Basf Corporation, Plaintiff-Appellant/cross-Appellee v. Old World Trading Company, Incorporated, Defendant-Appellee/cross-Appellant , 41 F.3d 1081 ( 1994 )

CUDAHY, Circuit Judge.

BASF and Old World Trading Company, a privately held Illinois corporation, are competitors in the antifreeze market. In the 1980s, Old World advertised that its antifreeze met certain industry specifications; BASF contended that Old World had never actually tested its antifreeze to determine whether it met the specifications. BASF brought suit for false advertising under the Lanham Act, 15 U.S.C. § 1125(a) (1982), and the Illinois Consumer Fraud and Deceptive Business Practices Act, 815 ILCS 505 et seq. Old World counterclaimed for common law commercial defamation and product disparagement. The district court found Old World liable under the Lanham Act and state law, and awarded BASF $2,498,726 in lost profits, $1,737,778 in prejudgment interest, and nearly $275,000 in costs. The district court found for Old World on its counterclaim, and awarded $1. The district court granted in part and denied in part BASF’s motion for reconsideration and to amend the judgment. BASF appealed, seeking a tenfold increase in the damage award. Old World cross-appealed, contending that its advertisements were not false under the Lan-ham Act, and that in any event BASF had not proven any damages. After the appeal was filed, the district court awarded attorneys’ fees to BASF, but left the final amount undetermined pending the parties’ agreement on an appropriate methodology for calculating fees. Old World also appeals from the order awarding BASF attorneys’ fees. We affirm the district court in all respects.

I. FACTS

A. Old World’s antifreeze

Antifreeze, as the name implies, remains liquid at temperatures where water freezes, and is used as a water substitute to dissipate automotive engine heat. Antifreeze consists primarily of ethylene glycol (EG) mixed with water and chemicals added to prevent corrosion. Different metals corrode at different rates, so when auto manufacturers began using more aluminum components in the early 1980s, antifreeze manufacturers had to spend a considerable amount of money tinkering with anti-corrosion chemicals to meet the auto manufacturers’ new specifications. “Specifications” are minimum performance standards adopted by each auto manufacturer. Whether an antifreeze meets an auto manufacturer’s specifications is determined by its performance in a battery of specific (and, for the most part, complicated and expensive) tests, developed in part by the American Society of Testing and Materials (ASTM).

In 1982, after the auto manufacturers published new specifications, Old World began selling a new aluminum protective antifreeze. Old World advertised that its antifreeze met all of the auto manufacturers’ specifications as well as the “Cummins specification,” a special heavy duty truck standard.¹ No other antifreeze manufacturer could make the same claims. BASF, for example, claimed only that it was approved by manufacturers for factory fill, but did not claim to meet the Cummins specification. Thus Old World had an advantage, since a customer buying for resale would only need to buy one antifreeze to meet all specifications.

Old World did not have its own manufacturing facilities, but like many antifreeze companies, contracted out its antifreeze blending. Old World’s antifreeze formula was developed by the Dearborn Chemical Company, a company with experience in manufacturing coolants but not in blending antifreeze. Dearborn either failed to perform or incorrectly performed certain tests required by several of the specifications Old World claimed to meet. Specifically, Dear-born failed to perform the Ford dynamome-ter test or fleet test, as required to meet Ford specification ESE-M97B44-A, and a General Motors 9066P test as required to meet GM specification 1825-M (1981). Dear-born also incorrectly performed the ASTM D-2809 test required by specifications SAE 11J1034, ASTM D-3306, GM 1899-M, and MCAT RP 302A.

Old World never requested that Dearborn perform a specific test or requested that Dearborn supply written test data. Instead, Old World claims it relied on Dearborn’s assurances that the formula met specifications, in particular Dearborn’s representation that it had conducted the dynamometer test required by the Ford specification, and that the Dearborn Formula met the Ford and GM specifications. But the district court found it incredible that Old World would have relied on Dearborn’s word. Old World’s chief technical officer, Jess Starkey, a long-time member of the ASTM committee that designed the industry specifications, had assisted in preparation of various test protocols and knew, for example, that Dearborn had not followed one of the protocols during its tests.

In June 1983, Old World and Dearborn modified the antifreeze formula. Old World claims that Dearborn’s representatives assured Old World that the new formula met the same specifications as the earlier formula.² Some time after 1983, Starkey altered the Dearborn formula even further by reducing the amount of EG and other chemicals. No additional tests were performed after these modifications, although Old World claims that once Dearborn was aware of the adjustments, it informed Old World that the formula met the specifications. This modified formula was sent to Old World blenders from 1983 until the introduction of a new formula for the 1987-88 antifreeze season.³

In October 1985, BASF wrote to Old World challenging the accuracy of Old World’s claim that it met the Ford specifications. Dearborn antifreeze was first subjected to rigorous performance testing under the protocols in January 1986, when a BASF sales representative purchased some Old World antifreeze and ran the ASTM D-2809, D-1384, D-4340, D-2580 and similar Ford tests. Old World passed all of the tests. BASF attempted to conduct a Ford dyna-mometer test, but the test was never concluded. In June 1986, Dearborn stopped making advertising claims about the Ford specification. BASF brought the present action on May 21, 1986.

In December 1986, Old World purchased antifreeze manufacturing facilities and a new formula (“Peak”) from Northern Petro. The new formula met all auto manufacturers’ specifications, but not the Cummins standard. Old World stopped advertising that it met all of the specifications. The district court found (although BASF vigorously disagrees) that Old World sold no more Dear-born formula after March 31, 1987.

Following a bench trial, the district court issued an opinion on May 21, 1992 holding that Old World’s representations violated § 43(a) of the Lanham Act, 15 U.S.C. § 1125(a). The district court found that BASF had produced overwhelming evidence that Old World’s testing was insufficiently reliable to justify its claims that the antifreeze met the auto manufacturers’ performance specifications. The district court also found that Old World’s advertisements were literally false, since the antifreeze had not been subjected to tests which must be passed to meet the specifications. Numerous experts had testified that Old World could not truthfully represent that its antifreeze met the specifications unless all of the required tests had been run and passed. The district court also found that purchasers were actually misled by Old World’s representations. In addition, the court found Old World liable under the Illinois Act.

B. Damages

The antifreeze market is highly price-driven, resulting in a very competitive, low margin business. In 1983, the antifreeze market was dominated by eight companies, although by 1987 both Dow and Northern Petro had left the market.⁴ Although the antifreeze industry was relatively stagnant during the mid-1980s, both Old World and BASF gained market share. Old World’s market share increased from 4% to 15%, with Old World’s 1986 acquisition of Northern Petro accounting for 6% of this increase. Meanwhile, BASF’s market share increased from 15% to 18%.

BASF and Old World competed in the “private label” segment of the antifreeze market. This is the section of market where antifreeze is sold under a customer’s trade name (for example, BASF would sell to Quaker State, who marketed the antifreeze under its own label). From 1983-1987, BASF was the largest producer of private label antifreeze. By the winter of 1986, several competitors (Northern Petro, Union Carbide, Conoco and Shell) had dropped out of the private label business, leaving only BASF, Old World and Texaco in the market.

Old World contends that its market share grew more rapidly than BASF’s because of Old World’s increased price sensitivity. The price of antifreeze is determined in great part by the price of EG, its primary component. BASF produces its own EG rather than purchasing it on the open market; Old World buys its EG on the market, and thus could take advantage of declining EG prices in the mid-1980s. Old World also contended that BASF locked itself into a high-pricing strategy trying to market its own trademark in the mid-1980s, and could not react quickly to market changes.

