This copyright action involves the infringement of a study of the robotics industry. The principal issue on appeal concerns the appropriate measure of damages where a late-registering plaintiff fails to prove out-of-pocket losses to itself or conventional profits to the infringer.

Appellee Business Trends Analysts, Inc. ("BTA"), the successor to Predicasts, Inc. ("Predicasts"), claims that Predicasts' study of the robotics industry was infringed by a publication of appellants The Freedonia Group, Inc., now dissolved, and The Freedonia Group, Incorporated (collectively "TFG"). Judge Conboy held that TFG had infringed BTA's copyright in the Predicasts study. Because BTA registered its copyright only after the infringement, it was limited to a damage award for its losses or TFG's profits and was not entitled to statutory damages. Judge Conboy's damage award included TFG's actual profits and also included an amount for lost market share or, alternatively, the "value of use." We affirm the award of actual profits, but we vacate the award for lost market share or value of use as not authorized by the Copyright Act of 1976, 17 U.S.C. Secs. 101-914 (1982).

Until 1984, Predicasts was in the business of publishing economic and marketing research reports on various industries. A pioneer in the business, Predicasts had an established reputation and customer list. Late in 1984, Predicasts began disbanding its Research Group, the division responsible for industry studies. As of January 1, 1985, Predicasts ceased publishing such studies, transferring its inventory and licensing the use of its name to BTA.

Earlier, in April 1982, William Weiss, then president of Predicasts' Research Group, had left Predicasts. After the expiration of a non-competition agreement in 1983, Weiss had formed The Freedonia Group, Inc., now dissolved. The Freedonia Group, Incorporated was formed in early 1985, when two other former Predicasts employees, T. Kevin Swift and Robert Baumgartner, joined Weiss. In May 1985, TFG hired another Predicasts employee, Andrea Fetsko-Louie.

Throughout 1985 and early 1986, TFG and BTA competed in the industry studies market, each trying to capitalize on its Predicasts connections. BTA thus marketed studies under the Predicasts name, while TFG publicized the connections of its personnel to Predicasts. Throughout 1985, BTA marketed, at a price of $1,500, a Predicasts study entitled 3547 Robotics (Markets and Competitors), produced by Neil Digeronimo in December of 1984. Meanwhile, Swift and Fetsko-Louie composed a similar robotics study for TFG, entitled Industry Study 113: Robotics. After TFG began marketing its study in August 1985, BTA applied for federal copyright registration of the Predicasts study and received a certificate of registration.

Initially, the TFG study was priced at $1,500, the same price charged for the Predicasts study. In January 1986, confronted with negligible sales, TFG offered the TFG study, along with two other studies, for $150 each during a special two- or three-month offer period. The purpose of this price-cutting was to expand TFG's customer base by exposing as many prospective customers as possible to TFG's work. The $150 price was a nominal one designed only to ensure that purchasers had a serious interest in industry studies and were thus potential customers. TFG eventually sold 37 of its robotics studies at or near the discount price.1

On May 5, 1986, BTA filed this action in the Southern District, alleging copyright infringement, misappropriation of Predicasts' customer list and other proprietary information, and violations of the Lanham Act, 15 U.S.C. Sec. 1125(a), as well as violations of New York's Anti-Dilution Statute, Sec. 368-d of the New York General Business Law. BTA sought a preliminary injunction and damages.

On January 5, 1987, Judge Weinfeld denied BTA's motion for a preliminary injunction. See Business Trends Analysts v. Freedonia Group, Inc., 650 F. Supp. 1452 (S.D.N.Y.1987). After Judge Weinfeld's death, a bench trial was conducted before Judge Conboy. In addition to presenting evidence of infringement, BTA called witnesses who testified that BTA had suffered sales losses as a result of TFG's study.

