SO THAT’S WHAT “RAND” MEANS?: A Brief Report on the Findings of Fact and Conclusions of Law in Microsoft v. Motorola

Guest Post By Jorge L. Contreras, Associate Professor of Law, American University Washington College of Law

In a meticulous 207-page opinion released on April 25, Judge James Robart in the Western District of Washington has crafted the first-ever judicial determination of a “reasonable and nondiscriminatory” (RAND) royalty rate for patents essential to industry standards. To some observers, the dense opinion (captioned “Findings of Fact and Conclusion of Law”) may be nothing more than another bit of procedural arcana in the interminable litigation over smart phone patents (summarized here), this time in the battle between Microsoft and Motorola (now owned by Google). But for followers of industry standards, Judge Robart’s opinion was a highly-anticipated and desperately-needed attempt to establish basic guidelines for the interpretation of the RAND licensing commitments that pervade industry standardization bodies.

Patents and RAND Licensing

As I discussed here last year, so-called “interoperability standards” enable products and services offered by different vendors to work together invisibly to the consumer (e.g., WiFi, USB, and the pervasive 3G and 4G telecommunications standards). Once standards are broadly adopted, markets can become “locked-in” and switching to a different technology can be prohibitively costly. Because patent holders have the potential to block others from deploying technology covered by their patents after lock-in occurs, the industry associations that develop standards (“standards development organizations” or “SDOs”) often require that companies participating in standards-development license patents that are essential to the standard to others on terms that are “reasonable and non-discriminatory” (RAND, also known as “FRAND” when the equally ambiguous term “fair” is added to the mix).

Microsoft, Motorola and the RAND Wars

The current litigation between Microsoft and Motorola relates to two common industry standards, the H.264 video coding standard developed at the International Telecommunications Union (ITU) and the 802.11 “Wi-Fi” standard developed at the IEEE Standards Association (IEEE). These standards are used in thousands of products on the market today, including Microsoft’s popular X-Box 360 game consoles, personal computers running the Windows operating system and a variety of smart phones. According to the pleadings, in 2010 Motorola offered to license Microsoft its patents essential to the implementation of these standards. A disagreement arose, however, over the royalty calculation that Motorola proposed. Under the rules of ITU and IEEE, royalties for patents covering H.264 and 802.11 must comply with RAND requirements. According to Microsoft, however, Motorola’s initial royalty demands were anything but “reasonable” and would have resulted in royalty payments in excess of $4 billion per year. Microsoft responded by suing Motorola for breach of contract in the Western District of Washington. The crux of Microsoft’s complaint is that Motorola reneged on its RAND commitments by offering a royalty rate that was manifestly unreasonable. Last October, Judge Robart ruled in the case that the applicable RAND royalty rate must be determined before a finding can be made regarding Motorola’s alleged breach of contract. A bench trial was held in November, 2012, and Judge Robart’s Findings of Fact and Conclusions of Law were released on April 25, 2013.

Defining RAND

Judge Robart’s opinion is important, not only because it resolves several highly contentious issues between Microsoft and Motorola, but because if provides a more general framework for analyzing RAND disputes in the future. At its heart, the bulk of Judge Robart’s opinion is a fairly conventional Georgia-Pacific analysis of the “reasonable royalty” rates applicable to Motorola’s patents. He spends a considerable amount of time analyzing comparable licensing transactions and determining their applicability to a hypothetical licensing negotiation between the parties. But Judge Robart makes significant modifications to the traditional Georgia-Pacific analysis in order to adapt it to the assessment of RAND royalty rates (which are related to, but different than, the “reasonable royalties” that serve as a measure of damages in patent infringement suits) (Para. 87). Here are some of the important observations that Judge Robart makes in this regard:

1.    Broad Industry Benefits from Standardization – Judge Robarts is explicit about the potential benefits that standards can confer on the overall economy through increased production and price competition (Para. 12). He observes that a primary goal of SDOs is the widespread adoption of standards “because the interoperability benefits of standards depend on broad implementation” (Para. 13). Likewise, RAND commitments are intended to “encourage widespread adoption” of standards (Para. 51, 70). Judge Robart recognizes the public benefit of standards and the public interest in ensuring that royalty rates for standardized technology enable broad implementation. Thus, unlike most patent licensing negotiations, the licensing of standards-essential patents takes on a public character. It is not merely a closed-door negotiation between two private parties. It must be conducted, and reviewed, with these public benefits in mind.