1. Lost profits

During the years in question (1983-87), BASF never made a profit on its sales of antifreeze. However, BASF’s antifreeze sales did provide sufficient revenue to meet all the variable costs of production and to contribute toward fixed costs. To calculate lost profits, BASF’s damages expert first calculated the contribution to overhead per gallon by subtracting total variable costs from sales revenues. The overhead contribution ranged from $.71/gallon to $.97/gallon.⁵

BASF next calculated the alleged shortfall in its sales caused by Old World’s misrepresentations, based on certain key assumptions: that but for Old World’s misrepresentations, Old World would not have made certain sales; that BASF would have made these sales; and that BASF would have made the same volume of sales as Old World. The shortfall was thus the equivalent of Old World’s sales to former BASF customers lost through false advertising. Multiplying the shortfall in sales by the contribution to overhead per gallon, BASF estimated lost profits of $22,713,000.

In calculating the shortfall, BASF contended that it lost ten customers to Old World. As to the six smallest of these customers, whose pin-chases accounted for $1,670,063 of BASF’s lost profits figure, the district court found that BASF had presented insufficient evidence that it would have made these sales but for Old World’s misrepresentations. The district court considered independently the evidence about the remaining four largest customers:

• Quaker State’s TAG division had purchased antifreeze from BASF and Texaco for years, but in 1986 switched to Old World. A TAG official testified that BASF’s prices were too high, but also testified that it would buy only from a supplier who met the Ford and GM specifications. BASF estimates that Quaker State’s move resulted in lost profits of $405,200.

• Until 1985, Valvoline had purchased antifreeze solely from BASF. Valvoline added Texaco as a supplier in 1986 because BASF’s prices were too high, and in 1987 added Old World, based on Old World’s representations that it met all specifications. In 1988, Valvoline dropped BASF, claiming that it was not sufficiently sensitive to price. BASF calculated lost profits of $1,230,801 through loss of Valvoline.

• Citgo had relied on BASF to meet all its antifreeze requirements in 1983 and 1984, but in 1985 bought from both BASF and Old World. In 1986, Citgo terminated its agreement with BASF over BASF’s perceived price insensitivity, and subsequently purchased all requirements from Old World. Citgo required its vendors to meet the auto manufacturers specifications. BASF claimed lost profits of $6,108,033.

• Phillips had usually purchased antifreeze from Dow, which left the private label market in 1985. For 1986, Phillips bought equal amounts from BASF and Old World. Phillips had difficulty selling BASF antifreeze due to its high prices, and informed BASF it would only be kept as a supplier if it lowered its prices. When BASF refused to do so, Phillips switched to Old World for 1987. BASF claims lost profits of $10,-081,519.

After considering the evidence from the various accounts, the district court found BASF’s calculations of the shortfall due to lost sales “grossly inflated.” The district court found that BASF’s calculations did not take into account that BASF’s pricing policies drove away customers. Moreover, BASF had not established that the business Old World gained would otherwise have gone to BASF; Texaco remained a viable competitor the entire time, and at least until the winter of 1986 there were other competitors in the private label market as well. BASF’s own pricing information — the forms “807” that BASF sales personnel filled out to request a price adjustment to meet competition — indicated that the majority of BASF’s price competition came from suppliers other than Old World. Since the evidence showed that BASF would not have made all of the sales absent Old World’s misrepresentations, the district court rejected BASF’s calculation of the shortfall.

On the other hand, the district court agreed that, despite customer complaints over pricing, BASF would have made some of the sales but for Old World’s misrepresentations. It concluded that BASF’s share of sales lost to Old World would equal its share of the non-Old World antifreeze market, since BASF would have competed with these other manufacturers absent Old World. The court adopted BASF’s calculations of gallon sales to the ten customers and the overhead contribution per gallon. It then multiplied this figure by BASF’s share of the non-Old World market in each year to determine BASF’s lost profits award. The district court arrived at a lost profits figure of $2,498,726.⁶

On September 9, 1992, the district court reconsidered its market share analysis. The district court had calculated damages based on BASF’s share of the entire antifreeze market (branded and private label), rather than merely the private label market. BASF contended that its share of the private label market was significantly higher than its share of the general market — 28% of private label in 1985 rising to 34% in 1988, instead of the 15.6% of all antifreeze sales rising to 21.2%, as calculated by the district court. But the district court denied BASF’s motion to amend the judgment, finding that BASF had adduced no proof to support its claims about the shares breakdown of the private label market. Absent any proof of the comparative size of the private label market shares, the district court found the general antifreeze market to be the best approximation.

2. Adjustments to damages

After calculating lost profits of $22,713,010, BASF adjusted its figure to arrive at its final damages claim of $34,275,869. BASF claimed various adjustments to account for elasticity in demand resulting from BASF’s higher prices; BASF’s lack of capacity to manufacture the entire volume of its requirement of EG and the consequent necessity of buying on the open market; price erosion from Old World’s lower prices; and prejudgment interest. The district court found no basis to adjust the damages figure upward to account for price erosion or downward to account for price elasticity. The district court did award BASF $1,737,778 in prejudgment interest.

3. Equitable remedies

In addition to its own damages, BASF requested disgorgement of Old World’s profits. The district court determined that the damages award was large enough to serve as an effective deterrent, and therefore declined to order disgorgement. BASF also argued that its damages award should be enhanced under the Lanham Act’s provision allowing a district court to enhance an award up to three times the amount of damages. 15 U.S.C. § 1117(a). The district court denied the enhancement, finding that the damages award adequately compensated BASF and that an enhancement would be punitive.

4. State law

The district court found that BASF’s damages were the same under the Illinois Act as under the Lanham Act. The district court refused to award punitive damages, finding that Old World’s conduct was not sufficiently egregious to merit punitive damages, and that there was no need for further deterrence.

5. Attorneys’ fees

The Lanham Act provides for attorneys’ fees to a prevailing plaintiff in “exceptional cases.” 15 U.S.C. § 1117(a). In its original order, the district court concluded that, although Old World’s actions were not malicious, they were deliberate. Accordingly, the district court ruled that BASF was entitled to “at least some” of its attorneys’ fees, but did not determine the amount of fees. BASF requested just over $4 million in attorneys’ fees. On August 30, 1993, the district court rejected about $1.5 million of the $4 million requested by BASF, and ordered the parties to develop an appropriate mechanism by which the court could properly assess the remaining fees.

C. Issues on appeal

BASF appeals the district court’s award of damages, arguing that the district court: (1) incorrectly calculated lost profits; (2) erred by refusing to require disgorgement or to enhance the damages award; (3) neglected to include damages for the 1988 antifreeze year; and (4) erred by refusing to award punitive damages under state law. Old World cross-appealed, claiming that the district court applied an erroneous legal standard to its determination of liability under the Lanham Act. Old World also argued that the district court erred in calculating BASF’s damages, maintaining that BASF had not shown any lost profits and had used an incorrect measure of BASF’s profits; and that BASF was not entitled to any attorneys’ fees. We consider first Old World’s cross-appeal on liability, and both the parties’ damages issues together. We review the district court’s conclusions of law de novo, and its findings of fact for clear error. Sandoz Pharmaceuticals Corp. v. Richardsonr-Vicks, Inc., 902 F.2d 222, 229 (3rd Cir.1990).

II. LIABILITY UNDER THE LANHAM ACT

A. Literal falsity

Section 43(a) of the Lanham Act prohibits the use of “any false description or representation” in advertising.⁷ To establish liability, a plaintiff must show that the challenged advertisement is literally false, or, if literally true or ambiguous, that it is “misleading in context, as demonstrated by actual customer confusion.” Abbott Laboratories v. Mead Johnson & Co., 971 F.2d 6, 13 (7th Cir.1992); Castrol Inc. v. Pennzoil Co., 987 F.2d 939 (3rd Cir.1993); but see Pennzoil, 987 F.2d at 951 (Roth, J., dissenting) (cogent argument that courts must consider customer confusion even when advertisement is literally false). But literal falsity is the only issue here.

