Sunday, May 13, 2012

Oracle and Google make competing proposals for determining copyright damages

On Saturday afternoon/evening local time, Oracle and Google filed different motions relating to how the court can efficiently arrive at a determination of copyright damages including the disgorgement of infringer's profits. As a consequence of the jury's failure to decide on "fair use" and the judge's refusal to fill out the resulting gap (and his decision to postpone the answer to the question of copyrightability), this is somewhat complex. It won't be trivial for reporters to get the implications of those proposals right, especially since many of them got confused by the related $150,000 story that I'll also address further below.

I'll start with a summary of the parties' proposals and their pro's and con's, followed by further detail on some of the terms and issues involved.

Google wants dismissal of certain disgorgement claims -- Oracle proposes to clarify all liability issues before talking remedies

The basis for all of this is that copyright will either be the bigger issue at the remedies-related (damages, disgorgement, willfulness) stage of the trial (Phase Three) than patents, or it will be the only one -- the parties may very well be able to settle patent damages quite easily after any liability finding since they are no longer too far apart from each other on those. The court would like to eliminate or narrow Phase Three.

Google wants the court to throw out some of Oracle's claims for disgorgement of copyright infringer's profits but would potentially leave the most important set of issues (API liability and related damages and disgorgement) unanswered at the end of this ongoing trial. Each party has yet to comment on the other's proposal. Oracle will oppose Google's motion by 10 PM local time on Sunday.

Oracle advocates a postponement of the damages-related third trial phase until after another jury has decided on Google's "fair use" defense concerning the 37 APIs, an issue on which the current jury wasn't able to reach unanimity but which must be resolved one way or the other, sooner or later. It's unknown how Google will respond to Oracle's proposal, and Google would benefit from the requested dismissal of claims regardless of whether Oracle's proposal is adopted or rejected by the court, but chances are that Google will oppose since the parties would otherwise have filed a stipulation. The proposals are incompatible in the sense that Google wants decisions with respect to parts of Oracle's copyright infringement claim, while Oracle insist that it brought one claim that needs to be evaluated as one.

For the court, Google's proposal may be tempting in some ways but it would not be a good choice if adopted in addition to Oracle's proposal, and it would even be a clearly bad choice if the court considered it a substitute:

  • Oracle's proposal is just about case management, not the parties' rights, and would, therefore, be in the discretion of the court. Realistically, there's no way Google could successfully appeal such a scheduling-related decision. By contrast, Google's proposal is aggressive. Given that the related copyrighted works have already been considered more than de minimis and that Oracle brought one coherent copyright claim (which should be evaluated at the remedies stage as one) as opposed to several disjointed copyright claims, Oracle simply has certain rights that I believe it can't just be deprived of. I doubt that the potential benefits of Google's proposal, which are limited, outweigh the downside.

  • Wholly apart from the legal issues it raises, Google's proposal is also inefficient in the mid term since it would only take care of relatively small issues without proposing a way forward for the far more important API structure, sequence and organization (SSO) issues.

    Judge Alsup will rule for or against the copyrightability of the 37 asserted APIs:

    • If the judge deems the API SSO copyrightable, a new trial for "fair use" of the API SSO will have to take place: the court and both parties agree on this much. And if Oracle overcomes the "fair use" defense, remedies will have to be determined.

    • If the judge doesn't deem the API SSO copyrightable, I personally believe that Oracle will nevertheless pursue this matter (rather than settle on sweet terms for Google) through an appeal, and if the appeal succeeds, there will have to be a new jury trial. The new trial would have to address fair use unless the appeals court determined in the meantime that Oracle was entitled to a judgment as a matter of law against that Google defense. At the very least, the new trial would have to address copyright damages (including disgorgement) involving the API SSO.

  • Oracle has a point when it says that this jury shouldn't try copyright remedies. Oracle's argument is that the court overruled the jury with respect to eight decompiled files. If the jury then has to assume at the remedies stage that there was an infringement of those eight files, it will still be influenced by its own unanimous (but erroneous and overturned) decision that there wasn't one. Under Oracle's proposal, a new jury would no longer have to look at infringement but simply be instructed that certain items have been found infringed. The current jury was overruled by Judge Alsup on only one item, and on that one, the jury found in Google's favor although Judge Alsup argued that no reasonable jury could have arrived at that conclusion.

    It's not too unlikely that the same mistake that led to the erroneous verdict on eight decompiled files, and/or a related mistake, prevented the jury from unanimously dismissing Google's "fair use" fairy tale.