In ruling in BTA's favor on the copyright infringement issue, Judge Conboy reviewed in detail the similarities between various passages in the Predicasts study and corresponding passages in the TFG study. See Business Trends Analysts v. Freedonia Group, Inc., 700 F. Supp. 1213 (S.D.N.Y.1988). He also noted that a copy of the Predicasts study found in TFG's files contained handwritten notations by Fetsko-Louie that corrected erroneous or outdated information and that stated, inter alia, "delete" and "start here". This document was understandably described by Judge Conboy as a "smoking gun" demonstrating direct copying of the Predicasts study by TFG. Judge Conboy found that there was substantial similarity in the respective discussions of "the phenomenon, characteristics, utility and potential of robots, and more to the point, of particular kinds of robots." Id. at 1218-19. He also found, however, that there was no substantial similarity between the two studies in their respective discussions of "highly generalized aspects of economic conditions, historical development and trends in industrial automation, and the role of government in the industrial arena." Id. at 1218. Because the TFG study included elements of substantial originality, he held against BTA on its Lanham Act, state law unfair competition, and misappropriation of trade secrets claims.

Because TFG published its study before BTA's copyright was registered, Section 412(2) of the Copyright Act barred recovery of statutory damages as provided in Section 504(c). BTA was thus limited to actual damages or TFG's profits calculated under Section 504(b). Judge Conboy calculated those damages to be $54,028.35. He found that BTA had failed to establish any actual damages from lost sales of the Predicasts study but determined that TFG's gross profits on the sale of its robotics study were $9,745.00. From this, he deducted TFG's proven expenses in the amount of $5,666.65. He then went on to consider the "novel" question of whether a competitor like TFG, which had engaged in drastic price-cutting in order to gain market advantage "can inferentially be held to have made non-cash profit which, if quantifiable, can be awarded to the infringed competitor." Id. at 1237. He answered that question affirmatively and held that TFG had gained market advantage or "value of use," id. at 1238, in the amount of $1,500, the full list price, less $150 for each copy sold. Thus, TFG's market advantage or value of use per copy was $1,350. Multiplying $1,350 by 37, the number of copies he found had been sold at the discount price, he arrived at a damages figure of $49,950 in market advantage or value of use. Adding gross profits ($9,745) to lost market advantage ($49,950) and subtracting TFG's expenses ($5,666.65), he arrived at a total damage award of $54,028.35. Finally, Judge Conboy declined to apportion TFG's profits between the infringing and noninfringing segments of the TFG study.

Only the copyright issues are before us on this appeal, as BTA has not cross-appealed on its Lanham Act, state law unfair competition, and misappropriation of trade secrets claims. We agree that TFG's study infringed the Predicasts study. We also agree with the award of TFG's actual profits less expenses. However, we reverse the award of damages for lost market advantage or "value of use."

Copyright plaintiffs can prove infringement by demonstrating the defendants' access to the copyrighted work and a substantial similarity between the two works. See Reyher v. Children's Television Workshop, 533 F.2d 87, 90 (2d Cir.), cert. denied, 429 U.S. 980, 97 S. Ct. 492, 50 L. Ed. 2d 588 (1976); 2 M. Nimmer, Nimmer on Copyright Sec. 141.2 at 613 (1988). Access is conceded in the instant case. Whether there has been infringement thus depends on whether the two studies exhibit substantial similarity.

In reviewing a determination of substantial similarity in copyright appeals, we apply a de novo standard of review. See Malden Mills, Inc. v. Regency Mills, Inc., 626 F.2d 1112, 1113 (2d Cir.1980). Such de novo review is appropriate because "[w]here ... the determination of similarity rests solely on a comparison of the works in issue rather than on credibility of witnesses or other evidence only for the factfinder, 'we are in as good a position as the trial judge' to determine the issue." Reyher, 533 F.2d at 90 (quoting Millworth Converting Corp. v. Slifka, 276 F.2d 443, 446 (2d cir.1960)).

Our comparison of the two texts confirms the finding of infringement. Indeed, the evidence of direct copying is unmistakable. Judge Conboy's opinion includes a lengthy analysis, with which we agree, of a large number of "accused passages." 700 F. Supp. at 1221-29. For illustrative purposes, we reproduce here only the passage designated as representative in the briefs of both parties. The relevant paragraph in the Predicasts study reads:

Modern CAD uses analytical software to conduct the analyses and utilizes its graphic capabilities to represent the results pictorially rather than numerically. Recent innovations and continuing developments which facilitate the design process are color, 3-dimensional views, hidden line removal, solid modeling, and texture representation.

