Publications

Greetings, Court Fans!

OT18 has been marked by its unanimity, with 13 out or 18 decisions so far coming down without dissent. This week, we can add two more to that tally, but we’ll begin with the split decision.

On its surface, BNSF Railway v. Loos (No. 17-1042) raised an answered an extremely narrow and uncontroversial question: Whether a railroad’s payment to an employee of damages for lost wages caused by the railroad’s negligence is taxable “compensation” under the Railroad Retirement Tax Act (RRTA). By a 7-2 margin, the Court answered that question in the affirmative. But what was more interesting, perhaps, is was left unsaid in the reasoning behind that answer.

Michael Loos was an employee of BNSF Railway Company when, as a result of the railroad’s negligence, he fell into a hidden drainage grate and injured his knee. As a result of his injury, he missed several months of work, and continued to have injury-related absences after he returned. When the company sought to fire him for his repeated absences, he flipped the switch and filed a lawsuit seeking damages for his injuries. A jury awarded him over $100,000, including $30,000 designated for lost wages. But rather than pay that full amount, BNSF sought to withhold $3,675, contending that amount was owed to the IRS as taxable “compensation” under the RRTA. Enacted in 1937, along with the Railroad Retirement Act (RRA), the RRTA imposes a payroll tax on both railroads and their employees. The taxes owed under the RRTA and benefits paid under the RRA are both measured by the employee’s “compensation,” which is defined in both statutes as “any form of money remuneration paid to an individual for services rendered as an employee.” The District Court and the Eighth Circuit both held that damages for lost wages did not constitute “compensation” and was therefore not taxable. Ever desirous to pay the taxman, BNSF brought its case all the way to the Supreme Court, which agreed that, yes, the $30k is taxable compensation—meaning that the finally victorious BNSF . . . must now pay its own portion of the payroll tax. (As will be seen, the majority and dissent had differing opinions on why BNSF would want to fight so hard for the right to pay taxes.)

Writing for the majority, Justice Ginsburg drew an analogy between the railroad retirement system and the Social Security system. As the Court has previously held, the railroad retirement system was set up to provide benefits to railroad workers “correspond[ing] . . . to those an employee would receive were he covered by the Social Security Act.” Just as the Federal Insurance Contributions Act (FICA) taxes employers and employees to fund Social Security benefits, the RRTA taxes employers and employees to fund railroad retirement benefits. Both FICA and the Social Security Act (SSA) impose taxes on “wages,” the equivalent of “compensation” in the RRTA and the RRA. The terms “wages” and “compensation” are similarly defined by their respective statutes to cover remuneration for services performed by an employee. Therefore, the Supreme Court’s previous interpretation of the term “wages” in the social-security context informed its interpretation of the term “compensation” in the railroad-retirement context. In two cases from the 1970s, the Court made clear that “wages” under the SSA and FICA included pay for active service as well as pay received for periods of absence from active service, and specifically that backpay for time lost due to “the employer’s wrong” counted as “wages.” In order to maintain a corresponding construction of the RRTA and RRA, the Court held that “compensation” under the RRTA encompasses not only pay for active service but also pay for periods of absence from active service—provided that the remuneration in question stems from the “employer-employee relationship.” Because Loos recovered lost wages as a result of a period of absence from active service stemming from the employer-employee relationship, the portion of his damages that corresponded with lost wages was taxable under the RRTA.

Justices Gorsuch and Thomas were having none of this. Taking the lead, Justice Gorsuch began by asking just why in the world BNSF cared so much about withholding $3,675 from Loos’s judgment (which, by the way, would mean that it also had to pay its $3,675 portion of the payroll tax), that it was willing to fight all the way to the Supreme Court. While the majority had accepted BNSF’s explanation that it needed to do its part to ensure that the railroad retirement system remained solvent, Gorsuch accepted the view of some amici that BNSF really wanted to establish a precedent that lost-wages damages are taxable so that it could extract more favorable settlements from its injured employees: “Forgo trial and accept a lower settlement, they will tell injured workers, and in return we will designate a small fraction (maybe even none) of the payments as taxable lost wages.” (It did not seem to bother Gorsuch that this is already the state of affairs for virtually all other settlements that might include lost wages, given the Court’s earlier construction of “wages” under FICA.) Turning to the merits, Gorsuch made the rather obvious statutory argument that “[i]nstances of being ‘compensation’ for ‘services rendered as an employee,’ it seems more natural to say that the negligence damages BNSF paid are ‘compensation’ to Mr. Loos for his injury.” Looking further to the statutory history (as opposed, he was quick to note, to dreaded “legislative history”), Justice Gorsuch observed that an early version of the RRTA defined “compensation” to include remuneration “for time lost,” but that portion was removed from the statute in 1975. Turning finally to the main basis for the majority’s decision, Justice Gorsuch rejected the idea that prior decisions construing the SSA and FICA are binding in this context. Those cases “concerned a different statute, a different legal claim, and a different factual context.” What’s more, there have been suggestions in ensuing years that the SSA holdings were “motivated more by a policy concern with protecting the employee’s full retirement to Social Security benefits than by a careful reading of the Social Security Act.”

