# Ericsson v. D-Link: How RAND Royalties Are Actually Tried

> The Federal Circuit's first RAND damages decision: juries must hear the patentee's actual RAND commitment and award only the incremental value of the invention.

Topic: Patents  |  Author: Lidiia Levitska  |  Source: Intellectual Property Law (outsideipcounsel.com)
Canonical: https://outsideipcounsel.com/blog/ericsson-v-dlink-rand-apportionment-jury-instructions/


*Ericsson, Inc. v. D-Link Systems, Inc.*, 773 F.3d 1201 (Fed. Cir. 2014), decided December 4, 2014, is the Federal Circuit's foundational statement on how reasonable royalties are litigated when the asserted patents are encumbered by a commitment to license on reasonable and non-discriminatory (RAND) terms. Writing for the panel, Judge O'Malley confronted a question the court had never squarely answered: what a jury must be told, and what it may not be told, when the patents in suit are essential to an industry standard. The opinion reshaped standard essential patent (SEP) damages practice and remains the doctrinal backbone that every SEP trial team works from.

## At a glance

- **Case:** *Ericsson, Inc. v. D-Link Systems, Inc.*, 773 F.3d 1201 (Fed. Cir. 2014), No. 13-1625
- **Decided:** December 4, 2014; opinion by Judge O'Malley, before Judges O'Malley, Taranto, and Hughes. Judge Taranto filed an opinion dissenting in part, joining all of the opinion except Part II.A.2, a claim-construction and infringement holding unrelated to RAND
- **Holding:** In a trial over RAND-encumbered SEPs, the jury must be instructed on the patentee's actual RAND obligation, given only the relevant and RAND-adjusted *Georgia-Pacific* factors, and told that damages must be apportioned to the incremental value of the invention rather than the value of the standard
- **Status:** Final. Affirmed in part, reversed in part, vacated in part, and remanded; the jury's damages award and the ongoing royalty award were both vacated

## The standard-setting bargain and the problem it creates

Modern wireless and networking technology runs on standards. The IEEE 802.11 (Wi-Fi) standard at issue in *Ericsson* lets devices from different manufacturers interoperate, but only because thousands of engineering choices were locked in by a standard-setting organization (SSO). Many of those choices are patented. Once a patented technique is written into a standard, every compliant product must practice it, and the patent becomes essential.

That creates leverage the patent would not otherwise carry. A patentee whose invention is baked into a standard could, in theory, hold an entire industry hostage, demanding royalties that reflect not the merit of the invention but the cost of abandoning the standard. This is the concern the literature calls patent hold-up. To contain it, SSOs require members to commit, in advance, to license essential patents on RAND (or FRAND, adding "fair") terms. Ericsson had made exactly such a commitment to the IEEE.

The doctrinal difficulty is that patent damages law was built for ordinary patents. The canonical framework, the fifteen *Georgia-Pacific Corp. v. United States Plywood Corp.*, 318 F. Supp. 1116 (S.D.N.Y. 1970), factors, models a hypothetical negotiation between a willing licensor and licensee. Several of those factors assume the patentee holds an unconstrained right to exclude. That assumption is false for a RAND-committed SEP, and *Ericsson* is the case that forced the framework to account for it.

## The facts and the jury verdict

Ericsson asserted patents essential to the 802.11(n) Wi-Fi standard against D-Link and other manufacturers of routers, laptops, and other networking products. A jury in the Eastern District of Texas found infringement and awarded Ericsson roughly $10 million, approximately fifteen cents per infringing device, a figure the defendants attacked as untethered from the incremental value of Ericsson's specific contributions to the standard. D-Link argued that Ericsson's damages expert had relied on licenses that were themselves keyed to the value of whole end products, and that Ericsson's counsel had invited the jury to capture the value of a laptop rather than the value Ericsson's patents actually added.