If one looks beyond the current trial and considers the possible impact of appeals and retrials, Oracle's proposal is the logical way forward, while the adoption of Google's proposal would only save court resources if Judge Alsup also held the API SSO non-copyrightable and if, subsequently, Oracle either offered Google a cheap settlement (which won't happen, at least not on that basis and at this stage) or lost all appeals. In any scenario outside of that set of permutations, Oracle's proposal would streamline and simplify the process, while Google's would serve to stall and complicate.

The question is now what the court's calculus will look like. Here are my thoughts on that:

  • Judge Alsup was very dismissive ("height of ridiculousness") of Oracle's insistence on disgorgement in connection with the nine-line rangeCheck function and eight decompiled code files. But if Oracle subsequently prevailed on the API SSO issue, then it would actually come in handy to have evaluated the disgorgement issue (Google needs to prove deductions from its Android-related revenue). In a recent brief, Oracle said it would be fine with letting the judge handle disgorgement without the jury (but Google would have to consent to this).

    If Oracle convinced the appeals court that its copyright claim must be evaluated as one claim at the remedies stage, then there's also a risk of significant or substantial duplicative effort. Ultimately Judge Alsup may just decide based on what Oracle's rights are, as opposed to what he considers ridiculous or not.

  • In the past, Judge Alsup tried all sorts of things to get Oracle to soften its stance. None of them succeeded. Oracle was willing to drop patents of then-uncertain validity in order to get a quick trial, but it kept insisting on Android becoming fully compatible with Java (I conclude that from various pleadings in the public record as well as a statement Oracle's lawyers made to reporters shortly before the trial started). Oracle can afford to bide its time and go for the big strategic win. There isn't much room for bargaining. Judge Alsup has a lot of influence over the case, but he's dealing with two parties that appear totally prepared to appeal (one of them already has).

  • The judge may assume that Oracle has only a limited incentive to appeal a dismissal of its smaller disgorgement claims in case of a finding of copyrightability of the API SSO. But Oracle's latest brief stresses the concept of one copyright claim -- it looks almost like Oracle's equivalent of the One-China Policy.

  • Oracle's proposal would help to clean up the mess that this case currently is. There would be much less of a mess if the jury had been able to decide unanimously on the "fair use" defense, but things would also be much simpler if the judge had thrown out the "fair use" defense as a matter of law, which he could have -- and in my personal view should have. The problem this case is facing is that it's like a mathematical equation with too many unknowns. Oracle proposes a step-by-step approach to answer all of the outstanding questions. Uncertainty didn't lead to a settlement. Clarity will. In the end.

    I hope Judge Alsup will choose the most efficient path to an adjudication of all of the issues over further attempts to get the parties to settle in the days or weeks ahead. With adjudication happening or being near, they will settle (maybe not before an appeal, but at some point). By now there's empirical evidence that game theory doesn't apply to this case, at least not with respect to the really important issues.

The $150,000 fallacy

I regret to say that there have been very misleading headlines even by journalists who usually don't propagate spin. Forget any stories about this being merely a $150,000 issue.

Yes, Judge Alsup said something that appeared to validate the spin that Google and its astroturfer friends told the press, but the judge simply did not express himself as accurately from the bench as he would in writing. He made a statement that related only to a subset of the possible scenarios and outcomes. Let's face it: none of us is equally precise irrespectively of circumstances and method of communication, but most of us don't speak out authoritatively on issues where incomplete representations correspond to billion-dollar differences. Also, don't forget that this is the very same judge who has previously made several attempts to get Oracle to soften its stance, including that he obligated the companies' CEOs to meet, but Oracle stands firm. Maybe this was another try to pressure Oracle. Whatever the reason or combination of reasons was, everyone, including the journalists who picked up and amplified that misrepresentation, would have gone home already if this case was about nothing more than chump change for companies of this size.

The most fundamental misconception is that some still think, or claim, that this is a case in which damages are the most important part. They are not. This is about Oracle's stated goal "to bring Android back into the Java fold" more than anything else. The path to that is a technically impactful injunction. The infringement findings are already broad enough to serve as a basis for such an injunction, and everything that's missing could easily be resolved as a matter of law (Judge Alsup will rule on copyrightability and equity, and even though he decided Google's "fair use" tale can't be dismissed as a matter of law, the appeals court may very well conclude that it can and, in fact, should). Wholly apart from injunctive relief, even if this were all about damages, the $150,000 story would be off-base -- in fact, Oracle's proposal for the way forward makes this very clear, while Google, being Google, obviously sticks to its spin.