This case poses an interesting question of what proof is necessary to show that Old World’s statement that its antifreeze “meets the Ford and GM specifications” is false. Old World claims that BASF was required to generate affirmative proof that the statement was literally false. BASF, on the other hand, argues that it only needed to show that the tests supporting Old World’s statement were not sufficiently rehable to support the claim.

Both parties’ arguments find support in overlapping strands of case law from the Second and Third Circuits. In Sandoz Pharmaceuticals, 902 F.2d at 229, the Third Circuit articulated the “literal falsity” standard advocated by Old World. Sandoz rejected a plaintiffs argument that it could prove an advertisement false by showing that it was not substantiated. 902 F.2d at 227-30. The court held “it is not sufficient for a Lanham Act plaintiff to show only that the defendant’s advertising claims ... are inadequately substantiated ... the plaintiff must also show that the claims are literally false or misleading to the public.” Id. at 229; see also Pennzoil, 987 F.2d at 943-45; Toro Co. v. Textron, 499 F.Supp. 241, 251 (D.Del.1980) (Stapleton, J.).

BASF argues that Sandoz’s burden of affirmative proof of literal falsity does not apply to completely unsubstantiated claims. The Sandoz court had observed that “in such a case, there is a plausible argument that the claim is literally false because the advertiser has absolutely no grounds for believing that its claim is true.” 902 F.2d at 228 n. 7. But Pennzoil, 987 F.2d at 939, suggests a narrower definition of “literal falsity.” In Pennzoil, the court held that Pennzoil had violated the Lanham Act by advertising that its oil “outperforms any leading motor oil against viscosity breakdown.” Id. Pennzoil pointed to tests supporting its claim, but the district court found that these tests were not the industry standards for measuring viscosity breakdown. There is no suggestion in the case that Pennzoil had conducted the appropriate tests. But Pennzoil’s liability hinged not on the demonstrated inadequacy of its own testing, but on Castrol’s evidence that, when subjected to the appropriate tests, Pennzoil oil performed no better and possibly worse than Castrol oil. Id. Pennzoil reaffirmed Sandoz’s requirement that a plaintiff must affirmatively prove literal falsity rather than rely on proof that the defendant’s claims are unsubstantiated.

BASF, on the other hand, relies on Procter & Gamble Co. v. Chesebrough-Pond’s Inc., 747 F.2d 114, 119 (2d Cir.1984) and its progeny to support its position that falsity can be established by a successful attack on the underlying data supporting a defendant’s claim, rather than by requiring proof negating the claim (in this case, proof that Old World in fact could not meet specifications). See also ALPO Petfoods v. Ralston-Purina Co., 913 F.2d 958 (D.C.Cir.1990); McNeil-P.C.C. v. Bristol-Myers Squibb Co., 938 F.2d 1544 (2d Cir.1991); Mylan Laboratories v. Matkari, 7 F.3d 1130, 1138 (4th Cir.1993). In Procter & Gamble, the court held that a plaintiff seeking to prove false the statement “tests show x” could meet its burden of proof “by showing that the tests referred to ... were not sufficiently reliable to permit one to conclude with a reasonable certainty that they established the proposition for which they were cited.” Id. at 119.

Importantly, however, Procter & Gamble underscored the proposition that a Lanham Act plaintiff is required to prove that an advertisement is false, “not merely that it is ■unsubstantiated by acceptable tests or other proof.” Id. In Procter & Gamble, the parties’ claims failed on these grounds: although the parties pointed to serious flaws in their opponent’s empirical support for their advertisements, the district court denied relief since neither had proven the other’s claims false.⁸ Similarly, in McNeil v. Bristol-Myers, a plaintiff proved false the defendant’s claim that “studies showed” its product superior, not only by pointing out that the studies could not establish this claim, but by showing that these same studies conclusively established that the products were identical. 938 F.2d at 1549-50. ALPO appears to be one of the few exceptions to the generally narrow application of the Procter & Gamble standard. 913 F.2d at 958. Alpo claimed that Ralston had no empirical support for its claim that Ralston Puppy Chow “lessened canine hip dysplasia.” Although Alpo presented no evidence that the claim was false, and apparently did not argue that Ralston’s own evidence proved the claim false, the court nonetheless upheld liability based on Ralston’s insufficient data in support of its claim. Id. at 962.

Upon closer examination, however, these two lines of cases may not conflict but may be explained by the nature of the respective advertisements at issue. Both Procter & Gamble and McNeil v. Bristol-Myers, for example, involved challenges to “establishment claims,” or advertisements in the form “tests show x.” Such a statement can be proven “literally false” by showing that the cited tests do not, in fact, establish the proposition claimed. The same “literal falsity” characterizes a statement such as “4 out of 5 doctors....” In Sandoz and Pennzoil, by contrast, the challenged advertisement made no reference to tests or surveys, but made independent claims (Vicks cough syrup starts to work the instant you swallow), 902 F.2d at 227, and comparison claims (of the form “our product works faster than theirs”), 987 F.2d at 943. In such “non-establishment cases,” proof that the advertiser had no support for its statement would not necessarily prove falsity; thus the Sandoz and Pennzoil courts, addressing claims meeting this description, required affirmative proof of falsity to establish liability.

The Second Circuit has recognized this distinction, effectively limiting the holding in Procter & Gamble to “establishment” claims while requiring affirmative proof for “non-establishment claims.” Castrol Inc. v. Quaker State Corp., 977 F.2d 57, 63 (2d Cir.1992). “A plaintiffs burden in proving literal falsity thus varies depending on the nature of the challenged advertisement. Where the defendant’s advertisement claims that its product is superior, plaintiff must affirmatively prove defendant’s product equal or inferior. Where ... defendant’s ad explicitly or implicitly represents that tests or studies prove its product superior, plaintiff establishes its burden by showing that the tests did not establish the proposition for which they were cited.” Id.

Quaker State’s analysis is also consistent with the only case in our Circuit to consider this issue. In Abbott Laboratories, Abbott sought to enjoin Mead Johnson from making certain claims about “Ricelyte,” an infant rehydrating solution competing with Abbott’s “Pedialyte.” 971 F.2d at 9. Abbott challenged Mead’s claims that Ricelyte was a “rice-based” solution, and that tests showed it was superior to Abbott’s Pedialyte solution in fluid absorption. Id. at 15. Affirming the district court’s findings of liability, we held that the district court had properly concluded that Mead’s claim that Ricelyte was “rice based” was literally false, since it contained no rice. Id. at 14. With respect to Mead’s claim that Ricelyte was superior to Pedialyte in fluid absorption, we affirmed the district court’s application of a different standard, finding liability because Ricelyte’s claim lacked a “firm scientific grounding.” Id. at 15. The district court opinion in Abbott makes clear that Mead’s comparative claims of fluid absorption were explicitly based upon one study that did not show Mead’s superiority; since the district court characterized Mead’s claim as an establishment claim, Abbott’s proof that the study was invalid was sufficient to prove Mead false. Abbott Laboratories v. Mead Johnson & Co., No. IP 91-202C, 1991 U.S.Dist. Lexis 21010 (N.D.Ill. Oct. 10, 1991), Findings of Fact ¶¶ 112-115; Conclusions of Law ¶ 7. Although we did not expressly consider Abbott’s standard of proof with respect to the different claims, Abbott Laboratories is consistent with the Quaker State model.

