Judge Davis's Damages Calculation in CSIRO v. Cisco

Today's post is a guest post by Professor Norman Siebrasse of the University of New Brunswick.

CSIRO v. Cisco, Case No. 6:11-cv-00343-LED  (E.D. Tex. 2014)

Judge Davis’s decision on damages in CSIRO v. Cisco is one of a handful of decisions to assess the quantum of a reasonable royalty for a standard essential patent, with infringement and validity having been stipulated. CSIRO's `069 patent relates to IEEE 802.11 wireless standards, in particular 802.11a, 802.11g, 802.11n and 802.11ac, but excluding 802.11b. The standards in question embodied the core technology of the ‘069 patent (p2), and that technology was an important factor contributing to the commercial success of the 802.11 products (p27). A RAND commitment was made with respect to the 802.11a standard, but not with respect to the other standards in issue (p8). However, Judge Davis made nothing of this point, as CSIRO has shown itself willing to license to all comers in any event.

Judge Davis largely rejected the evidence of the expert witnesses for both sides. CSIRO’s expert attempted to assess the value of the ‘069 technology by comparing the profitability of the infringing products with that of similar 802.11b products at a similar point in time (p11). This approach is sound, in that it represents an attempt to determine the value of the patented technology in comparison with the best non-infringing alternative, but it was rejected by Judge David largely for methodological reasons, such as inadequate sample size (p13). Cisco’s model was also rejected, primarily because it used as a basis a royalty that was negotiated in a non-arms length relationship, in which the royalty “was only one small part of the relationship created by the agreement” (p21). Consequently, the licence was not sufficiently comparable even to serve as a starting point for further adjustment (p20).

In the absence of satisfactory guidance from the parties, Judge Davis evaluated the evidence directly. He used as a baseline the rates actually discussed by CSIRO and Cisco in licensing discussions which ultimately did not come to fruition, $0.90 to $1.90 (p25). It was also clear that the parties would have a negotiated a rate varying with volume according to a discount schedule. Judge Davis then applied the Georgia-Pacific factors determine what adjustments to these initial rates were appropriate (p25). He ultimately concluded that while some factors favored an upward adjustment and other a downward adjustment, the sum of the factors was “essentially in equipoise,” and accordingly “no overall adjustment is needed to the baseline rates and a range of $0.90 to $1.90 is the appropriate outcome of the hypothetical negotiation here” (p28).

While the overall outcome was even, I am not sure I entirely agree with Judge Davis’ application of a number of the Georgia-Pacific factors. The first non-neutral factor was Factor 3, which considers the nature and scope of the license. Judge David remarked that “Although not explicit, it is reasonable to infer that both Cisco’s suggested royalty rate and CSIRO’s Voluntary Licensing Program assumed a worldwide license. Since the Court is considering damages only for infringing sales in the United States,  this factor favors adjusting the baseline rates downward” (p26). It is not clear to me why a worldwide licence would warrant a higher royalty rate. Obviously, there would be more sales with a worldwide licence, but that is reflected in the sales volume to which the royalty is applied, and also in the discount schedule. I don’t see why it should affect the rate to which the discount schedule is applied.

With respect to Factor 4, which looks at the licensor’s established policies and marketing programs, Judge Davis stated that

CSIRO was very willing to license the patented technology, sending offer letters to many wireless industry firms. Further, under its RAND obligation, CSIRO had a binding commitment to license the ’069 Patent with regard to 802.11a products. The Court agrees with both experts that this factor favors a downward adjustment of the baseline rates (p26-27).

Since the baseline rate was the rate offered by CSIRO to Cisco, this means a downward adjustment from the rate that CSIRO was willing to offer to Cisco. I don’t see the rationale for this adjustment. If Patentee A does not licence, and makes its money by exploiting the patent itself, and the closest comparison product is made by Patentee B, which exploits its patent rights by licensing, the rate charged by B should be adjusted upwards before it can be imputed to A, on the view that profit margins from exploiting the patent itself are higher. But why would the fact that CSIRO was willing to licence require a downward adjustment from the rate that CSIRO itself was willing to licence at?

I have much the same concern regarding Factor 5, which considers the commercial relationship existing between CSIRO and Cisco. Judge Davis noted that:

In this case, the parties are not competitors. CSIRO is a government organization focused on research and development. Cisco and Linksys engage in the marketing and sale of information technology products. CSIRO needed to license the ’069 Patent in order to commercialize and monetize it. Accordingly, this factor favors a downward adjustment of the rates (p27).

At the time CSIRO made the offer to Cisco, it was a government R&D organization, and Cisco was a product vendor, and the same is true at the time of infringement. Why is a downward adjustment necessary when the actual offer and the hypothetical negotiation took place in exactly the same circumstances? Again, if the comparison licence was different—the parties to the comparison licence were head-to-head market competitors—I can see the need for adjustment, but not here. The court makes essentially that point with respect to Factor 11 (p28), and I don't see why it is different with these factors.

Another concern I have is with the baseline rate itself. These offers and counter-offers, in the range of $.90 to $1.90, were discussed before the validity of the patent has been tested in court. Judge Davis started his analysis by noting that it is established law that “A hypothetical negotiation assumes the asserted patent claims are valid and infringed. Accordingly, the Court need not ‘discount’ the hypothetical rate due to uncertainty regarding validity or infringement.” (p23-24). But this implies that conversely, negotiations, whether consummated or not, that take place before validity has been established in litigation, are implicitly discounted to allow for the possibility of invalidity. (Non-infringement would not be an issue for a SEP). Maybe the reason that Cisco’s offer was lower than CSIRO was willing to accept was because Cisco was skeptical of the patent’s validity. It strikes me that this consideration is not neutral, but warrants an upwards adjustment of the baseline royalty. This is particularly a concern because Judge Davis noted that chip prices were depressed in the damages period because of “rampant infringement which occurred in the wireless industry” (p22). This is point is also relevant to some of Judge Davis’ criticisms of CSIRO’s expert. He says “Illogically, [CSIRO’s expert] proposes that at the hypothetical negotiations CSIRO and Cisco would have agreed to prices higher than CSIRO’s asking price.” I don’t see why this is illogical – it follows directly from the fact that the hypothetical negotiation takes place with validity established, while the actual negotiation did not.