Opinion analysis: Court limits fee awards in copyright cases

The desultory argument in Rimini Street v. Oracle USA (discussed in my earlier post) suggested a consensus hostile to the broad fee award approved by the lower court, so nothing in today’s unanimous opinion reversing the court of appeals comes as a surprise.

The case involves the problem of litigation expenses in copyright litigation. Specifically, it calls for interpretation of Section 505 of the Copyright Act, which defines the expenses that a prevailing party in copyright litigation can obtain as including “full costs.” All agree that “costs” in Section 505 include the six narrow categories of “taxable costs” defined in the general provisions of the Judicial Code (28 U.S.C. § 1920). The question is whether an award of “full” costs can include costs beyond the costs that are “taxable” under Section 1920. In this case, for example, when respondent Oracle prevailed in its claim that petitioner Rimini Street had infringed Oracle’s copyright, the lower courts awarded Oracle $36 million in copyright-related damages, $29 million in attorney’s fees, $3.4 million in taxable costs and $13 million in additional nontaxable costs (covering expert witnesses, e-discovery, jury consulting and the like). The court held yesterday that the $13 million award of nontaxable costs was a mistake.

Justice Brett Kavanaugh’s opinion is succinct and linear, taking the approach that the court’s prior treatment of Section 1920 compels a narrow reading of Section 505. For him, Section 1920 and related provisions “create a default rule and establish a clear baseline against which Congress may legislate.” As he explains, when “Congress wants to authorize awards … beyond the six categories, Congress may do so,” and has done so on numerous occasions. “But absent such express authority, courts may not award litigation expenses” beyond those categories.

To document the point that the court’s earlier cases “have consistently adhered to that approach,” Kavanaugh summarizes three cases that require an “explicit” statutory reference to justify an award of expert-witness fees when a statute authorizes an award of costs. For Kavanaugh, the “clear rule” from those cases is that a “statute awarding ‘costs’ will not be construed as authorizing an award … beyond the six categories … absent an explicit statutory instruction to that effect.” Because the Copyright Act does not “explicitly authorize” such an award, the lower courts erred in providing one. (If you sense unwarranted repetitiveness in my summary – the word “explicit” in one form or another appears seven times in Kavanaugh’s opinion.)

The opinion closes with a summary rejection of Oracle’s principal arguments. First, Oracle argued that the statute’s reference to “full” costs takes Section 505 out of the general rule. Kavanaugh disagrees, characterizing “full” as “an adjective that means the complete measure of the noun it modifies,” which thus can “not alter the meaning of the word ‘costs.’” For him:

The word “full” operates in the phrase “full costs” just as it operates in other common phrases: A “full moon” means the moon, not Mars. A “full breakfast” means breakfast, not lunch. A “full season ticket plan” means tickets, not hot dogs. So too, the term “full costs” means costs, not other expenses.

Kavanaugh also rejects Oracle’s argument that historical practice predating the Copyright Act establishes that “full costs” is a “historical term of art” that goes beyond the six listed categories from Section 1920. For one thing, in his view the line of cases interpreting Section 1920 means “that courts should not undertake extensive historical excavation to determine the meaning of costs statutes,” because they should bear the same meaning “regardless of when individual subject-specific costs statutes were enacted.” Moreover, he states that Oracle has failed to “persuasively demonstrat[e] … an established meaning … that covered more than the full amount of the [listed] costs.” Among other things, Kavanaugh points out that “none of the more than 800 available copyright decisions from 1831 to 1976 … awarded expenses other than those specified by the applicable state or federal law.”

Finally, Kavanaugh rejects Oracle’s argument that the court’s reading leaves “full” an “unnecessary surplusage” in the statute. Kavanaugh points to the period from 1831 to 1976, when awards of costs in copyright litigation were mandatory: “During that period, the term ‘full’ fixed both a floor and a ceiling for the amount of ‘costs’ that could be awarded.” Kavanaugh acknowledges that fee awards have been discretionary since 1976, so “full” no longer bears that weight, because “district courts could award any amount of costs up to 100 percent.” Kavanaugh ridicules the idea that the shift from mandatory to discretionary fee awards (apparently designed to cut back awards) changed the meaning of the word “costs” (in a way that implicitly would expand awards): “It makes little sense to think that Congress in 1976, when it made the award of full costs discretionary rather than mandatory, silently expanded the kinds of expenses that a court may otherwise award as costs in copyright suits.”

Additionally, Oracle’s argument would render superfluous an entire sentence of Section 505 – the sentence in Section 505 that explicitly authorizes an award of attorney’s fees (justifying the $29 million attorney’s fee award in this case). Kavanaugh’s final answer directly challenges Oracle’s reliance on statutory surplusage. For Kavanaugh, “[r]edundancy is not a silver bullet.” Rather, he explains, redundancy is “hardly unusual,” because “[s]ometimes the better overall reading of the statute contains some redundancy.”

The unanimous opinion underscores that, as the oral argument suggested, the justices saw this as an easy case. I do not see it as breaking important doctrinal ground. In the end, this case probably will serve as a visible precedent lower courts will cite to justify a narrow reading of the “costs” available under the various federal statutes that specify fee awards in federal litigation.