Quaker State recognizes a fundamental point: since the conditions under which a statement is true or false vary according to the statement at issue, a plaintiffs burden of proof may be satisfied in different ways with respect to different statements. Statements in the form “tests show x” are literally false if tests do not establish “x.” But more general comparative claims can only be proven false by affirmative evidence of falsity. Of course, claims of product superiority are a close cousin to claims that “tests show x,” for “implicit in most assertions of superiority ... is that some type of comparison between products has been made.” Accu-Sort Systems, Inc. v. Lazerdata Corp., 820 F.Supp. 928, 932 (E.D.Pa.1993). Quaker State agreed that implicit establishment claims might give rise to liability upon proof that the underlying data cannot support the stated proposition.

We adopt the logic of Quaker State and hold that a Lanham Act plaintiff bears the burden of proving literal falsity, but that the proof sufficient to meet this burden will vary depending upon the statement made. If the challenged advertisement makes implicit or explicit references to tests, the plaintiff may satisfy its burden by showing that those tests do not prove the proposition; otherwise, the plaintiff must offer affirmative proof that the advertisement is false.⁹

B. Old World’s representations

While most false advertising claims arise from advertisements involving simple comparison or performance claims, the representations at issue here are somewhat more ambiguous. Old World claimed that its antifreeze met the Ford and GM specifications. BASF contends that this meant that Old World had subjected the antifreeze to various protocols and tests and that it has passed. But in Old World’s view, “meeting specifications” only means that, if tested, Old World would pass. The meaning of a given advertisement is a question of fact, see Abbott Laboratories, 971 F.2d at 13, and the district court rejected Old World’s interpretation. “All experts who testified in this case ... stated that Old World could not truthfully represent or claim that its AF met or exceeded the performance specifications ... unless all required tests had been run and passed.” Findings of Fact, ¶ 30. The district court thus concluded that “meeting specifications” means “we have run specific tests and achieved specific results.” Since Old World had not run all of the appropriate tests, the district court concluded that its statement was literally false. Old World has not shown that this finding was clearly erroneous. Id. at 14.

Nonetheless, Old World maintains that BASF’s proof was insufficient to support the district court’s conclusion that its statements were literally false. Incomplete testing cannot be proof of literal falsity, contends Old World. But that would be true only under Old World’s interpretation of “meeting specifications.” The district court found that meeting specifications means that all requisite tests were performed and passed. Since Old World did not perform the tests, its claim was literally false, and the district court did not err in finding Old World liable under the Lanham Act or state law.¹⁰

III. REMEDIES

The Lanham Act allows the district court a wide range of legal and equitable remedies for a violation of the Act.

[T]he plaintiff shall be entitled, ... subject to the principles of equity, to recover (1) defendant’s profits, (2) any damages sustained by the plaintiff, and (3) the costs of the action. The court shall assess such profits and damages or cause same to be assessed under its direction. In assessing profits the plaintiff shall be required to prove defendant’s sales only; defendant must prove all elements of cost or deduction claimed. In assessing damages the court may enter judgment, according to the circumstances of the ease, for any sum above the amount found as actual damages, not exceeding three times such amount. If the court shall find that the amount of recovery based on profits is either inadequate or excessive the court may in its discretion enter judgment for such sum as the court shall find to be just, according to the circumstances of the ease. Such sum in either of the above circumstances shall constitute compensation and not a penalty. The court in exceptional cases may award reasonable attorney fees to the prevailing party.

15 U.S.C. § 1117(a). Remedies are intended to make violations of the Act unprofitable, but not to act as a penalty. Otis Clapp & Son, Inc. v. Filmore Vitamin Co., 754 F.2d 738, 742 (7th Cir.1985). As the Act makes clear, the district court has wide discretion to fashion an appropriate remedy. Accordingly, we review the adequacy of the district court’s remedy for abuse of discretion. Badger Meter, Inc. v. Grinnell Corp., 13 F.3d 1145, 1157-58 (7th Cir.1994); Otis Clapp, 754 F.2d at 744.

A. Lost profits

BASF contends that the district court elearly erred when it awarded BASF only $2,498,726 in lost profits rather than the requested $22,713,000 (excluding adjustments). BASF argues that the district court erred by rejecting BASF’s lost profits determinations and choosing a market share analysis, and that it further erred in calculating BASF’s market share. Old World, on cross-appeal, argues that BASF was entitled to no damages since it had not proven that its loss was caused by Old World’s advertisements. We consider these arguments together.

1. BASF’s loss

Although the district court accepted BASF’s method of calculating its profits per gallon of antifreeze, the district court did not adopt BASF’s calculation of lost sales. In particular, the district court disagreed with BASF’s key contention that if Old World had not made the sales, BASF would have. Old World’s argument is essentially the mirror image of BASF’s position: Old World argues that BASF failed to prove that any customers left BASF because of Old World’s misrepresentations. Instead, Old World contends that all customers were pushed away due to BASF’s high prices. We find that the district court did not err in finding that BASF would have made some, but not all, of Old World’s sales but for Old World’s misrepresentations.

Clearly there was sufficient evidence in the record for the district court to reject BASF’s position that it would have retained all of the accounts that went to Old World. Representatives of Citgo and Phillips, the two largest accounts (comprising about $16 million of BASF’s damages claim), testified that they left BASF due to its high prices. Citgo in particular testified that it would not have bought from BASF at such high prices even if Old World were not in the market. Although the Citgo and Phillips representatives also testified that they needed antifreeze that met specifications — and thus had they known the truth about Old World would probably not have bought from them — both these clients could have gone to Texaco or, earlier in the game, other competitors. Moreover, the district court found that BASF offered no proof with respect to the six smallest customers, whose purchases from Old World to-talled about $1.6 million. The district court’s finding that BASF was at times sensitive to price does not contradict its finding that, on significant occasions, BASF lost customers due to its pricing policy.

But Old World’s challenge to the district court’s findings also fails, since the district court had a reasonable basis to conclude that Old World’s misrepresentations gained it customers. Old World argues that BASF has failed to establish a link between Old World’s misrepresentations and BASF’s lost sales. See Grove Fresh Distributors, Inc. v. New England Apple Products Co., Inc., 969 F.2d 552, 556 (7th Cir.1992); Otis Clapp, 754 F.2d at 745. Rather, Old World contends that all of BASF’s lost sales can be attributed to its inflexible pricing.

Old World relies primarily on the testimony from Citgo and Phillips that they would not have purchased from BASF at all due to its price (Valvoline and TAG complained about BASF’s prices, but were less adamant). But the district court properly considered this testimony in the broader context of the effect of Old World’s misrepresentations on the antifreeze market. BASF had some price flexibility, and in a market where Old World was not making misrepresentations, BASF could have been more competitive. Cf. Bigelow v. RKO Radio Pictures, Inc., 327 U.S. 251, 264, 66 S.Ct. 574, 579-80, 90 L.Ed. 652 (1946).¹¹ There was evidence that many customers would not have gone to Old World had they known the truth; if Old World were thus eliminated from the market, BASF’s prices may have looked eomparatively better, or BASF might have adopted a different pricing strategy. There also could have been more players in the market had Old World not been able to make its misrepresentations. Thus the district court was entitled to discount the Phillips and Citgo testimony to account for the fact that the market would have been quite a different place if Old World were not drawing in customers.