2.    Royalty Stacking – For years, commentators have observed that the aggregation of royalty demands by multiple patent holders can result in significant and unsupportable royalty burdens on standardized products. This is the problem of “royalty stacking”, which is very real in the case of H.264 (2,500 essential patents held by 35 U.S. entities, plus 19 entities with unknown numbers of patents) and 802.11 (developed by more than 1,000 companies). Throughout his opinion Judge Robart notes the threat of royalty stacking, and insists that any RAND royalty payable to Motorola must take into account royalties payable to other holders of patents covering the standards in question (Paras. 72, 92, 112, 456). This observation is significant, as it confirms that RAND royalties cannot be computed “in a vacuum” on the basis of isolated bilateral negotiations between licensor and licensee, without regard to the broader industry context in which such negotiations take place.

3.    Relative Value of Patented Technology – Judge Robart confirms yet another point that has been advanced by commentators for years: that the royalty associated with a particular patented technology should be commensurate with the actual value that technology adds to the overall standard and to the product in which it is implemented (Paras. 80, 104). Undertaking this lengthy and complex analysis, he determines that, by and large, Motorola’s patented technology added relatively little to either the H.264 or 802.11 standards or to Microsoft’s products (e.g., Paras. 289, 299, 384, 394, 457, etc.)

RAND-arithmetic

The bulk of Judge Robart’s opinion is devoted to a detailed, patent-by-patent, standard-by-standard analysis of the relative contribution made by Motorola’s technology, followed by an equally detailed Georgia-Pacific style analysis of licensing “comparables” a derivation of the RAND royalty rates for each standard. In the case of H.264, the primary comparable is the MPEG-LA H.264 patent pool. Judge Robart constructs a hypothetical negotiation between Microsoft and Motorola, assuming that Motorola’s patents are included in the pool and then calculating the royalties that Motorola would have earned from Microsoft as a result (¢0.185/unit). He then doubles this figure and adds it to the royalty base to account for the hypothetical license benefit that Motorola would have gained from being a member of the pool, yielding a RAND royalty rate for Motorola’s H.264 patents of ¢0.555/unit. In the case of 802.11, Judge Robarts finds the Via Licensing 802.11 pool to be a less relevant comparable than the MPEG-LA H.264 pool. He thus uses as his primary 802.11 comparable a royalty rate determined in 2003 by industry analyst InteCap, which he then divides by 25 to reflect the small value actually attributable to Motorola’s patented technology. The RAND royalty rate for Motorola’s 802.11 patents is thus set at ¢3.471/unit.

Both of these rates are significantly lower than the ones Motorola urged. As reported in the press, Microsoft calculates that on the basis of these rates, it would owe Motorola approximately $1.8 million per year, as opposed to Motorola’s original 2010 demand of approximately $4 billion, or the annual amount it requested at trial, which was approximately $400 million.

Judge Robart also calculated “ranges” of RAND royalties for Motorola’s patented contributions to each standard, presumably to help establish whether Motorola’s initial offers to Microsoft constituted “good faith” offers for purposes of Microsoft’s breach of contract claims. In both cases, the computed RAND royalties fall at or near the low end of these ranges (¢0.555 to ¢16.389 for H.264 and ¢0.8 to ¢19.5 for 802.11).

Significance and What the Future Holds

Judge Robart’s April 25 opinion is important to the standards world primarily because it sets out, for the first time, a logical and consistent methodology for computing a RAND royalty. This methodology is based on a conventional Georgia-Pacific patent royalty analysis, as modified to give substantial weight to royalty stacking, relative value and public interest considerations. The opinion will thus be useful not only to other courts considering RAND issues, but also, and perhaps more importantly, to arbitrators, mediators and private parties seeking to adjudicate RAND disputes before litigation commences. Given the likely increase in arbitration of standards-essential patent disputes (prompted by the recent nudge in this direction by the FTC’s Consent Order with Google), such guidance can only be helpful for the industry.

This being said, it is not clear that Judge Robart’s methodology offers the optimal means for resolving disputes over RAND royalties. It is, at best, complex and time consuming. At worst, it may be criticized as somewhat arbitrary. It may also be less useful for standards that are less broadly-adopted than H.264 and 802.11. With these standards, Judge Robart was able to choose from several different “comparables” to develop RAND royalty rates and ranges. Most importantly, he found comparables that were not merely bilaterally negotiated licenses between private parties, which he acknowledges result in royalty rates higher than those of patent pools. Most other standards do not have so many non-bilateral comparables from which to choose, if they have any at all. Thus, RAND rates may not universally drop, as they did in this case. Finally, it may be quite difficult to assess the “value” of particular patented technology in a standard if it is not functionally discrete, as Motorola’s contributions to H.264 and 802.11 seem to have been (e.g., Motorola’s contribution to H.264 appears to have consisted primarily of “interlaced video” technology that is now relatively obsolete).

For these reasons, other proposals regarding the determination of RAND royalty rates may still be worth considering, particularly as this latest decision in Microsoft v. Motorola begins what is likely to be a lengthy and interesting appeals process.

27 thoughts on “SO THAT’S WHAT “RAND” MEANS?: A Brief Report on the Findings of Fact and Conclusions of Law in Microsoft v. Motorola

  1. 25

    IANAE, Huh?

    The whole point is to force the prior agreement in order to obtain a standard in the first place. If the industry goes forward KNOWING there is a standard essential patent out there, unlicensed, they cannot adopt the standard.

    As I said, there are numbers of good examples where firms have actually cooperated to form a standard, licensed patents, and shared in licensing revenue according to their contribution. It is not that hard. Not at all.

  2. 24

    There are examples galore of companies sharing patents to adopt a standard. The royalties are agreed to in advance.

    Could you provide one or two of these examples galore, where royalties were agreed to before a standard was adopted?

  3. 23

    Item 2 Will force 1 to complete prior to the adoption.

    Why would you, as a patentee, complete item 1 prior to adoption, when the alternative is getting a standard-essential patent that entitles you to injunctions against everybody else in the market?

    It’s amazing what lengths people will go to, to avoid any situation where they’re expected to come to agreements on reasonable terms.

  4. 22

    Matt, two points:

    1. Negotiations have to take place before a standard is adopted so that the adopters know what the royalty is. Those wishing to adopt the standard would simply have to pay the stated royalty.

    2. If the standard is adopted and the negotiations for a FRAND royalty not finalized, the best solution is that the patent is not encumbered by FRAND obligations. In other words, an injunction would be proper.

    Item 2 Will force 1 to complete prior to the adoption. This is only right because the automatic license makes the standard adoptable in the first place.

    There are examples galore of companies sharing patents to adopt a standard. The royalties are agreed to in advance. FRAND, in contrast, is a train wreck because no one will ever agree on what a FRAND royalty is, even despite this case, forcing the matters into court. The only people who prosper from this are the lawyers.

  5. 21

    Robart’s clear and detailed opinion, in my view, gives a lot of guidance for determining the relative value of the invention (which, I agree, ends up being the baseline consideration for valuing the SEP).

    First, he looks at the value of the patent to the standard. He uses several different tools to evaluate this, including what areas of the standard it touches or concerns; the relative importance of that area of the standard to implementation; the availability of alternatives that would have permitted the standard; the amount of work done by the invention in a hypothetical implementation of the standard; and the likelihood that a variant of the standard involving the SEP would be implemented in the general case.

    Second, he looks at the value of the invention itself, not the standard, in the particular implementation at issue. Although he gets at it in a few different ways, the basic outline of the analysis appears designed to answer “ignoring that it would make it standards non-compliant, how would eliminating the invention affect the demand (or equivalently, the market-clearing price) of the implementation?”.

    Those two rubrics, one with detailed and useful subparts, will guide FRAND negotiations.

  6. 19

    Standards bodies need to look at compulsory copyright licenses as perhaps a starting point or an possible end point.

    For example, standards bodies may require companies to identify patents and applications and to agree on relative value before a standard is approved. Then, anyone wanting to practice the standard could take out a license at a fixed annual fee that would gradually increase to a full rate after a couple of years to allow the royalty to match business revenue. The fixed fee really is necessary if one thinks about it for awhile.

    The collected royalties could then be shared according the value contributed, less expenses. If anyone was caught infringing without taking out a license, they could be sued by the standards body in the name of the contributing patent holders. The only remedy authorized would for the collection of past-due royalties, plus the costs of enforcement, plus increased damages for willful infringement in a proper case.

    If a standard-necessary patent or application is not identified, but necessary, it becomes unenforceable against the standard by agreement.

  7. 17

    form [oldstep]+[newthought]

    LOL – you mean that claptrap that no one else anywhere has ascribed to (unless you count your svckpuppets like Francis), and that has not and cannot be correct because you do not (have not, and cannot) support the fact that claims completely of [oldstep] are still very much eligible – with or without [newthought] AND you have not (YET? EVER?) provided and acknowledged the controlling state of law containing mental steps.

    LOL right back atcha.

    9-0? LOL – you dance once again at your own self-defeat.

    Keep the hits coming? You keep on doing it to yourself.