That evidentiary attack largely failed on appeal. The Federal Circuit held that the district court properly admitted the license testimony, reasoning that real-world licenses are rarely perfectly analogous and that imperfect comparability generally goes to weight rather than admissibility, and that because Ericsson's expert had explained the need to discount those licenses, the testimony violated neither *Garretson* apportionment nor the entire market value rule's evidentiary principle. The separate objection to counsel's references to the price of laptops was waived. The court vacated the damages award for a different reason: the jury instructions.

The trial court had given the jury all fifteen *Georgia-Pacific* factors, over objection, without considering their relevance to the record. It granted D-Link's request for a RAND instruction only in part, adding a sixteenth factor that told the jury it "may consider ... Ericsson's obligation to license its technology on RAND terms," and it refused to instruct on patent hold-up and royalty stacking. The defendants appealed the damages award and the jury instructions.

## What the Federal Circuit required

The panel vacated the damages award and, in doing so, laid down the ground rules for SEP royalty trials.

First, **apportionment is mandatory**. Building on *Garretson v. Clark*, 111 U.S. 120 (1884), and its own decisions in *VirnetX, Inc. v. Cisco Systems, Inc.*, 767 F.3d 1308 (Fed. Cir. 2014), and *LaserDynamics, Inc. v. Quanta Computer, Inc.*, 694 F.3d 51 (Fed. Cir. 2012), the court held that damages must reflect the incremental value the patented invention adds, not the value of the standard or the end product. For SEPs the apportionment must strip out any value attributable to standardization itself. A patent may be valuable partly because it was adopted into a widely used standard, but that added value flows from the collective decision to standardize, not from the invention, and the patentee is not entitled to capture it.

Second, **the jury must be instructed on the actual RAND commitment**. A generic reference to "RAND terms" is not enough, because RAND terms vary from case to case. Ericsson had pledged to the IEEE that it would grant a license under reasonable rates to an unrestricted number of applicants worldwide, on terms demonstrably free of unfair discrimination, and the court held the jury should have been told about that specific promise rather than the district court's abstract sixteenth factor. The court also faulted the permissive framing: a jury must be informed of its obligation, not merely its option, to take the commitment into account. A RAND commitment is not a mere background fact. It constrains the hypothetical negotiation.

Third, **the *Georgia-Pacific* factors must be pruned and adjusted**. The court refused to bless the reflexive recitation of all fifteen factors. Many, it explained, are irrelevant to a RAND-encumbered patent, several are contrary to RAND principles, and others require modification. A trial court must give the jury only the factors that fit the case and must not read factors that would let the jury award the holdup value the RAND commitment was meant to foreclose.

Fourth, **hold-up and royalty stacking require evidence**. The court declined to require an anti-hold-up or anti-stacking instruction as a matter of course. A jury may consider those theories only where the accused infringer introduces actual evidence that hold-up or stacking is present in the specific case. Fear of hold-up in the abstract does not entitle a defendant to an instruction; concrete proof does.

## What the decision changed

*Ericsson* converted RAND from a contractual abstraction into a set of concrete trial rules. Before it, courts improvised. After it, an SEP damages case has a recognizable shape: identify the RAND commitment, apportion to the incremental value of the invention, cull the *Georgia-Pacific* factors, and hold hold-up and stacking theories to a record-evidence standard.

The opinion also struck a deliberate balance. It rejected the defense-side wish for automatic hold-up and stacking instructions, and it rejected the patentee-side practice of feeding the jury the full end-product value dressed up in the standard's importance. The court's later decision in *CSIRO v. Cisco Systems, Inc.*, 809 F.3d 1295 (Fed. Cir. 2015), refined the apportionment analysis further, confirming that *Ericsson* left room for varied damages methodologies so long as they genuinely apportion.

## Open questions

*Ericsson* set the framework but left the hardest number-crunching to future cases. It did not fix a single required methodology for calculating a RAND royalty, and it declined to make the smallest salable patent-practicing unit an absolute rule, leaving trial courts to police the royalty base and rate case by case. It also did not resolve how apportionment interacts with comparable-license evidence, which often reflects portfolio-wide or standard-wide rates rather than the incremental value of the specific asserted patents. Courts continue to wrestle with how to admit real-world licenses without reimporting the very value *Ericsson* said must be stripped out. Finally, the opinion addressed United States damages law, not the growing body of global FRAND rate-setting in foreign courts, which now shadows every major SEP dispute.