As the district court acknowledged, its analysis is necessarily somewhat more speculative than in an ordinary case where market forces (or independent explanations for a drop in sales) can be analyzed separately from the wrongful conduct. See, e.g., Nikkal Industries, Ltd. v. Salton, Inc., 735 F.Supp. 1227, 1230-32 (S.D.N.Y.1990) (defendant’s lost sales of ice cream makers not necessarily caused by plaintiffs advertising in a market characterized by an increasing number of participants and declining prices); Isaksen v. Vermont Castings, Inc., 825 F.2d 1158, 1165 (7th Cir.1987), cert. denied, 486 U.S. 1005, 108 S.Ct. 1728, 100 L.Ed.2d 193 (1988) (small player had little effect on wood-burning stove market); Schiller & Schmidt, Inc. v. Nordisco Corp., 969 F.2d 410, 415 (7th Cir.1992) (plaintiff attempted to attribute all lost sales to misappropriation of an office supplies catalog, when sales drop was clearly attributable to personnel changes and entry into market of a powerful, legitimate competitor). But the evidence showed that Old World was able to adopt lower prices because of its false advertising; that some private label business left BASF and went to Old World; that these customers would not have gone to Old World had they known the truth; and that alternative options were limited. Given the partieu- ‘ lar nature of Old World’s role in shaping the private label antifreeze market; we cannot conclude that the district court erred in finding that BASF sustained some lost sales due to Old World’s false advertising.

2. Market share analysis

Since the parties’ all-or-nothing theories of damages were inconsistent with the district court’s finding that BASF proved that it would have made some of the sales absent Old World, the district court was required to adopt a different method of calculating damages. The district court’s choice of market share analysis logically follows from its finding that although prices drove some customers away from BASF, ordinary market forces would have pulled some sales back had Old World competed fairly in the market. We cannot conclude that the court erred by determining lost profits according to market shares.

Nevertheless, BASF argues that the district court improperly conducted the analysis. The district court calculated BASF’s lost profits based on its share of the total non-Old World antifreeze market, which included both private and manufacturer-brand labels. BASF contends that the district court should have looked only to the private label market in which BASF and Old World competed. BASF points out that while Union Carbide dominated the total antifreeze market with its branded “Prestone” label, BASF was the biggest player in the private label antifreeze business. In its Rule 59 motion, BASF contended that its share of the private label market was 28% in 1985 rising to 34% in 1988, rather than the 15.6% rising to 21.2% as calculated by the district court. BASF arrived at this figure by deducting Union Carbide’s share and recomputing its own share, reasoning that Union Carbide owed its strong position in the total antifreeze market to a product it did not sell in the private label market. But the district court found that “the evidence disclosed that Union Carbide was a strong player in the private label market and did not exit this portion of the antifreeze market until near the end of the 1987 antifreeze year. Thus during the damage period ... Union Carbide was a strong competitor of BASF in the private label market.”¹² On appeal, BASF does not repeat its claim to 28%-34% of the private label market, but continues to argue that Union Carbide must have held a lesser share of the private label market than did BASF. BASF points to nothing in the record, however, that contradicts the district court’s conclusion that Union Carbide was a significant player or explains why BASF employees noted competition from Union Carbide in the 1987 antifreeze year.

The district court therefore knew that Union Carbide had a share of the private label antifreeze business through 1987, but had no additional information about the composition of the private label market. The district court concluded that the entire antifreeze market offered the best approximation of the private label market, and that any other measure would involve “a great deal of speculation.” Under these circumstances, we cannot conclude that the district court erred given the record before it. BASF failed to introduce sufficient evidence to support its claims about the breakdown of the private label market. That this was BASF’s burden is clear: although “a plaintiff is held to a lower burden of proof in ascertaining the exact amount of damages” as opposed to the existence of damages, the plaintiff still bears some burden of proof. Otis Clapp, 754 F.2d at 745. The rule in Otis Clapp is not to the contrary, holding only that a Lanham Act plaintiffs failure to prove a precise measure of damages cannot be the basis for a finding that it sustained no damages. Otis Clapp, 754 F.2d at 745; accord Broan Mfg. v. Associated Distributors, Inc., 923 F.2d 1232, 1238 (6th Cir.1991). BASF excuses its failure by claiming it was surprised by the district court’s use of market shares to determine a remedy. But market share was clearly relevant to the case — and notwithstanding BASF’s argument that the only relevant market is the private label market — BASF also points to Old World’s share of the total antifreeze market as evidence of Old World’s wrongdoing. Plaintiff-App. Br. at 3, 5, 7. As the district court explained, “the short of the matter is that BASF presented damage opinion evidence that gave the court no alternative short of total victory, to which it was clearly not entitled. The court attempted to fashion as fair an award as possible under the circumstances and the evidence.” Mem. Op. September 8, 1992. This the district court was permitted to do: although damages may not be purely speculative, a factfin-der may “make a just and reasonable estimate of the damage based on relevant data” and may “act upon probable and inferential, as well as direct and positive, proof.” Bigelow v. RKO Pictures, 327 U.S. at 265, 66 S.Ct. at 580; Otis Clapp, 754 F.2d at 745. The district court’s attempt to steer a course by the somewhat obscure light of the record was permissible, and because the record does not convince us that the district court was mistaken, the district court’s findings are not clearly erroneous.

B. Equitable remedies

1. Disgorgement

BASF argues that the district court erred in denying disgorgement of Old World’s wrongful profits. The district court, noting that disgorgement is an equitable remedy, reasoned that disgorgement is most appropriate when damages are nominal and the defendant would not otherwise be deterred. The court found the $2.5 million damages awarded here to be a sufficient deterrent.

BASF makes the threshold argument that the district court clearly erred by failing to calculate the amount of Old World’s wrongful profits. BASF estimated Old World’s wrongful profits to be $17.6 million from 1983-87, and $25.6 million if calculated to 1988. At oral argument, BASF stated that these figures represented revenues minus direct costs, but with no deduction for indirect costs.¹³ However, the burden lay with Old World to produce such cost figures. 15 U.S.C. § 1117(a) (“In assessing profits, the plaintiff shall be required to prove defendant’s sales only; defendant must prove all elements of cost or deduction claimed.”) Old World, by contrast, claims net profits of only $600,000 from 1985-87, the years included in the damages formula. This figure, however, does not relate solely to Old World’s antifreeze operations but includes all of Old World’s operations, including several less profitable divisions. Old World also contends that BASF’s calculations include profits derived from Old World’s sales of antifreeze bearing its own labels (Full Force and Advance) rather than merely sales of private label antifreeze.

The district court made no specific finding as to the amount of Old World’s wrongful profits. BASF argues that without a finding, the district court could not determine whether disgorgement of the wrongful profits was appropriate. Moreover, BASF contends that Old World acquired wrongful profits over a longer time period (1982-88) than that measured by the lost sales calculations (1984-87). But we do not agree that the district court’s failure to adopt a specific profits figure constitutes clear error. The district court here made detailed findings of facts regarding sales and profits, albeit only for the years after 1984. Of course, lost profits of the plaintiff are quite a different matter than the defendant’s profits, Web Printing Controls Co., Inc. v. Oxy-Dry Corp., 906 F.2d 1202, 1205 (7th Cir.1990), and thus need to be calculated independently; but given the relatively few customers in the market and fairly detailed sales information, the district court possessed a sufficient understanding of the amount of profits involved to determine whether disgorgement would be appropriate. Cf. Louis Vuitton S.A. v. K-Econo Merchandise, 813 F.2d 133, 134-35 (7th Cir.1987) (remand appropriate when “no facts were unearthed, no law propounded, and no reasons for the ruling explained”).