  8. 16

    Your lack of ability to notice, let alone appreciate the nuances in patent law

    Nuances? Like the ineligibility of claims in the form [oldstep]+[newthought]? LOL.

    9-0. Still loving it. Keep the hits coming, anon.

  9. 15

    Malcolm,

    Who is “our”?

    Do you have a clue as to what the issue is, besides throwing out another vacuous soundbyte?

    Your lack of ability to notice, let alone appreciate the nuances in patent law is what is “Classic.”

    Way to underwhelm yet again.

  10. 14

    FRAND is a twisted concept. One needs to be “sick” or dishonest to accept something like that. We must keep on fighting this FRAND ridicule.

    We will fight to the last man against anybody who wants us to be reasonable.

    Sounds reasonable to me.

  11. 13

    I find #2 more problematic from an artificial forced constraint point of view. The forced assumption that the sum of the parts cannot be greater than the whole has an unthinking appeal. But it is a false appeal – an invention plays on the merits of itself – not such extrinsics as what else may also be present in a device.

    Our favorite patent t–b-gger complaining about “artificial forced constraints.” Classic.

  12. 12

    Whenever Laws enable ambiguity, legal decisions will take time, effort and many, many pages. Energy gets spent in semantics, not in engineering, and the result may be rather disappointing.

    It’s like when a project saves on planning and the energy that is “saved” by not planning is spent later, times 10, in mitigation of problems.

    FRAND is a twisted concept. One needs to be “sick” or dishonest to accept something like that. We must keep on fighting this FRAND ridicule.

  13. 10

    What should parties expect when they enter into agreements in which the most important provision of the contract is as clearly ambiguous and undefined as “reasonable and nondiscriminatory” (RAND) [or FRAND]? Absent a compulsory arbitration provision, or an impartial industry rate setting organization, their RAND agreements have “passed the buck” to some unknown future judge, and they must face the consequences of what they created, including wasting a lot of money on litigation like this.

  14. 8

    But what about this being standard essential? I agree with your #2 point in the broader sense, but I am assuming that they are saying that you need the patent to implement the standard, which I think changes the analysis since FRAN forces a license.

  15. 7

    That’s my point: you easily swallow the false “commonsense” appeal without thinking about what that actually does.

    It’s an artificial constraint. One that escapes critical thinking because of the ‘commonsense’ notion that a whole cannot be any more than a sum of its parts.

    Such should not be common, as it does not make any true sense. Not if you understand patents.

  16. 6

    I think you are right in the sense that it tends to force #3 into a bargaining position. But, there is a commonsense appeal to #2 that if the product goes for x dollars and it is practicing 1000 patents, then they all couldn’t be worth so much to the market value of the product. That is true, but number 3 gets to the relative value. 999 could be worth nearly nothing with 1 could be worth 10 percent royalty.

    The bigger point of this is that again we see the tough question of what is that worth to the product being buried around layers of stuff.

  17. 5

    I find #2 more problematic from an artificial forced constraint point of view.

    The forced assumption that the sum of the parts cannot be greater than the whole has an unthinking appeal.

    But it is a false appeal – an invention plays on the merits of itself – not such extrinsics as what else may also be present in a device. If a device cannot ‘afford’ to have some invention in it – than use something else or invent your own replacement – the ‘block’ here is a feature, not a bug.

  18. 4

    And, then the judge hides behind the mountain of impartial and usually fair process. But, let’s be real. The one finding of fact in number 3 is the whole ballgame.

    This is part of the zeitgeist where people in power think they can just create a large analysis where 90 percent is built on fair and impartial reasoning and the 10 percent is where they do what they want and slice someone in the back (and often at the state level take some money.) If you read 1984, you see this type of control.

    Bizarre world we live in. I am sure that this opinion helps, but let’s be real…number 3 is still where all the action is and not much from what I could see was developed in this area.

  19. 3

    As in so much of law (which is a depressing topic), there is always this illusion of impartiality and process. But, as in most of law, almost all of the important part is done in 3 and the rest of this is just common sense adjustments. So, the important bit: how much is that patent worth really wasn’t done. And what scares me about this sort of opinion and (which is true in almost all of law) is that one of those findings of fact is usually the whole ballgame and the judges build mountains of process and fairly impartial analysis on top of the one finding of fact that makes all the difference.

  20. 2

    Seems like the devil is No. 3 as it is evaluating how much the patent is worth to the technology. So, in some ways No. 3 is probably the starting point and all the rest is merely an adjustment to no. 3. And, did anything he do help to understand how to evaluate no. 3?

Comments are closed.