The corresponding passage in the TFG study reads:

Modern CAD technology employs analytical software to conduct the design analysis (iterations of formulae, modeling, simulation, etc.) and utilizes its inherent capabilities to present the results graphically. Recent innovations in the latter include color, 3-dimensional views, hidden line removal, solid modeling, and texture representation.

We view the similarities as inexplicable other than as the result of copying and agree with Judge Conboy that the differences amount only to a "crude effort to give the appearance of dissimilarity" and are thus themselves evidence of copying. 700 F. Supp. at 1230 (quoting Meier Co. v. Albany Novelty Mfg. Co., 236 F.2d 144, 146 (2d Cir.1956)).

Of course, we recognize that, evidence of intentional copying notwithstanding, factual information is not protected by the copyright laws. Hoehling v. Universal City Studios, Inc., 618 F.2d 972, 974 (2d Cir.), cert. denied, 449 U.S. 841, 101 S. Ct. 121, 66 L. Ed. 2d 49 (1980); Rosemont Enters., Inc. v. Random House, Inc., 366 F.2d 303, 309 (2d Cir.1966), cert. denied, 385 U.S. 1009, 87 S. Ct. 714, 17 L. Ed. 2d 546 (1967). Nevertheless, the Predicasts study is a "compilation" afforded protection by the Act. 17 U.S.C. Sec. 101 ("A 'compilation' is a work formed by the collection and assembling of preexisting materials or of data that are selected, coordinated, or arranged in such a way that the resulting work as a whole constitutes an original work of authorship."); cf. Eckes v. Card Prices Update, 736 F.2d 859, 863 (2d Cir.1984); Wainwright Sec. Inc. v. Wall Street Transcript Corp., 558 F.2d 91, 95-96 (2d Cir.1977), cert. denied, 434 U.S. 1014, 98 S. Ct. 730, 54 L. Ed. 2d 759 (1978). It was not only Predicasts' specific language, but also Predicasts' "select[ion], coordinat[ion and] arrange[ment]" that TFG appropriated. We therefore affirm the finding of infringement.

We turn, accordingly, to the damages issues. Section 504 of the Copyright Act of 1976 provides two methods for determining damage awards. First, a plaintiff may prove actual damages and profits under Section 504(b). That provision reads in pertinent part:

The copyright owner is entitled to recover the actual damages suffered by him or her as a result of the infringement, and any profits of the infringer that are attributable to the infringement and are not taken into account in computing the actual damages. In establishing the infringer's profits, the copyright owner is required to present proof only of the infringer's gross revenue....

17 U.S.C. Sec. 504(b). Second, copyright plaintiffs may elect at any time before final judgment to receive statutory damages under Section 504(c), a method useful where proof of actual damages or profits is insufficient. Statutory damages are not available, however, for "any infringement of copyright commenced after first publication of the work and before the effective date of its registration, unless such registration is made within three months after the first publication of the work." 17 U.S.C. Sec. 412(2). Because BTA did not register the Predicasts study until after the infringement, BTA is not eligible for statutory damages and is limited to actual damages and profits under Section 504(b).

We need not tarry over Judge Conboy's award of TFG's actual profits. We also affirm as not clearly erroneous his findings that BTA failed to prove any lost profits on its part. Although BTA offered opinion testimony indicating that it had suffered sales losses, Judge Conboy was free not to credit this evidence on the grounds that it lacked any foundation in past sales of BTA or Predicasts studies and was otherwise unconvincing. 700 F. Supp. 1232-33.

Judge Conboy also found, however, that TFG received profits in the form of a market advantage or value of use from the good will established by using the discounted robotics study as a means of familiarizing potential customers with TFG's products. The parties have sacrificed some trees and much ink over the issue of whether enhanced good will and market recognition may be treated as a profit under Section 504(b). The gravamen of TFG's argument is that such an award is impermissible because it is always nonquantifiable. However, we see no legal barrier to such an award under Section 504(b) so long as the amount of the award is based on a factual basis rather than "undue speculation." See Abeshouse v. Ultragraphics, Inc., 754 F.2d 467, 470 (2d Cir.1985). Although we recognize that proving the value received from an infringing product used to enhance commercial reputation may be difficult, we are unpersuaded that the difficulty is so universal that an award for such gains may never be made as a matter of law.

Nevertheless, the evidence in this case does not support such an award. Although some gain in market recognition was conceded by TFG's president, that gain cannot be attributed to the robotics study alone because other noninfringing studies were discounted as part of the same market strategy. Moreover, Judge Conboy found that the TFG study had not affected sales of the Predicasts study, even though the former had been distributed at a price that was only ten percent of the price of the latter.

Judge Conboy acknowledged the paucity of evidence on this issue. Nevertheless, he held that the value of enhanced market recognition could be quantified as "the difference between the prevailing market price of the study, $1500.00, and the actual sales price of $150.00," 700 F. Supp. at 1232, multiplied by the number of copies sold by TFG. We disagree.

There is no reason to believe that TFG would have sold, at a price of $1,500, the same number of robotics studies as it sold at the cut-rate price of $150. Indeed, the evidence is overwhelmingly to the contrary. The infringing study simply was not selling at all, and continuing to offer it at $1,500 was a profitless gesture. Moreover, the discounted price of $150 was not arrived at by calculating present losses on discounted sales against hoped-for future profits. The price of $150 was a nominal one designed solely to ensure that purchasers were seriously interested in industry studies and were thus potential customers.

Judge Conboy offered an alternative theory for an award of $1,350 multiplied by the number of sales based upon Deltak, Inc. v. Advanced Systems, Inc., 767 F.2d 357 (7th Cir.1985). Deltak involved two competitors, Deltak and Advanced Systems, Inc. ("ASI"), in the business of data-processing training. As part of a data-processing kit, Deltak marketed a pamphlet intended as a guide to its publications for customers. On one page, the pamphlet listed data-processing tasks, and on the facing page, the pamphlet listed Deltak publications intended as aids to the performance of those data-processing tasks. See id. at 358. The list price of Deltak's kit was $5,000. See id. at 360. Deltak's competitor ASI distributed an infringing pamphlet that included a verbatim transcription of Deltak's list of data-processing tasks, but that substituted the names of its own data-processing publications for the names of Deltak's data-processing publications. ASI distributed this infringing pamphlet without charge. See id. at 358-59. As in the instant case, Deltak had not registered its copyright before the infringement, and statutory damages were not available. See id. at 359. Because Deltak suffered no reduction in sales, its losses were zero. Because ASI had not charged for its infringing pamphlet, it had no profits.

Faced with the seemingly inequitable prospect of awarding no damages, the Seventh Circuit determined that an award for the value of use of the infringing studies was appropriate. By copying the Deltak pamphlet, the court held, ASI had saved itself the cost of purchasing the Deltak kit. "Each of the copies ASI distributed had a value of use to it equal to the acquisition cost saved by infringement instead of purchase, which ASI was then free to put to other uses." Id. at 361 (citing Aitken, Hazen, Hoffman, Miller, P.C. v. Empire Construction Co., 542 F. Supp. 252, 262-63 (D.Neb.1982); Nucor Corp. v. Tennessee Forging Steel Serv., Inc., 513 F.2d 151, 152-53 (8th Cir.1975); Atlantic Monthly Co. v. Post Pub. Co., 27 F.2d 556, 560 (D.Mass.1928)). In the court's view, ASI had saved itself "what a willing buyer would have been reasonably required to pay to a willing seller for [the infringed party's] work." Deltak, 767 F.2d at 362 (quoting Sid & Marty Krofft Television Productions, Inc. v. McDonald's Corp., 562 F.2d 1157, 1174 (9th Cir.1977)). Although the court declined to award the full $5,000 list price of Deltak's kit as a measure of the value of use because actual sales of the kit appeared to have been at much lower prices, it remanded to the district court for a finding of the fair market value of the Deltak kit.