We also do not conclude that the district court abused its discretion by refusing to award Old World’s profits. Disgorgement initially developed as a remedy to pro-vide a plaintiff relief in equity, to serve as a proxy for damages, or to deter the wrongdoer from continuing his violations. George Basch Co. v. Blue Coral, Inc., 968 F.2d 1532, 1537-40 (2d Cir.), cert. denied, — U.S. —, 113 S.Ct. 510, 121 L.Ed.2d 445 (1992) (thorough discussion of purposes of the wrongful profits award); Texas Pig Stands, Inc. v. Hard Rock Cafe Int’l, 951 F.2d 684, 699 (5th Cir.1992) (Johnson, J., dissenting). The variety of circumstances in which a court may award disgorgement by no means indicate that the district court is required to award disgorgement. See Champion Plug Co. v. Sanders, 331 U.S. 125, 131, 67 S.Ct. 1136, 1139, 91 L.Ed. 1386 (1947) (disgorgement not required upon proof of violation); Texas Pig Stands, 951 F.2d at 695-96, 699. Contrary to BASF’s assertions that disgorgement of wrongful profits is the “norm” in a Lanham Act case, disgorgement is most appropriate if damages are otherwise nominal, Playboy Enters., Inc. v. Baccarat Clothing Co., 692 F.2d 1272, 1274 (9th Cir.1982); accord Louis Vuitton v. Lee, 875 F.2d 584, 588-89 (7th Cir. 1989) (either damages or disgorgement is appropriate). “While damages directly measure the plaintiffs loss, defendant’s profits measure the defendant’s gain. Thus, an accounting may overcompensate for a plaintiffs actual injury and create a windfall judgment at the plaintiffs expense.” George Basch Co., 968 F.2d at 1540. However, the district court may consider the need to deter further violations to protect the public at large. Id.

Damages in this case were significant— $4.2 million including prejudgment interest— albeit far less than what BASF requested. But the district court heard evidence on Old World’s profits, and concluded that the judgment would be sufficient to render Old World’s violations unprofitable. Equity did not require the district court to award disgorgement, particularly after it awarded damages equivalent to its calculation of BASF’s losses. Indeed, disgorgement might have required the district court to engage in undue speculation as to the amount. See, e.g., Sands, Taylor & Wood v. Quaker Oats Co., 978. F.2d 947 (7th Cir.1992), cert. denied, — U.S. —, 113 S.Ct. 1879, 123 L.Ed.2d 497 (1993). We therefore cannot conclude that the district court abused its discretion by failing to require disgorgement of Old World’s profits.

2. Enhancement

BASF argues that the district court erred in refusing to enhance the damages award pursuant to the Lanham Act’s provision allowing a district court to triple the amount of damages. 15 U.S.C. § 1117(a). Enhancement of an award of damages is clearly committed to the district ' court. Badger Meter, 13 F.3d at 1157; Otis Clapp, 754 F.2d at 744. BASF relies solely on ALPO Petfoods v. Ralston Purina Co., 778 F.Supp. 555 (D.D.C.1991), affd in part and remanded, 997 F.2d 949 (D.C.Cir.1993) {“ALPO II ”) to argue that the district court abused its discretion by refusing to enhance an award. ALPO II listed factors a court should consider in deciding whether enhancement is appropriate: “the impact of the unlawful conduct directly on the plaintiff competitor, the systemic distortion which the wrongdoer’s conduct has upon the particular product market ... and the costs incurred by the competitor in its effort to mitigate the damages.” Id. at 564. In ALPO II, however, the court awarded neither attorneys fees nor lost profits, finding the latter too difficult to ascertain, so relied on the flexible remedy of the enhancement provision to fashion a just award. Id. The present case is similar to ALPO II only with respect to the second factor, market distortion. Although Old World tries to argue that it did nothing to distort the market, its position is identical to the defendant’s in ALPO II, in which the fraudulent advertising concerned a quality that would determine a buyer’s preference in that market and in fact did cause a market shift. 778 F.Supp. at 565. However, this one equitable factor in BASF’s favor does not require a conclusion that the district court’s decision was an abuse of discretion. And in any event, the Act specifically provides that enhancement is only available to ensure that the plaintiff receive compensation, Badger Meter, 13 F.3d at 1157, and the district court found that BASF was adequately compensate ed by the damage award. “A discretionary award is just that — discretionary. The judge made no error of law in his analysis, and his refusal to enter such an award has not been shown to be an abuse of discretion.” Badger Meter, 13 F.3d at 1158.¹⁴

C. Damages for 1988

The district court found that Old World’s false advertising ceased on March 31, 1987, and thus did not include any damages attributable to the 1988 antifreeze year (April 1,1987-March 31,1988) in its damages award. Old World closed on its purchase of Northern Petro’s “Peak” formula on December 31,1986, and the district court found that it “terminated its business relationship with Dearborn on March 31, 1987.” Old World purchased no more formula from Dearborn after that date, but commenced using the Peak formula. Accordingly, the district court found that Old World’s misrepresentations ceased at the end of the 1987 antifreeze year. BASF argues that the district court’s findings were clearly erroneous, and that BASF sustained damages into the 1988 antifreeze year.

Upon consideration of BASF’s Rule 59 motion arguing that the district court should have considered the 1988 losses, the district court explained more thoroughly its basis for concluding that damages ceased in 1987. The district court relied upon:

(1) Testimony of George Beck, Dearborn’s salesman in charge of the Old World account, who testified that Dearborn lost the Old World account for the 1988 season, when Old World went exclusively to Peak;

(2) Testimony of Richard Tumm, Dear-born’s director of sales, in a similar vein;

(3) Testimony of John Hurvis, Old World’s chairman, that the relationship with Dear-born ended on that date;

(4) The absence of any direct evidence of sales of the Dearborn formula to customers in 1988;

(5)Evidence that Old World’s customers were aware of the BASF suit and the charge that Old World did not meet specifications, particularly a memo to Larry Birch of Citgo advising him of the lawsuit.

BASF contends that these findings were erroneous. BASF first argues that the testimony of Dearborn and Old World officials of the duration of their relationship did not support the district court’s conclusion. We disagree. George Beck, the Dearborn sales representative, testified that Dearborn last supplied Old World with an inhibitor package in the 1987 antifreeze season (which Beck understood ran from April 1 to March 31), Tr. 1231, but that Dearborn lost the account for the 1988 season. Tr. 1225. Ordinarily, there would have been purchase orders coming in at the beginning of the season, but Beck received none for the 1988 season. He testified that thereafter Old World switched “100%” to the Peak formula. Tr. 1221. Although Beck recalled one meeting between Old World and Dearborn at the beginning of April 1987, he recalled no further meetings. Tr. 1222-23. Richard Tumm provided similar testimony. Tumm testified several times that the Dearborn-Old World relationship ended in late March or early April 1987, Tr. 444, 497, although Dearborn attempted to regain Old World’s business into May 1987. Tr. 501. Nonetheless, the relationship did not continue into the 1987-88 season, when Old World switched to Peak. Tr. 458. John Hurvis, Old World’s chairman, also testified that Old World terminated its relationship with Dearborn in the first quarter of 1987. Tr. 633.

BASF argues that testimony about the end of the relationship does not firmly establish the lack of sales in 1988. But the district court was entitled to make reasonable inferences from the record, and absent other proof of sales, the testimony reasonably supports the conclusion that there were no further sales.¹⁵ This interpretation is strengthened when the transcript is read as a whole rather than in isolated fragments. The witnesses consistently testified that Old World switched from Dearborn to the Peak formula at the beginning of the 1988 antifreeze season. Moreover, any perceived lack of specificity in the witnesses testimony about the end of the relationship (i.e., whether this meant that no blending or sales continued), is a result of the general nature of the questions asked, which did not require the witnesses to describe in detail what it meant for the parties to terminate the relationship. The district court did not err in relying in part on this evidence to conclude that there were no more sales after the relationship had ended.