For the remedies/damages phase of the trial (Phase Three), copyright is the more controversial part than patents because of a concept called disgorgement of copyright infringer's profits, which could result in a disagreement between the parties of (theoretically) up to $700 million, while the court has limited Oracle's patent damages demand to a level at which the parties may indeed be able to settle that part of the dispute once liability has been established. I guess it would be easy for them to arrive at the right number -- the most important inhibition may be that Google won't like the notion of admitting liability for infringement.

The $150,000 figure was based on the assumption of Oracle going for statutory damages (for an explanation of the term, check out the next section) and only related to a remedies phase of the current trial prior to resolving, one way or the other, the "fair use" issue. As Oracle's proposal shows, the logical approach would be to clarify all liability matters, including the "fair use" thing, before even going into the remedies phase. A hung jury doesn't mean that Google has proven its "fair use" defense. It means that this has to be resolved later (except if copyrightability is denied by Judge Alsup and the appeals courts). The $150,000 idea was the notion of making the third step (damages and other remedies) before the first step (copyright liability) has been completed, with the possibility of completing the first step later and then sort of going back to the third step. That would be chaos. This litigation needs structure. It needs it badly, for the sake of an efficient process and equitable outcome -- and as a side effect, it would make things easier for those reporting on the case, but that's less than secondary.

Terminology: damages; disgorgement; statutory damages

Judge William Alsup and the jury have identified a plurality of copyright infringements (the state of affairs is specified further below), and their collective findings entitle Oracle to damages including disgorgement of infringer's profits.

In colloquial use and even in many legal documents, disgorgement is often considered to be another one of several categories of damages, though in some contexts the term "(total) damages" is defined more narrowly, i.e. limited only to what the right holder actually lost, thus not meant to include disgorgement -- a potentially confusing inconsistence in usage (judges and lawyers appear to know what "damages" means based on context, just like the ordinary person will usually know when a "seal" is an animal or a stamp). You can assume that most of the time people include disgorgement when they talk about copyright "damages", simply since there's no commonly-used word that comprises the two concepts (one could refer to them collectively as "monetary remedies", maybe).

Another important term is "statutory damages". Under U.S. law, a copyright holder who prevailed on a liability claim has the choice between seeking statutory damages or actual damages. Statutory damages are sort of "flat-rate" damages, though there is a range, with an upper end of $150,000 per item. It's easier for the court to decide on statutory damages without involving a jury. And between these parties, those statutory damages are unimportant.

Oracle has made clear that it's not interested in statutory damages, not even with respect to smaller items. It seeks actual damages as well as disgorgement. And further above I talked about Oracle's position that it brought one copyright claim. One that needs to be evaluated as one at the remedies stage.

The plurality of infringements identified

  1. by partial jury verdict: the structure, sequence and organization of 37 Java APIs

    The essential questions of copyrightability, which the judge reserved for a later stage, and the "fair use" defense, which has neither been proven by Google nor overcome by Oracle due to a jury impasse, also have to be answered in Oracle's favor in order for the infringement finding to matter. Google argues that a retrial should also involve a new verdict on infringement (obviously, Google is highly uncomfortable with the current state, regardless of what its official statements indicate) while Oracle advocates efficiency. The key thing to understand is that, figuratively speaking, the jury is still out on this. Literally, this jury is no longer out. And Judge Alsup didn't want to decide either. But there will either be a new jury or an appeals court or a judgment as a matter of law at the level of the district court, or any sequence or combination of some or all of this.

  2. by partial jury verdict: the nine-line rangeCheck function, which is small but executed all the time (2,600 times when an Android device is switched on)

  3. by judgment as a matter of law (judge overruled jury): eight decompiled files

    This blog was first to talk publicly about six of them and to show tables that juxtaposed the infringing code to the original code. The jury verdict on these files was negative. That part of the verdict was inexplicable. There can be no reasonable doubt about copying, and testimony was also very clear. The judge has fixed this problem.

It's always hard to predict jury decisions, and this jury has now already been overruled once because of an exceedingly defendant-friendly position, but I think there's a reasonable chance that the jury will also conclude that Google infringed Oracle's asserted patents. We will know soon. Hopefully the patent phase of the trial won't complicate things even further.

If you'd like to be updated on the smartphone patent disputes and other intellectual property matters I cover, please subscribe to my RSS feed (in the right-hand column) and/or follow me on Twitter @FOSSpatents and Google+.

Share with other professionals via LinkedIn: