportunity to present evidence on the basis of the existing record from which its post-combination and postcomputation membership dues claim may be reduced to reflect the decrease in the number of producers.34

E. Adjustments Not Included in the Basic Damage Estimate

NFO's damage estimate included two elements in addition to checkoff fees and dues. These were bargaining expense checkoff fees and post relief damages.

At various times during the damage period, NFO imposed a bargaining expense checkoff fee in addition to the three-cent per hundredweight basic checkoff fee. This checkoff fee was originally instituted in 1974 at two cents per hundredweight and was increased to six cents per hundredweight in 1976. This checkoff represents an additional element of NFO's claimed damages.

With respect to the two cent per hundredweight bargaining checkoff, NFO's damage calculation contained the assumption that, but for the restrictive practices, it would have imposed the checkoff in early 1972, some two and one-third years earlier than it was actually imposed. Therefore, it claims lost bargaining expense checkoff damages, not only for the additional milk it would have marketed absent the unlawful conspiracy, but also for the milk that it actually did market during the two and one-third year period.

In calculating the former volumes--the milk it would have marketed but for the restrictive practices--NFO relied on the same figures it used in calculating its basic checkoff losses (the three cent per hundredweight checkoff). This volume totaled 77,438,179 hundredweight. See NFO Trial Exhibit No. 1040.

In computing the volume of milk actually marketed, NFO used its own figures for the years 1972-74.35 This volume totaled 54,214,592 hundredweight. See Id.

NFO then multiplied the additional and actual volumes by the hypothetical two cent bargaining expense checkoff fee. This was done consistent with NFO's contention that, absent the appellees' unlawful conspiracy, the two cent checkoff would have been imposed at the beginning of 1972 instead of after one-third of 1974.

Thereafter, until 1976, the two cent checkoff applied only to the additional volume of milk NFO would have marketed but for the restrictive practices, reflecting the fact that NFO actually imposed the two cent checkoff in 1974. During 1976, NFO actually raised the bargaining expense checkoff fee to six cents per hundredweight. Reflecting the fact that the 1976 increase did not occur at the beginning of the year, NFO converted the 1976 checkoff fee to an annual figure of 5.7 cents per hundredweight.36 Finally, from 1977 to 1980, NFO used the full six cent per hundredweight figure, reflecting the assumption of static market conditions after 1977. See NFO Trial Exhibit No. 1005.

As a final step, after multiplying the actual and estimated volumes by the relevant checkoff fees, NFO deducted any additional expenses associated with obtaining and marketing the additional volumes.37 Using this method, NFO estimated bargaining expense losses of $2,394,605. See NFO Trial Exhibit Nos. 1040, 1041.38

The appellees challenge as unsupported by the record NFO's contention that, but for the unlawful conspiracy, it would have imposed the additional two cent checkoff two and one-third years earlier than it actually did. NFO, on the other hand, cites the testimony of Gene Potter, an administrative coordinator with NFO in support of its contention. The testimony indicates that NFO waited until 1974 to impose the two cent checkoff because the growth of the dairy marketing program was slower than had been anticipated. See Trial Transcript at 7985.

While it is clear that a failure to grow may serve as the basis for an antitrust damage award, see, e.g., T.V. Signal Co. of Aberdeen, 617 F.2d 1302, 1308 (8th Cir.1980) (allowing antitrust damages for sales that would have been made as the result of natural business expansion), in this case Potter's testimony simply lacks sufficient information to form the basis of the award NFO claims. Specifically, the testimony does not indicate the earlier time at which NFO would have instituted the checkoff nor does it indicate that the additional checkoff would have been two cents had it been imposed at an earlier time. In short, apart from a generalized allegation of slower than anticipated growth, the record is devoid of evidence supporting NFO's contention in this regard. Thus, the contention may not properly form the basis for damages arising out of the hypothetical two cent checkoff fee NFO claims it would have imposed in 1972.39

With respect to NFO's other bargaining expense checkoff claims, the record reflects that it actually imposed a two cent per hundredweight checkoff in 1974 and increased that checkoff fee to six cents in 1976. We can discern no reason from the record why this additional checkoff should not be recoverable on the additional volume of milk NFO claims it would have marketed but for the unlawful conspiracy.

A final element in NFO's damage claim is for the post-judgment period it would require to regain the market position it would have occupied but for the unlawful conspiracy. In calculating such damages, NFO assumed that relief would be granted in 1980 and that it would regain one-third of its mature share in each of the succeeding two years.40 If, during the two-year period, the 1980 share plus the regained share exceeded the estimated share, no damages were claimed.41 See NFO Trial Exhibit Nos. 1021, 1038.

Appellees challenge NFO's post relief damage claim as unsupported by the record. In assessing the claim, it is important to note that it is based on the same assumption that this Court did not completely accept with respect to NFO's post computation claim--that conditions on the orders would remain essentially static after 1977. See supra at 1303-1305. The 1980 figures, which served as the base from which one-third of the mature share was gained, necessarily involve the same assumptions as the post computation claim because they are actually 1977 figures which NFO assumed would remain unchanged through 1980. See id; see also, Trial Transcript at 9667.

The post relief claim is, however, more suspect than the post computation claim for a number of reasons. First, because it is dependent entirely on post computation figures, any error in those figures will necessarily be carried forward and compounded in the post relief figures. Second, the unlawful acts in furtherance of the illegal conspiracy occurred well before 1980. Thus, the post relief claim arises out of a time period even more distant from the unlawful conduct cited by this Court. Accordingly, NFO's post relief claim is not properly included as an element of its damages.42

Appellees finally contend that NFO's damage theory is defective because it fails to account for the effect on NFO of lawful competition. In support of their contention, appellees cite MCI Communications v. American Telephone & Telegraph Co., 708 F.2d 1081 (7th Cir.), cert. denied, 464 U.S. 891, 104 S. Ct. 234, 78 L. Ed. 2d 226 (1983). In that case, MCI filed an antitrust action against AT & T containing four counts and alleging twenty-two types of unlawful acts. At the close of MCI's case in chief, the district court directed a verdict in favor of AT & T on seven of the twenty-two alleged unlawful acts. The remaining fifteen were submitted to the jury.

As part of its proof at trial, MCI submitted damage evidence consisting of a lost profits study purportedly comparing the revenues it actually received or would receive with the revenues it expected to receive if it had been "undamaged." The figures for the "undamaged" MCI were derived from original business plans for the company.43 After adjusting for taxes, MCI claimed damages in the amount of $900,000,000.

The jury found in favor of MCI on ten of the fifteen alleged unlawful acts and awarded $600,000,000 in damages. The trial court trebled the award to $1,800,000,000. On appeal, the Seventh Circuit reversed with respect to three of the ten alleged unlawful acts on which the jury had ruled in favor of MCI.

Considering the sufficiency of the damage evidence, the Seventh Circuit remanded for a recalculation of MCI's damages because MCI's damage proof improperly attributed all of its losses to AT & T's illegal acts. In other words, the Court rejected MCI's damage computation because it did not "establish any variation in the outcome depending on which acts of AT & T were held to be legal and which illegal." MCI Communications, 708 F.2d at 1163.

Building upon MCI, the appellees argue that NFO's damage claim must be completely rejected because it was not adjusted on remand to reflect the fact that certain alleged unlawful acts were later found to be lawful.44 Such a position is at odds with the action of the MCI Court and cannot be accepted. While the damage award may have to be reduced so as to reflect only the damages caused by the unlawful conspiracy, it cannot be eliminated entirely.

At base, an antitrust plaintiff's damages should reflect the difference between its performance in a hypothetical market free of all antitrust violations and its actual performance in the market infected by the anticompetitive conduct. See Bigelow, 327 U.S. at 264, 66 S.Ct. at 579 (approving of comparison of profits and values affected by the conspiracy "with what they would have been in its absence under freely competitive conditions"); Dolphin Tours, Inc. v. Pacifico Creative Service, Inc., 773 F.2d 1506, 1511 (9th Cir.1985) ("In economic terms, the amount of damages is the difference between what the plaintiff could have made in a hypothetical free economic market and what the plaintiff actually made in spite of the anticompetitive activities."). Cf. Rose Confections, Inc. v. Ambrosia Chocolate Co., 816 F.2d 381, 394-95 (8th Cir.1987) (rejecting damage theory because it assumed that plaintiff would have received the benefit of alleged price discrimination).

Thus, if a plaintiff's business or property is diminished as a result of a competitor's lawful acts, no antitrust damages are recoverable for those acts. See MCI Communications, 708 F.2d at 1161. That is not to say, however, that a plaintiff's failure to prove with precision the amount by which the award should be reduced to reflect lawful competition will be fatal to its claim. Such a rule would ignore the Supreme Court's repeated recognition that, once the fact of injury is established, the plaintiff need only present evidence as to the amount of its damages sufficient to allow the finder of fact to make a just and reasonable estimate not based on speculation or guesswork. See, e.g., J. Truett Payne Co., 451 U.S. at 566-67 n. 5, 101 S. Ct. at 1930 n. 5 (citing Story Parchment Co., 282 U.S. at 562, 51 S.Ct. at 250).

An antitrust plaintiff's damage claim is not, as the appellees suggest, rendered speculative or unreasonable merely because it fails to provide for a specific reduction in the event that allegations of certain unlawful conduct are rejected. In other words, if an antitrust plaintiff alleges that the defendant engaged in unlawful acts A, B, C, and D, and acts C and D are later rejected as a basis of liability, it does not automatically follow that the damage award must be reduced. Rather, if acts A and B support the entire damage award, it must be sustained.45

Since the determination whether the unlawful acts can support the damage claim is largely a factual one, we must remand the case to the district court. On remand, the court shall afford the parties an opportunity to point out evidence in the existing record from which it may determine whether NFO's damage claim approved of herein should be reduced and, if so, the amount of such reduction. In considering such a reduction, we offer several guidelines to the district court.

First, in our prior opinion, we stressed the Supreme Court's admonition in Continental Ore Co. v. Union Carbide & Carbon Corp., 370 U.S. 690, 699, 82 S. Ct. 1404, 1410, 8 L. Ed. 2d 777 (1962), to avoid tightly compartmentalizing the plaintiff's proof. NFO II, 687 F.2d at 1193-94, 1208. We recognized that some of the appellees' conduct, if viewed in isolation, was lawful, yet formed part of the mosaic that constituted the unlawful conspiracy. See NFO II, 687 F.2d at 1193.

Separating the consequences of the unlawful from the consequences of the lawful will most certainly prove difficult. As the Supreme Court has stated, "[t]he vagaries of the marketplace usually deny us sure knowledge of what the plaintiff's situation might have been in the absence of the defendant's antitrust violation." J. Truett Payne Co., 451 U.S. at 566, 101 S.Ct. at 1929.

Yet, the fact that the appellees' illegal conspiracy was composed of lawful and unlawful conduct so tightly intertwined as to make it difficult to determine which portion of the damages claimed were caused by the unlawful conduct should not diminish the recovery. Similarly, the Court should recognize that the harmful consequences of certain unlawful conduct may have been exacerbated by otherwise lawful conduct. In such a situation, the fact that lawful conduct contributed to additional injury should not prohibit recovery for that injury.

Second, this case presents a situation in which one of the defendants, AMPI, ordered the destruction of documents allegedly relevant to this case. The destruction was characterized as involving "admitted bonfires." NFO II, 687 F.2d at 1205. In our prior opinion we stated:

We can only describe AMPI's conduct as outrageous. Obviously, the relevance of and resulting prejudice from destruction of documents cannot be clearly ascertained because the documents no longer exist. Under the circumstances, AMPI can hardly assert any presumption of irrelevance as to the destroyed documents. On this record, the district court properly could have imposed the most severe sanctions upon AMPI--dismissal of its claims and default judgment against it on NFO's claim. Nonetheless, we cannot say it was an abuse of discretion not to do so. It was error, however, not to draw factual inferences adverse to AMPI on matters undertaken in or through offices and individuals involved in the destruction of documents.

Id. at 1205-06 (citations omitted).

Therefore, the fact that AMPI may have engaged in conduct in addition to the antitrust conspiracy outlined in our prior opinion which made proof of the amount of damage more difficult must be considered on remand.

Third, review of the record and our prior opinion establishing the fact of injury leave this Court with the definite and firm conviction that the amount of damages suffered by NFO as a result of the illegal conspiracy was substantial. The record simply will not support a finding that NFO is entitled to only nominal damages. In short, the appellees must not be allowed to "clip NFO's wings and then escape liability on the grounds that NFO cannot fly." Id. at 1210.

Finally, we emphasize that, in assessing the evidence as to the amount of damage, precision is not required. Rather, the district court must weigh the existing record evidence and arrive at a just and reasonable figure. We realize the task the district court is called upon to perform is not easy and is made all the more difficult by the formidable record in the case. Nonetheless, it is one that must be undertaken, lest the appellees be allowed to profit from what can only be characterized as a broad conspiracy undertaken in blatant disregard of the antitrust laws.

In addition to the damage issues, we remanded for consideration of injunctive relief. On remand, the district court was to consider, inter alia, the extent to which consent decrees obtained by the United States in actions against Mid-Am and AMPI already provided NFO with sufficient relief.46

On remand, the appellees argued that injunctive relief is inappropriate because there is no evidence of any impending violation of the antitrust laws, and the injunction proposed by NFO merely duplicates sufficient relief already provided by consent decrees explicitly covering Mid-Am and AMPI and implicitly covering CMPC.

NFO responded that the consent decrees do not provide sufficient relief and, in any event, are not binding on CMPC. In particular, NFO argued that the decrees do not contain adequate enforcement mechanisms and that it is entitled to restorative injunctive relief designed to place it in the competitive position it would have enjoyed absent the illegal conspiracy.

The district court, with one exception, denied NFO's proposed injunctive relief as unwarranted and duplicative of the Mid-Am and AMPI consent decrees. The exception had to do with a proposed prohibition of sham litigation. With respect to sham litigation, the district court entered an injunction, despite the fact that it was satisfied the consent decrees were broad enough to cover sham litigation. It did so because the issue had never actually been decided and our prior opinion placed great emphasis on violations involving sham litigation. See 614 F. Supp. at 801-02.

Under the Clayton Act, 15 U.S.C. Sec. 26, the district court has the duty and broad authority to fashion relief that will terminate the unlawful acts found to have occurred, ensure that they will not recur, and eliminate their consequences. Paschall v. Kansas City Star Co., 695 F.2d 322, 335 (8th Cir.1982), rev'd on other grounds, 727 F.2d 692 (1984) (en banc) (quoting National Society of Professional Engineers v. United States, 435 U.S. 679, 697, 98 S. Ct. 1355, 1368, 55 L. Ed. 2d 637 (1978); United States v. United Shoe Machinery Corp., 391 U.S. 244, 250, 88 S. Ct. 1496, 1500, 20 L. Ed. 2d 562 (1968)); see also United States v. Glaxo Group Limited, 410 U.S. 52, 64, 93 S. Ct. 861, 868, 35 L. Ed. 2d 104 (1973) ("The purpose of relief in an antitrust case is 'so far as practicable [to] cure the ill effects of the illegal conduct, and assure the public freedom from its continuance.' "). In carrying out its duty, the district court has "large discretion to model [its] judgment to fit the exigencies of the particular case." Paschall, 695 F.2d at 335 (quoting International Salt Co. v. United States, 332 U.S. 392, 400-01, 68 S. Ct. 12, 17, 92 L. Ed. 20 (1947)).

On appeal, NFO claims that it is entitled to injunctive relief even though the conduct complained of may be prohibited by the consent decrees. We disagree. Our prior opinion clearly recognized that NFO is not entitled to injunctive relief to the extent other existing relief is adequate. See NFO II, 687 F.2d at 1210, n. 38, 39. We have carefully reviewed the AMPI and Mid-Am consent decrees and have compared them with the injunction proposed by NFO. We conclude that, with the exception of the prohibition against sham litigation noted by the district court, they offer NFO substantially the same relief. In short, provided the provisions of the consent decrees can be reasonably enforced by NFO, there is nothing to be gained by entering an injunction that substantially duplicates the relief already available. Cf. Hawaii v. Standard Oil Co., 405 U.S. 251, 261, 92 S. Ct. 885, 890, 31 L. Ed. 2d 184 (1972) ("While the United States Government, the governments of each State, and any individual threatened with injury by an antitrust violation may all sue for injunctive relief against violations of the antitrust laws, and while they may theoretically do so simultaneously against the same persons for the same violations, the fact is that one injunction is as effective as 100, and concomitantly, that 100 injunctions are no more effective than one.").

Alternatively, NFO contends that the enforcement mechanisms in the consent decrees are inadequate to afford it meaningful relief. Pursuant to the consent decrees, NFO may bring an enforcement action by filing a request with the Assistant United States Attorney General in charge of the Antitrust Division and the district court alleging noncompliance, the injury resulting therefrom, and the relief requested. The Assistant Attorney General then has twenty days to respond. A copy of the response must be sent to NFO and to the district court. If NFO is not satisfied with the government's response, it may petition the court to exercise its independent jurisdiction and to direct enforcement proceedings on its own motion. See United States v. Associated Milk Producers, Inc., 394 F. Supp. 29, 57-58 (W.D.Mo.1975); United States v. Mid-Am, 1977 Trade Cases 61,508. Although the enforcement provision places an administrative burden on NFO and may involve a small delay, in view of the fact that the decrees clearly contemplate full and complete private enforcement, we cannot say that the district court erred in finding that the consent decrees offered NFO adequate relief.

NFO also claims that the consent decrees are inadequate because they are not binding on CMPC. The district court rejected this claim as "untenable." NFO III, 614 F. Supp. at 801. CMPC is a federation of cooperatives of which AMPI is a member and the sole marketing agent. The consent decree entered against AMPI is binding on "all persons in active concert or participation" with AMPI.47 United States v. Associated Milk Producers, Inc., 394 F. Supp. at 50. Accordingly, we affirm the district court in this regard.

Finally, NFO contends it is entitled to the restorative relief. The proposed relief consisted of five parts. Part one required the appellees, for a period of one year, to allow their members to terminate marketing agreements at any time on thirty days notice in order to join NFO. Part two required Mid-Am and AMPI to purchase milk from NFO under certain specified terms and conditions for a period of five years. Pursuant to part three, if the appellees acquired a milk receiving or processing plant which bought milk from NFO during the twelve months preceding the acquisition, they would be required to purchase the same volume of milk NFO sold the plant in the preceding twelve months at the same price and terms as other milk received at the plant. Part four required CMPC to allow NFO to participate in its dairy surplus disposal program for the next five years under the same terms and conditions as other CMPC members. Part five prohibited the appellees from block voting their members' stock if the effect would be to terminate or amend any existing federal market order. NFO Proposed Injunction, Brief of Appellant, Addendum II, part p. at 7-10.

The district court carefully considered NFO's proposed restorative relief and found it unwarranted, duplicative, or inappropriate. We agree and affirm the district court's rejection of the requested restorative relief.

In addition to the appeal brought by NFO, nine individual NFO members appeal the district court's refusal to certify a class of NFO member dairy farmers seeking to recover damages.

Shortly after this case was filed, it became apparent that issues concerning class certification would play an important role. As early as June of 1971, issues involving requests by both NFO and Mid-Am to certify a class of their respective members were discussed. Subsequently, the district court expressed to the parties on numerous occasions its view that other devices superior to a class action were available and that it was not inclined to grant the class certification motions. Accordingly, the district court requested the parties to try to reach an agreement on alternative methods for resolving standing and damage distribution issues. Attempts to reach such an agreement failed, however, and at a pretrial conference in 1974, the district court denied NFO's request for class certification "without prejudice to later suggestions as to [the resolution of] the problems of distribution of damage." Transcript of Pretrial Conf. No. 20 at 31.48

After additional discussion, NFO moved the district court to reconsider the denial or, in the alternative, to certify the issue for interlocutory review pursuant to 28 U.S.C. Sec. 1292(b). The district court denied the motion.49 On January 21, 1981, the district court entered a final judgment in phase II of the case (involving NFO's counterclaim). At that time, no further motion with respect to class certification had been raised. The judgment stated, "It is Ordered and Adjudged that the plaintiff, The National Farmers Organization, Inc., take nothing and that this action be dismissed on the merits." Alexander v. National Farmers' Organization (W.D.Mo.1981).

NFO thereafter appealed to this Court from the judgment. Neither the individual class members nor the putative class appealed either the judgment or the previous denial of class certification. On appeal, this Court examined the standing and damage issues and held that NFO could not recover damages resulting from the lower price it received for its milk. Instead, we held that such damages could be recovered only by the individual NFO members. NFO II, 687 F.2d at 1209.

On remand, the individual NFO members again brought a motion to certify a class. The district court denied the motion, finding that NFO made a deliberate election to claim the price reduction damages as belonging to it and not to the class and that the failure to appeal the district court's original interlocutory order and judgment barred consideration of the class certification issue. Alternatively, the district court held that class certification would be improper because individual factual and legal issues concerning the amount and fact of damage predominated.

We agree with the district court that consideration of the class certification motion is barred. In Federated Department Stores, Inc. v. Moitie, 452 U.S. 394, 101 S. Ct. 2424, 69 L. Ed. 2d 103 (1981), the Supreme Court addressed a situation in which two parties failed to join in a direct appeal from an adverse judgment and, in a later collateral proceeding, sought to take advantage of a reversal obtained by the appealing parties. The Court ruled against the non-appealing parties, holding that a final judgment on the merits, even if wrong or based on a subsequently overruled legal principle, precludes the parties or their privies from relitigating issues that were or could have been raised in the original action. Id. at 398-99, 101 S.Ct. at 2427-28.

The putative NFO class argues that Moitie is not applicable to this case because the 1981 judgment, by its terms, was entered against NFO and not against the class. We disagree. The fact that the 1981 judgment did not specifically run against the class claimants did not preclude their appeal from it. In United Airlines, Inc. v. McDonald, 432 U.S. 385, 97 S. Ct. 2464, 53 L. Ed. 2d 423 (1977), the Supreme Court addressed a situation in which a United Airlines flight attendant attempted to bring a class action under Title VII of the Civil Rights Act of 1964, 42 U.S.C. Sec. 2000e et seq., challenging a rule requiring female flight attendants to remain unmarried as a condition of employment. The district court denied class certification, finding that the class could properly consist of only those attendants who, upon discharge for marriage, had filed claims under a fair employment statute or grievances pursuant to United's collective bargaining agreement. Thus, because the class so limited contained only thirty members, the district court refused certification on the ground that the class was not sufficiently numerous to render joinder of all members impracticable. The case then proceeded to a judgment in favor of the individual plaintiffs.

After entry of the judgment, McDonald, a member of the flight attendant class, as originally defined, who failed to challenge her discharge either to the EEOC or under the collective bargaining agreement, sought leave to appeal the denial of class certification despite the fact that individual plaintiffs had decided not to appeal. The district court denied the request as untimely. The Court of Appeals reversed, finding both that the request was timely and that the district court erred in narrowly defining the class. United appealed only the timeliness issue. The Supreme Court affirmed, finding that, "the entry of final judgment made the adverse class determination appealable." Id. at 394, 97 S.Ct. at 2469-70.

The basic premise of McDonald --that a member of a class denied certification may appeal that denial upon entry of a final judgment in the case in which certification was denied--is directly applicable to this case. As in McDonald, the NFO class claimants were free to appeal the adverse class certification ruling, as well as any other rulings of the district court, once final judgment was entered on NFO's claims. Their failure to do so is preclusive.

To hold otherwise defies common sense. It would only be redundant for the district court to enter a final judgment against the class after it had already refused to recognize it. Moreover, since the claims of the class and of NFO arise out of many of the same operative facts and require application of the same laws, considerations of judicial economy and consistency argue strongly in favor of requiring the appeals of the class and of NFO to be heard at the same time. Accordingly, we affirm the district court's refusal to certify a class of NFO dairy farmer members.

The antitrust laws provide that a plaintiff who recovers damages or who "substantially prevails" in an action for an injunction shall recover the costs of suit and a reasonable attorney's fee. 15 U.S.C. Secs. 15 and 26. Mindful of the admonition in Hensley v. Eckerhart, 461 U.S. 424, 437, 103 S. Ct. 1933, 1941, 76 L. Ed. 2d 40 (1983), to avoid a "second major litigation," the district court denied, without prejudice, NFO's motion for attorney's fees and costs pending appeal to this Court.50

Specifically, the district court noted that under the standards set forth in Hensley and in this Court's application of Hensley in Williams v. Mensey, 785 F.2d 631 (8th Cir.1986), it must award attorney's fees in an amount that is reasonable in relation to the results obtained. Alexander v. National Farmers' Organization, 637 F. Supp. 1487, 1499-1500 (W.D.Mo.1986). In this regard, the district court stated:

[A]ny attorney's fee award that this Court may make at this time, would necessarily have to be based on the results that NFO has up to this point obtained in the district court. Any award of attorney fees that would be based on judgments that are subject to immediate appellate review would be nothing more than an exercise in futility unless it is assumed that this Court will eventually be affirmed on the judgments that will be entered on both the damage issue and equitable relief issue. Based on its experience with the parties, the Court is confident that the parties will not be in agreement that such an assumption should be made. We are accordingly satisfied that the extent of NFO's success or lack of success should be finally determined on appeal before this Court embarks on what promises to be a second major litigation in connection with NFO's request for an attorney's fee award.

Id. at 1500.

Thus, rather than undertaking the time-consuming task of determining a reasonable attorney's fee, only to see the effort overturned on appeal, see, e.g., Copperweld Corp. v. Independence Tube Corp., 467 U.S. 752, 104 S. Ct. 2731, 81 L. Ed. 2d 628 (1984) (overturning entire attorney's fee award); Paschall v. Kansas City Star Co., 727 F.2d 692 (8th Cir.1984) (en banc) (same), the district court wisely deferred ruling on attorney's fees and costs pending appeal.

In light of our decision herein finding that NFO is entitled to a substantial damage award in addition to the injunctive relief obtained in the district court, it is clear that the time has now come for the district court to address the question of attorney's fees and costs. Accordingly, we remand to the district court with directions to it to determine the amount NFO should be awarded for reasonable attorney's fees and costs it incurred in the district court.51

Finally, in accordance with our prior opinion in this matter, see Alexander v. National Farmers' Organization, Inc., 696 F.2d 1210 (8th Cir.1982), NFO may move this Court for an award of expenses incurred in pursuing this appeal. Such expenses may include attorney's fees, court costs, and other miscellaneous costs.

In summary, from our prior opinion and the record in this case it is clear that NFO has established that the appellees engaged in a broad conspiracy in violation of the antitrust laws and that the conspiracy caused it a substantial amount of damage. Accordingly, we remand to the district court for computation of the amount of damages. In computing that amount, the district court shall accept the "test market" theory propounded by NFO as a valid base from which to determine the amount of damages and shall use the $7,403,374 claimed under the theory as a starting point.

In accordance with this opinion, that figure shall be reduced by $1,604,098, reflecting our rejection of NFO's claim that it would have imposed a two cent bargaining expense checkoff fee in 1972, rather than in 1974, and by $764,618, reflecting disallowance of post relief damages. In addition, the district court shall make certain the damage award reflects the expenses associated with handling the additional volume of milk NFO would have marketed had it achieved the 10.5% test market share.52

The district court may also, in its discretion, further reduce the damage award to reflect the following:

(1) the voluntary checkoff fee contributions of NFO members pooling milk on the damage orders through milk marketers other than NFO;

(2) an appropriate reduction of NFO's post order combination and post computation membership dues claim due to an overall decrease in the number of producers; and

(3) an adjustment, if any is necessary, for the effects of alleged unlawful conduct later found to be lawful.

In considering the discretionary reductions, the district court shall keep in mind the admonition in the Supreme Court opinions and in this opinion that NFO need not prove its damages with unnecessary precision. Rather, the damages awarded need only be just and reasonable in light of the practical limits of proof in a case of this nature. Finally, once the district court has determined a just and reasonable award, that amount shall be trebled in accordance with section 4 of the Clayton Act, 15 U.S.C. Sec. 15.

We also deny NFO's requests for additional injunctive relief and to certify a class of its dairy farmer members. We remand to the district court for a determination and award of a reasonable attorney's fee and costs. NFO may make application to this Court for any attorney's fees and costs it reasonably incurred in pursuing this appeal.

Accordingly, the judgment of the district court is affirmed in part, reversed in part, and remanded to the district court for further proceedings consistent with this opinion.