BASF next attacks the district court’s reliance on the absence of any records showing sales of Dearborn formula after March 31, 1987. BASF contends that Old World destroyed the blending and sales records that would have proven 1988 sales, and that the district court erred by not presuming that these records would have proven BASF’s point.¹⁶ But although the district court may have been entitled to draw adverse inferences from Old World’s failure to produce documents, see, e.g., NLRB v. Dorothy Shamrock Coal Co., 833 F.2d 1263, 1268 (7th Cir.1987) (factfinder allowed to draw adverse inference); Brown & Williamson Tobacco Corp. v. Jacobson, 827 F.2d 1119, 1134 (7th Cir.1987) (Juries may presume actual malice from bad faith destruction of documents), “whether to draw the inference is a matter of discretion for the fact finder.” International Union, UAW v. NLRB, 459 F.2d 1329, 1339 (D.C.Cir.1972) (detailed discussion of adverse inference rule). The inference, “more a product of common sense than of the common law,” id. at 1335, is no different than the numerous other inferences the fact finders must always weigh; we cannot conclude that, when faced with conflicting evidence, the district court erred by failing to hinge liability on an inference drawn from missing documents. We are particularly reluctant to do so given that the district court made no finding that Old World willfully destroyed the 1988 sales records.

BASF also argues that the district court failed to consider notes from blenders indicating that some blending of the Dearborn formula occurred into July, 1987. BASF claims that the evidence — in the form of copies of notes from blenders indicating that they stopped blending Dearborn formula in June or July 1987 — was introduced into evidence as exhibits to the deposition of Jeff Grizzle (a blender). The documents do seem to indicate that some blenders mixed the Dearborn formula into July 1987. But the district court found that the blending records mentioned in the Grizzle deposition were not made part of the record during the trial, but first appeared as appendices to BASF’s Rule 59 motion and, in any event, did not tell to whom the antifreeze was sold. The district court evaluated the evidence with the Rule 59 motion, and concluded that it still did not prove that Old World continued to sell falsely advertised antifreeze during the 1988 season. The records contain no' indication whether the antifreeze was sold, let alone on the private label market. We cannot conclude that the district court was required to find that the blenders’ notes proved that BASF continued to lose customers as a result of Old World’s false advertising in the 1988 antifreeze season.¹⁷

D. Attorneys’ fees

The Lanham Act allows for fees in “exceptional” eases, 15 U.S.C. § 1117(a), which encompasses cases in which the acts of infringement are “malicious, fraudulent, deliberate or willful.” Roulo v. Russ Berrie & Co., Inc., 886 F.2d 931, 942 (7th Cir.1989), cert. denied, 493 U.S. 1075, 110 S.Ct. 1124, 107 L.Ed.2d 1030 (1990). On cross-appeal, Old World argues that the district court erred when if found that BASF was entitled to attorneys’ fees due to Old World’s deliberately false advertising. BASF argues that Old World’s challenge to the district court’s award of fees is premature, since the district court has not yet determined a final amount. Generally, BASF would be right, since a fee award that does not specify an amount or a means of determining the amount is not a final decision under 28 U.S.C. § 1291. John v. Barron, 897 F.2d 1387, 1390 (7th Cir.), cert. denied, 498 U.S. 821, 111 S.Ct. 69, 112 L.Ed.2d 43 (1990); Lac Courte Oreilles Chippewa Indians v. Wisconsin, 829 F.2d 601, 603 (7th Cir.1987). However, we have recognized an exception where the order awarding fees in an amount not yet determined can be consolidated on appeal with a final order. Bittner v. Sadoff & Rudoy Indus., 728 F.2d 820, 826 (7th Cir.1984); cf. Budinich v. Becton Dickinson & Co., 486 U.S. 196, 108 S.Ct. 1717, 100 L.Ed.2d 178 (1988) (consequences for timing of notice of appeal). In Bittner, the district court held that a party was entitled to fees, but did not determine the amount. Nonetheless, we held the order reviewable (although it is unclear whether on the basis of some form of pendent jurisdiction or by a determination that the fee award was final under § 1291), since “there would be no net judicial economy, but if anything a net diseconomy, if we held that we could not consider the merits of a fee order until the amount of fees to be awarded is fixed.” Bittner, 728 F.2d at 827; but cf. Cooper v. Salomon Bros., Inc., 1 F.3d 82, 85 (2d Cir.1993), cert. denied, — U.S. —, 114 S.Ct. 737, 126 L.Ed.2d 700 (1994); Becton Dickinson & Co. v. Dist. 65, UAW, 799 F.2d 57, 61 (3d Cir. 1986); Southern Travel Club, Inc. v. Carnival Air Lines, Inc., 986 F.2d 125, 129-30 (5th Cir.1993); Gates v. Central States Teamsters Pension Fund, 788 F.2d 1341, 1343 (8th Cir. 1986); Jensen Electric Co. v. Moore, Caldwell, Rowland & Dodd, Inc., 873 F.2d 1327, 1329 (9th Cir.1989). While concerns for “judicial economy” might suggest a different approach in the present case, where the amount of fees is still hotly contested in the district court, jurisdiction cannot turn on the eventual determination of an amount. It is well settled that we have jurisdiction to review the district court’s finding that BASF was entitled to fees. See Lac Courte, 829 F.2d at 603; U.S. Marine Corp. v. Szabo, 819 F.2d 714, 717 (7th Cir.1987); Vandenplas v. Muskego, 797 F.2d 425, 429 n. 1 (7th Cir. 1986); see also Exchange Nat. Bank of Chicago v. Daniels, 763 F.2d 286, 291-92 (7th Cir.1985).

A decision to award attorneys’ fees under the Lanham Act is firmly committed to the district court’s discretion, Otis Clapp, 754 F.2d at 747, and we do not find that the district court abused its discretion here. Old World, relying on ALPO, 913 F.2d at 971, argues that a plaintiff must show “willfulness or bad faith” to be entitled to a fee award, and that the defendant’s conduct “targeted” the plaintiff. But the law of this Circuit clearly allows the district court to award attorneys’ fees based upon the defendants’ “malicious, fraudulent, deliberate or willful” infringement. Otis Clapp, 754 F.2d at 746; Roulo v. Russ Berrie, 886 F.2d at 942. The district court concluded that although Old World’s conduct was not malicious, it was deliberate. See, e.g., Badger Meter, 13 F.3d at 1158-60. We will not disturb the district court’s finding that BASF is entitled to “some” attorneys’ fees.

IV. CONCLUSION

The decisions of the district court are

Affirmed.

1 . Old World advertised that it met the following standards: Cummins 85T8-2 1985; MCAT RP 302A; SAE 11J1034; ASTM D-3306; GM 1825-M (1982); GM 1899-M; Ford ESE-M97B44-A.

2 . Old World introduced a letter from Dearborn to Old World dated June 15, 1984 that read in part, "This is to certify that Full Force [Old World’s name for the Dearborn formula] meets specifications for performance testing by GM and Ford Motor Company.” The district court made no findings in connection with this letter, but found that Dearborn did not certify any test results to Old World.

3 . Antifreeze years run from April 1 of the previous year to March 31 of the year in question. For example, antifreeze year 1988 runs from April 1, 1987 until March 31, 1988.

4 . The respective market shares for the entire antifreeze market from 1984-1988 were:

1988 1984

32% Union Carbide 38%

Dow 10%

Shell 5% O

Texaco 12% ^

BASF 15% CO

Northern Petro 6%

Conoco 7%

Old World 4%

Others 3%

5 . BASF carries as its variable cost for ethylene and EG (antifreeze ingredients) its variable cost of its own production of each rather than the market value of each. This cost method of accounting does not include the opportunity costs associated with forgoing sales of EG on the open market. During the relevant time period, the market value of EG was greater than BASF's cost to produce it. Accordingly, BASF's cost of production was less than market value. Old World's damages expert testified that BASF’s method of accounting was contrary to the chemical industry practice for integrated producers, and that under a transfer pricing system, under which EG would be costed at its market value, BASF would have been able to make no contributions to fixed costs during any of the years in question. But the district court rejected Old World's analysis and adopted BASF's calculations of the contribution margin.