Justice Gorsuch closed his dissent with an interesting observation about what did not lead the majority to rule in BNSF’s favor. Although the IRS has for eighty years interpreted the term “compensation” in the RRTA to encompass remuneration for periods of absence from work, the majority opinion did not even mention the word “Chevron.” BNSF barely argued for Chevron deference at all in its brief, and at oral argument it came up only in the final seconds, when BNSF’s lawyer uttered the name with the caveat that he “hate[d] to cite it.” This, of course, has been a trend in recent terms—the silent overruling (or at least ignoring) of Chevron; and it’s a cause Justice Gorsuch in particular has championed (albeit a bit more loudly). So, while he and Thomas weren’t willing to join the majority’s opinion, Justice Gorsuch at least took heart that his “colleagues rightly afford[ed] the parties before us an independent judicial interpretation of the law.”

Next up, we have Fourth Estate Public Benefit Corp. v. Wall-Street.com, LLC (No. 17-571), in which a unanimous Court resolved a circuit split over what a plaintiff must do before bringing a copyright infringement suit. Fourth Estate, an online news organization, licensed some of its works to another news website, Wall-Street.com. The license required Wall-Street to remove any of Fourth Estate’s articles if it cancelled the license agreement, but Wall-Street ignored that, continuing to display the articles even after it terminated the license. Fourth Estate filed copyright applications for its articles with the Register of Copyrights, and then promptly sued Wall-Street for infringement. But when the suit was filed, the Register had not yet acted on Fourth Estate’s applications. The District Court dismissed Fourth Estate’s claims as premature, because 17 U.S.C. § 411(a) provides that “no civil action for infringement of the copyright in any United States work shall be instituted until . . . registration of the copyright claim has been made in accordance with” the copyright laws. The Eleventh Circuit affirmed, disagreeing with a Ninth Circuit decision holding that a plaintiff could sue so long as it had applied for registration, even if the application had not yet been granted by the Copyright Office.

Writing for a unanimous Court, Justice Ginsburg agreed with the Eleventh Circuit. (This despite the fact that she did not attend oral argument because she was still recovering from surgery at the time.) As noted, Section 411(a) requires that “registration of the copyright claim has been made” before a claimant can commence an infringement suit. But the statute does not explicitly say what “registration . . . has been made” means. Fourth Estate offered an “application approach,” meaning that so long as a claimant has submitted an application (and taken all steps necessary for the Copyright Office to act on it), the claimant can sue for infringement. As it did below, Wall-Street urged a “registration approach,” requiring the claimant to not only apply for registration but to obtain it before suing. Justice Ginsburg agreed with Wall-Street’s registration approach. For one, that interpretation is the more natural reading of the phrase “registration . . . has been made.” For two, other provisions of the copyright laws are more consistent with this interpretation. For example, another sentence of Section 411(a) establishes a procedure under which a claimant whose application has been denied by the Copyright Office can sue for infringement, a process that would be superfluous if the claimant did not have to wait for the Copyright Office to act. And Section 408(f) allows claimants to sue for infringement before attempting to register a copyright, where certain more-demanding circumstances are present. That process too would be unnecessary if the only prerequisite to suing under Section 411(a) was filling out an application.

Fourth Estate offered several statutory-interpretation arguments for its application approach, but the Court found these unpersuasive for various reasons (which we’ll mercifully ignore here). Its most potent argument, though, was the well-established principle of copyright law that an author’s exclusive rights in his or her work arise immediately with the work’s creation. How, Fourth Estate asked, can you square the idea that works are protected by copyright as soon as they are created with the statutory requirement that you have to successfully register the works with the Copyright Office before suing? Quite simply, the Court said: While a claimant must register the work before suing for infringement, the claimant’s remedies, if successful in the suit, can include pre-registration damages as well. True, the Court recognized, waiting until a copyright application is granted to sue may thwart claimants in some circumstances (e.g., where the risk of harm is immediate and can’t be remedied later or where the Copyright Office takes a long time), but those circumstances are rare. And to the extent they’re a problem, claimants can invoke procedures like Section 408(f), which allows pre-registration suits.

Continuing with the theme of unanimous copyright decisions, the Court held in Rimini Street, Inc. v. Oracle USA, Inc. (No. 17-1765) that successful litigants in a copyright infringement case can recover the same set of costs, but no more, that are recoverable in a typical non-copyright suit.

Oracle develops a variety of business software (databases, cloud systems, etc.). Its products are so widely adopted that whole industries have arisen to provide aftermarket support systems for Oracle’s software packages. One of these companies is Rimini Street, which sells products and services that let Oracle customers modify Oracle’s base offerings to meet their businesses’ specific needs. Eventually, the symbiotic relationship between Oracle and Rimini Street broke down, and in 2010 Oracle sued, alleging that Rimini Street’s products were infringing Oracle’s software copyrights. Oracle ultimately prevailed, and a jury awarded it tens of millions of dollars in damages. The District Court also awarded Oracle more than $15 million in costs. Some of this was for things like transcript fees and witness expenses, costs that are recoverable under the general federal cost statutes, Section 1821 and 1920. But the bulk of the award, nearly $13 million was for things like expert witness fees, contract attorney services, jury consulting, and other types of fees (other than attorneys’ fees) that are commonly incurred in complex civil litigation, but which fall outside the “taxable costs” recoverable under Sections 1821 and 1920. Applying Ninth Circuit precedent, the District Court reasoned that the costs recoverable in copyright suits are much more expansive than those recoverable in other federal-court litigation, because one of the copyright statues empowers district courts to award “full costs” to a party in copyright litigation. Other circuits have disagreed with the Ninth Circuit on this question, interpreting Section 505’s “full costs” language as simply referring to the same set of costs recoverable under Section 1821 and 1920. The Supreme Court granted certiorari to resolve this split (one that is consequential given that a significant share of high-value copyright litigation takes place in the Ninth Circuit, California specifically).

Writing for a unanimous Court, Justice Kavanaugh rejected the Ninth Circuit’s broad interpretation of “full costs.” More than two hundred federal statutes explicitly authorize the award of costs to prevailing parties in litigation. The Court has long held that in determining what costs are recoverable under these statutes, Sections 1821 and 1920 establish the baseline. While Congress can go beyond that baseline, by allowing for the award of a broader set of costs, it must do so explicitly. Justice Kavanaugh rejected each of the three arguments Oracle advanced for why Congress had explicitly authorized more expansive costs. First, Oracle argued that “full” means “full”: by using that adjective, Congress intended to award litigants all their costs. Kavanaugh agreed, but pointed out that “costs” also means “costs.” Saying that litigants could recover their “full costs” just means they can recover all the “costs,” as defined in Section 1821 and 1920; it does not gives “costs” a different meaning than it ordinarily has.  Next, Oracle argued that the history of the Copyright Act suggests a broader understanding of costs than those defined in Sections 1821 and 1920. But Kavanaugh was not having that either, because Congress’s basic purpose in drafting Sections 1821 and 1920 was to standardize the set of costs recoverable in litigation, dispensing with sometimes idiosyncratic differences based on the nature of the dispute. For these reasons, the Court has long refused to “undertake extensive historical excavation” to decide what “costs” means in a particular context, and it refused again to do so here. Finally, Oracle argued that the Court’s reading makes “full” superfluous: If it just allows litigants to recover the same costs defined in Sections 1821 and 1920, then there’s no need for the word “full” at all. But “redundancy is not a silver bullet,” Justice Kavanaugh noted. Instead, it is simply a “clue” that one interpretation of the statute that avoids redundancy might be better than a reading that results in it. But sometimes (here, for instance) that clue is not enough to overcome a much more persuasive reading of a statute that entails some limited redundancy.

That’ll do it for this week. Stay tuned for the latest when it arrives.

Dave and Tadhg