Applying Deltak to the instant case, TFG might be said to have profited from the value of use in the amount of the saved acquisition cost of the Predicasts study. This could be calculated as the number of discounted sales multiplied by $1,350 (the sales price of the Predicasts study less $150 in TFG's gross receipts already reflected in the award for actual profits). Deltak thus provided an alternative ground for the award by Judge Conboy.2

We decline to adopt Deltak's approach. As the Nimmer text has noted, it is based "on the most transparent of fictions." 3 Nimmer Sec. 14.02[A] at 14-16--14-17. TFG no more priced the BTA study and then decided to copy than a purse-snatcher decides to forgo friendly negotiations. BTA did not, because of the infringement, lose sales that it would have made to TFG, and TFG did not save money that it would have paid to BTA for copies of the Predicasts study. Moreover, the market value of the infringed study seems an anomalous measure of the value of use in the amount of the saved acquisition cost because that market value is in large part based on the good will attributed to the copyright plaintiff's trademark. In both Deltak and the instant case, the last thing the infringers wanted to buy and to sell was the actual material produced by their competitors under their competitors' name.

We see no room for such a speculative and artificial measure of damages under Section 504(b). The language of the provision speaks of "actual damages suffered by" the infringed party. That is hardly a reasonable description of the entirely hypothetical sales to TFG lost by BTA. Section 504(b) also authorizes damages for "profits of the infringer." True saved costs may well be counted as a gain in economic theory (although not by Deltak's market value test), but the statute had a more conventional view of profits in mind. It thus further states, "In establishing the infringer's profits, the copyright owner is required to present proof only of the infringer's gross revenue...." This language clearly indicates that Congress means "profits" in the lay sense of gross revenue less out-of-pocket costs, not the fictive purchase price that TFG hypothetically chose not to pay to BTA.

We believe, as does Nimmer, that Deltak is based on a perceived need to avoid "the anomaly of affording plaintiffs a right without a remedy." 3 Nimmer Sec. 14.02[A] at 14-14. It is surely true that where an infringer such as TFG sells the offending publication at a nominal price, and there is no evidence of lost sales of the infringed publication, a conventional profits test may seem inadequate. Nevertheless, we believe we must follow the statutory scheme. Our difficulty with expanding upon the design of the statute to remedy the seeming inadequacy derives not from a truncated sense of justice but from our conviction that Congress quite consciously limited the remedies available under Section 504(b). Although Section 504(b) may provide a conventional and narrow method of calculating damages and profits, the Copyright Act does not overlook the fact that this method, if the exclusive remedy, may lead to injustice and underenforcement of the copyright laws. Under Section 504(c), a copyright plaintiff may thus elect to recover statutory damages. The specific purpose of this section is to offer plaintiffs an alternative for use where Section 504(b) provides inadequate relief. The difficulty faced by BTA and the plaintiff in Deltak is that neither registered their copyright before the infringement and thus are not entitled to statutory damages. The issue before us, therefore, is whether we should in effect rewrite Section 504(b) because it provides inadequate relief to late registrants.

As the Nimmer text states in commenting upon Deltak,

Under the 1909 Act, the Supreme Court warned against ... a weak disincentive against purloining copyrighted material: "[A] rule of liability which merely takes away the profits from an infringement would offer little discouragement to infringers.... Even for uninjurious and unprofitable invasions of copyright the court may, if it deems just, impose a liability within statutory limits to sanction and vindicate the statutory policy."

Nonetheless, although Deltak's value of use standard serves the Supreme Court's goal of "discourag[ing] wrongful conduct," it is open to question whether it fits the Court's requirement of being "within statutory limits." In the context of the 1909 Act, for which the Court enunciated the foregoing language, the value of use theory would have been neither needed nor approved. For under the 1909 Act, precisely for those instances such as Deltak where "proof of damages or discovery of profits" is "difficult or impossible," Congress provided for statutory damages, an automatic measure of recovery to plaintiffs regardless of injury or profits. Moreover, such statutory damages were available in all infringement actions, even if the copyright had not been registered prior to the infringement. Moreover, given the availability of statutory damages, the courts under the 1909 Act rejected the "reasonable royalty" standard, a patent measure of damages that looks to the royalties customarily paid for the type of use to which the defendant has put the infringing material; the similarities between this reasonable royalty rule and Deltak's value of use theory are apparent.

Under the current Act, by contrast, plaintiffs such as Deltak who neglect to register their copyright before the infringement cannot recover statutory damages. Therefore, the question under the current Act becomes whether value of use is a species of actual damages falling within the statutory language of "the actual damages suffered by [the author] and any profits of the infringer." The value of use of the infringing material fits neither category....