## Implications for inventors and businesses

- **Document the RAND commitment and build damages around it.** Patentees asserting SEPs should expect the jury to be told about the commitment and should structure damages proof to survive apportionment, using comparable licenses that themselves reflect incremental value.
- **Implementers should invest in apportionment and evidence of hold-up.** Accused infringers gain the most by attacking the royalty base, insisting on incremental-value apportionment, and, where the facts support it, putting on concrete evidence of hold-up or stacking to earn the corresponding instructions.
- **Tailor jury instructions early.** Both sides should litigate the *Georgia-Pacific* factor list and the RAND instruction well before trial, because *Ericsson* makes those instructions a frequent basis for reversal.
- **Treat standardization value as off-limits.** The value a patent gains purely from being adopted into a standard belongs to the standard, not the patentee, and damages theories that ignore that line invite vacatur.

## Frequently asked questions

**What does a RAND commitment change about a damages trial?** The jury must be told what RAND commitment the patentee made and instructed that the royalty has to reflect that promise. The court must also prune the *Georgia-Pacific* factors, removing or adjusting the ones that assume an unconstrained monopoly, and the award must be apportioned to the incremental value the patented feature adds, not the value of the whole standard.

**Did Ericsson v. D-Link ban hold-up and royalty-stacking arguments?** No. The Federal Circuit held that a jury should hear hold-up or stacking theories only where the accused infringer puts in actual evidence that hold-up or stacking occurred in the case at hand. A generic instruction untethered to record evidence is improper.

**Does the smallest salable patent-practicing unit rule always apply?** The court reaffirmed apportionment as the governing principle but declined to make the smallest salable unit a rigid requirement in every case. The touchstone is that damages reflect the incremental value of the invention, and the royalty base and rate together must accomplish that apportionment.

## Authorities and sources

- *Ericsson, Inc. v. D-Link Systems, Inc.*, 773 F.3d 1201 (Fed. Cir. 2014), No. 13-1625 (Dec. 4, 2014). [Slip opinion, U.C. Berkeley Law copy](https://www.law.berkeley.edu/wp-content/uploads/2016/05/Ericsson-v-D-Link-773_F.3d_1201.pdf).
- [WIPO Lex entry for Ericsson v. D-Link, 773 F.3d 1201](https://www.wipo.int/wipolex/en/judgments/details/2215).
- Wiggin and Dana LLP, [In Ericsson v. D-Link the Federal Circuit Provides Guidance on Damages for Standard Essential Patents](https://www.wiggin.com/publication/in-ericsson-v-d-link-the-federal-circuit-provides-guidance-on-damages-for-standard-essential-patents/).
- Essential Patent Blog, [Federal Circuit gives guidance on litigating RAND royalty (Ericsson v. D-Link)](https://www.essentialpatentblog.com/2014/12/federal-circuit-gives-guidance-on-litigating-rand-obligation-ericsson-v-d-link/).
- American Antitrust Institute, [Federal Circuit Follows AAI Recommendation on RAND Royalty Jury Instructions](https://www.antitrustinstitute.org/work-product/federal-circuit-follows-aai-recommendation-on-rand-royalty-jury-instructions-ericsson-v-d-link/).
- Stout, [Takeaways From Ericsson v. D-Link, et al.](https://www.stout.com/en/insights/article/takeaways-from-ericsson-v-d-link-et-al).
- *Georgia-Pacific Corp. v. United States Plywood Corp.*, 318 F. Supp. 1116 (S.D.N.Y. 1970); *LaserDynamics, Inc. v. Quanta Computer, Inc.*, 694 F.3d 51 (Fed. Cir. 2012); *VirnetX, Inc. v. Cisco Systems, Inc.*, 767 F.3d 1308 (Fed. Cir. 2014).