6 . The district court calculated profits as follows:

LOST SALES 1985 1986 1987

Package Package Bulk Package Bulk

Ashland 0 9,576 265,471 10,919 250,232

Valvoline 0 0 5,017 638,849 0

Texstar 0 236,148 0 300,000 0

Phillips 625,344 2,918,399 504.534 4,213,978 381,022

Citgo 540,072 1,824,918 376,769 2,931,135 787,540

Southland 0 0 0 48,078 0

Chem Cent 0 8,580 45.534 10,554 79,535

Comm Dist 0 203,578 0 222,973 0

Felts 0 63,320 0 63,467 0

Denver 0 880 0 4,931 0

Total Gallons 1,165,416 5,265,399 1,197,325 8,445,844* 1,498,329

Contrib/Gallon $.79 $.76 $.89 $.71 $.75

Lost Profits $920,679 $4,001,703 $1,065,619 $5,996,577 $1,123,747

BASF Share non-OW market 17.0% 18.4% 18.4% 19.8% 19.8%

Lost Profits Awarded $156,515 $736,313 $196,074 $1,187,322 $222,502

TOTAL AWARD: $2,498,726

The district court found there were no bulk sales in 1985. The damages award also reflected a calculation error, left uncorrected here. The district court found that the total gallons lost for package sales in 1987 came to 8,445,884 rather than 8,444,884. Had the district court added correctly, the actual amount of lost profits would be $140 less than the district court's award. However, we will not disturb the district court's award despite this error in calculation, since the district court relied on its award in determining whether other remedies were appropriate.

7 . When the conduct at issue in this case took place, § 43(a) read in part:

Any person who shall affix, apply, or annex, or use in connection with any goods or services, or any container or containers for goods ... any false description or representation, including words or other symbols tending to falsely describe or represent the same, and shall cause such goods or services to enter into commerce ... shall be liable to a civil action ... by any person who believes that he is or is likely to be damaged by the use of any such false description or representation.

15 U.S.C. § 1125(a) (1982). The Trademark Law Revision Act of 1988, Pub.L. No. 100-667, § 132, 102 Stat. 3935 (1988), amended § 43(a) to make more explicit its focus on commercial advertising. The amendments primarily codified pre-1988 case law and did not work any substantive changes relevant to this case. See S.Rep. No. 515, 100th Cong., 2d Sess. 40, reprinted in 1988 U.S.C.C.A.N. 5577, 5603; Mylan Laboratories v. Matkari, 7 F.3d 1130, 1138 n. 11 (4th Cir.1993), cert. denied, -U.S. -, 114 S.Ct. 1307, 127 L.Ed.2d 658 (1994).

8 . Procter & Gamble’s advertisements claimed that "clinical tests” proved that its New Wondra hand lotion was superior to leading brands; Che-sebrough made the lesser, but inconsistent, claim of parity — that "no leading lotion beats Vaseline Intensive Care.” 747 F.2d at 116. The parties' brought cross-motions seeking injunctions. The district court denied both parties’ motions, finding that neither had established that the other’s advertisements were false. The Third Circuit affirmed.

9 . We reiterate that we are only concerned here with the first "prong” of the standard for Lan-ham Act liability. If a statement is literally true or ambiguous, a plaintiff may stUl recover upon proof of customer confusion. Abbott Labs, 971 F.2d at 13.

10 . To recover under the Illinois Consumer Fraud and Deceptive Business Practices Act, 815 ILCS 505/10a, a plaintiff must prove a deceptive act or practice and reliance on the deceptive act. Crowder v. Bob Oberling Enterprises, Inc., 148 Ill.App.3d 313, 101 Ill.Dec. 748, 499 N.E.2d 115 (1986). The district court found that Old World's claims were false and that consumers had purchased antifreeze at least in part as a result of these claims. Old World argues that the Illinois Act should be interpreted like the FTC Act and FTC precedent, which, Old World claims, require proof that an advertisement is "fraudulent or dishonest.” FTC v. Kitco of Nevada, Inc., 612 F.Supp. 1282, 1292 (D.Minn.1985). We do not need to consider Old World's argument that the two Acts are interpreted alike, since the district court's finding of literal falsity makes clear that Old World would lose under either standard.

11 . Describing Eastman Kodak Co. v. Southern Photo Co., 273 U.S. 359, 47 S.Ct. 400, 71 L.Ed. 684 (1927), and Story Parchment Co. v. Paterson Parchment Paper Co., 282 U.S. 555, 51 S.Ct. 248, 75 L.Ed. 544 (1931), the Court noted: "In each case we held that the evidence sustained verdicts for the plaintiffs, and that in the absence of more precise proof, the jury could conclude as a matter of just and reasonable inference from the proof of defendants' wrongful acts and their tendency to injure plaintiffs’ business, and from the evidence of the decline in prices, profits and values, not shown to be attributable to other causes, that defendants' wrongful acts had caused damage to the plaintiffs.... The tortious acts had in each case precluded ascertainment of the amount of damages more precisely, by comparison of profits, prices and values as affected by the conspiracy, with what they would have been in its absence under freely competitive conditions." 327 U.S. at 264, 66 S.Ct. at 579.

12 . The court based this finding on evidence drawn from BASF’s 807 forms. As late as the 1987 antifreeze year, BASF salespersons were still reporting on their 807 forms that they needed a price reduction of private label antifreeze to compete with Union Carbide. See Def.Ex. D.

13 . Based on these numbers, the antifreeze business would appear to be remarkably profitable for a commodity sold on a private label basis. The parties apparently agree that Old World's revenues were $38.9 million in 1986 and $49.1 million in 1987. BASF estimates that Old World’s profits for these years were $4.2 million and $5.8 million respectively, indicating profit margins well over 10%.

14 . BASF also argues that the district court erred when it denied BASF's request for punitive damages for violations of the Illinois Consumer Fraud and Deceptive Practices Act, but we disagree. The district court noted that punitive damages are appropriate only where conduct is "wilful, intentional and egregious," Crowder v. Bob Oberling Enterprises, Inc., 148 Ill.App.3d 313, 101 Ill.Dec. 748, 752, 499 N.E.2d 115, 119 (1986), and did not find "either the degree of egregiousness nor the need to deter sufficient” to warrant punitive damages. The court's finding that Old World had engaged in deliberate conduct does not automatically require punitive damages. Nor does it necessarily conflict with the district court's conclusion that BASF was entitled to attorneys' fees, since the Illinois courts will offset awards of fees against punitive damages. Ciampi v. Ogden Chrysler Plymouth, 262 Ill.App.3d 94, 199 Ill.Dec. 609, 634 N.E.2d 448 (1994). Moreover, the court had already made a finding that its damages award was sufficient to deter future violations.

15 .The district court also considered testimony from Larry Birch of Citgo interpreting a file memorandum he had written on August 3, 1987. The memo contained a notation "Last week in July Citgo inventory had 90,000 of Dearborn Chemical,” indicating that Old World had produced inventory for Citgo. Tr. 2598-99. But since there was no evidence that Citgo ever bought or took possession of this inventory, and the memo stated that Birch knew of the suit against Old World, the district court did not find that the Birch memorandum offered proof of 1987 sales.

16 . On September 9, 1992, the district court sanctioned Old World for turning over documents that were not authentic, but made no finding as to destruction of evidence.

17 . BASF points to other parts of the record in support of its claim, but we do not find that these change our determination. In particular, BASF claims that Ron Confer of the Queen City Terminal (a blending facility) testified that the Dear-born formula continued to be blended up until December 1987. But Confer testified that at least as of June 1987, the only Dearborn formula that continued to be blended was a special blend made exclusively for Chem Central and was not for use in the private label market. Confer Dep. at 143-44, 152-53, 176, 183-84.