Given that Congress deliberately excluded the recovery of statutory damages for unregistered copyrights under the current Act, the current statutory scheme suggests that Congress, in order to foster registration, chose that the penalty for failure to register is the loss of any recovery in situations such as Deltak.

Id. at 14-14--14-16 (footnotes omitted). That eminent text goes on to suggest that whether to follow Deltak involves a choice between "reason" (rejecting it) and "experience" (adopting it). We view the choice, however, as between carrying out or rejecting a rational, if strict, policy embraced by Congress. Our obligation in those circumstances not being in doubt, we decline to follow Deltak.

In doing so, however, we emphasize that we are not rejecting as a matter of law a view of "profits" under Section 504(b) that might include enhanced good will or market recognition. We hold only that the proof in the instant case is inadequate to support such an award. Where there is no evidence at all allowing an assessment of the value of such an enhancement, establishment of the fact of enhancement will not support an award. See Roy Export Co. v. Columbia Broadcasting System, 503 F. Supp. 1137, 1156-57 (S.D.N.Y.1980), aff'd, 672 F.2d 1095 (2d Cir.), cert. denied, 459 U.S. 826, 103 S. Ct. 60, 74 L. Ed. 2d 63 (1982) (benefit to infringer from unsponsored infringing broadcast not profit under Section 504(b) because it consists of unmeasurable good will with affiliates and increased prestige in industry). In the present case, for example, TFG's president admitted that discount sales of the infringing study along with two other discounted studies had led to increased sales of subsequent TFG studies. Had this issue been pursued and purchasers of the infringing study been matched with purchasers of later studies at proven prices, BTA might well have argued that revenues from such subsequent sales were "profits" under Section 504(b), thereby shifting the burden on expenses and causation to TFG. See Stevens Linen Associates, Inc. v. Mastercraft Corp., 656 F.2d 11, 15 (2d Cir.1981). The issue was not pursued, however. We therefore vacate the award of $49,950.

Finally, we turn to TFG's claim that the damage award must be reduced to reflect an apportionment between the value created by the infringing material and the value created by the non-infringing material. It is true that Judge Conboy found as a fact that only certain portions of TFG's robotics study infringed the Predicasts study. It is also true that "where an infringer's profits are not entirely due to the infringement, and the evidence suggests some division which may rationally be used as a springboard it is the duty of the court to make some apportionment." Orgel v. Clark Boardman Co., 301 F.2d 119, 121 (2d Cir.), cert. denied, 371 U.S. 817, 83 S. Ct. 31, 9 L. Ed. 2d 58 (1962); see also Sygma Photo News, Inc. v. High Society Magazine, 778 F.2d 89 (2d Cir.1985). The burden is on the infringer, however, to prove the portion of total profits attributable to non-infringing elements. Sheldon v. Metro-Goldwyn Pictures, 309 U.S. 390, 405-06, 60 S. Ct. 681, 686-87, 84 L. Ed. 825 (1940); Lottie Joplin Thomas Trust v. Crown Publishers, Inc., 592 F.2d 651, 657 (2d Cir.1978). Judge Conboy found that TFG had not carried its burden, and we agree. He found as a fact that the "heavily infringed" portions were the sections of both reports "that most gave them their value." 700 F. Supp. at 1241. That finding is not clearly erroneous. Moreover, we agree that here "infringed portions are so suffused and intertwined with non-infringing portions as to render [an apportionment] impossible" in the instant case. Id. Under such circumstances, no apportionment is appropriate. See Sheldon, 309 U.S. at 401-02, 60 S.Ct. at 684-85; Belford v. Scribner, 144 U.S. 488, 508, 12 S. Ct. 734, 740, 36 L. Ed. 514 (1892); Callaghan v. Myers, 128 U.S. 617, 665-66, 9 S. Ct. 177, 191, 32 L. Ed. 547 (1888); Alfred Bell & Co. v. Catalda Fine Arts, Inc., 86 F. Supp. 399 (S.D.N.Y.1949), aff'd as modified, 191 F.2d 99 (2d Cir.1951).

Affirmed in part, reversed in part.