Session 6: Standards
Moderator: Kirti Gupta
Jorge L. Contreras, “Private Ordering or Public Law? The Legal Character of Technical Standard Setting”: Private ordering structure arises when public enforcement mechanisms are unavailable (crime syndicates, rural settings) or less efficient/logical (credit rating, accounting standards). Standard setting: collaboration among competitors; technical focus, not legal; ideally the priority is on optimizing solutions—wi-fi, USB (my favorite USB joke). Most standard-setting bodies have formal policies about due process, and also about patents (disclosure requirement, FRAND licensing requirement—some bodies have one, some the other, some have both). Informal norms and practices also: tribal culture.
What did courts do when disclosure obligations were allegedly violated? Looked to informal norms of community. Rambus: written policy was quite vague and treated badly by Fed. Cir. (shockingly vague and imprecise) but still an affirmative obligation to disclose arising from practice; so too with Broadcom v. Qualcomm, where the written policy simply said parties are “encouraged” to disclose. FRAND: vague standard, but where challenged in litigation, looked at comparable licenses and patent pools to find reasonable range based on norms and practices in the industry. Private practice finds its way into private law disputes. That’s not unusual.
Federal recognition of standardization’s public character: DOJ/FTC report on IP from 2007 calls it one of the engines of th emodern economy. DOJ/PTO statement in 2013: voluntary consensus standards serve the public interest, fuel innovation. Debates over whether that should affect injunction against violator? There is a tendency to talk about the availability of injunctions in terms of public characteristics of standard-setting prices—harm to competition from allowing injunctions or exclusion orders on standard-essential patents in Apple v. Samsung (ITC). USTR says the same thing. That is public law.
But is that appropriate to regulate standard-setting under public law/antitrust framework? EU/US agencies seem to believe so, but dissenting voices say private law/contract enough, and overdeterrence of patent enforcement may chill innovation/SDO participation. Tentantive conclusion: public law frameworks should be applied w/caution after private ordering is permitted to address potential problems.
Commentator: Joseph P. Liu: Private ordering can mean different things to different people. More expansive understanding of private law might provide a way to respond.
Private ordering: at least 3 definitions. (1) Contracting around existing legal rules; ordering that results from private transactions. (2) The kinds of private dispute resolution systems identified by Ellickson (ranchers), diamond merchants, etc. With their own rules & customs, typically enforced through self-help. (3) Private promulgation of rules to govern internal organizations. Homeowners’ associations; sports league rules. Like (2) in that content is privately created, but unlike in that it’s more top down than bottom up and more reliant on enforcement from courts.
Paper places technical standard setting within (2) when it might be more like (3). Rules tend to be more top down. This might be important b/c if you put it in category (2), order w/o law, you get presumptive benefits of that category—superiority to judicial proceedings in some cases—when they aren’t warranted.
What turns on private ordering v. public law characterization? Paper sometimes suggests that if standard setting is private ordering it should be left to internal rules, whereas if it’s public then it should be subject to regulation via antitrust etc. Hard for me to see why degree of regulation should depend on characterization; private agreements are pervasively subject to public regulation. Real q: whether regulation is required to achieve the substantive policy goals we have—e.g., solving holdup problems—is antitrust or other regulation required to solve them?
Paper identifies issues arising from particular type of private behavior, coordinated activity by industry participants: ostensibly private/technical, but has a public character. Paper expresses doubt over whether antitrust/criminal law are the appropriate mechanisms. Consider third option, between private ordering and public law. Tools available in private law understood more broadly as the substantive doctrines of patent law. Paper discusses tort, contract and not property claims. Other private law doctrines like promissory estoppel might be relevant as well. Avoids the bind of choosing between “pure” private ordering or public law. Richer set of public interests into the considerations.
Janet Freilich & Jay P. Kesan, “Towards Patent Standardization”: We don’t have a good theory of what standardization would mean, if it’s desirable. What is standardization of patent content? We mean the most general sense: something that is agreed upon, whether it’s standard nomenclature or sections in the patent description. Beneficial for purposes of notice and disclosure, reducing search costs and fuzziness of patent boundaries; easier to read and identify information. Can improve databases, also helpful for disclosure.
How to make it happen? Congress could maybe try, but not going to happen in the real world. PTO can make rules but there are practical and political problems. Standardization could arise through voluntary measures—soft standardization could get us part of the way to the goal. WIPO standard for disclosure of nucleotide and amino acid sequences: PTO adopted it, using the rationales of quality and efficiency of examination—easier to compare w/prior art; conformity for scientific community, using language they’re used to; improved dissemination of information in electronic format. But most standardization in life sciences doesn’t come from regulations. Taxonomies, controlled vocabularies created by public/private institutions—new vocabularies are created for new technologies. Often fairly well-defined; FDA won’t let you use a drug name for something it doesn’t think the name encompasses.
In patent, you don’t have to use these standards; the patentee is her own lexicographer and can define a term differently from any standard. But outside the patent world, these standards usually do have to be used. Journals, for example, require authors to use standards. Thus patents can diverge from conversation in rest of life/science. So it would be easier to use the nomenclature in patents too. Case study: percent of granted patents that use celsius or fahrenheit—number using celsius is going up as a percentage of those who use either, simultaneous w/increase of use in celsius in the scientific community generally. 1988: Congress again said we should use metric, but we don’t have to; metric required in patent in 1995, but the trend in using celsius in patents started before that: a result of general scientific community convergence.
Possibilities for software: Representational languages: pseudocode; object-oriented languages; modeling languages—better comprehend the new/inventive features for which patent protection is sought in software. More useful and technically discernible software patent repository compared to the current problems in figuring out what prior art is. SSOs can encourage the detailed specifications needed. They discover the best technology/certification of the process. Use IPR policies that are fairly detailed, taking disclosure into account and penalties for noncompliance. They’re uniquely positioned to provide standardization guidance, especially since they’re dealing with patents that will be very valuable if the standard is accepted.
Templates: universities’ tech transfer offices could have templates, as Stanford and MIT do, as starting points. Possible problems: powerful interest groups dominate/exclude others; allow hidden patents that resurface at some point. Criticisms of standards: often that they are in fact reducing the scope of protection; we don’t think that’s the case. Standardized patents can be drafted broadly, and vague patents can be drafted narrowly.
Next: empirical studies to see if patents with standardized disclosures are likely to be more valuable.
Commentator: Michael J. Meurer: Like the data on the metric system, but it’s not clear it provides benefit in more disclosure or more notice/clarity. We need examples of success in disclosure/notice. Mendeleev: periodic table facilitated codification and spread of knowledge—Mokyr, von Hippel, Winter; Moser identified jump in patenting of chemicals after Mendeleev: codification, reverse engineering became easier and trade secret, which had been used, became less attractive. W/semiconductors, a de facto standard crept into the industry based on the TTL Data Book—might be a helpful examples.
Incentives of applicants to embrace standard language: Teaching/reduce transaction costs in licensing or assignment. Clarity of property rights/notice (borrowing ideas from literature on standard contract terms): good language might come from collective wisdom on the breadth I want/what design around opportunities are available; I might like the fact that courts have already interpreted what “anodize” means, but since courts interpret “a” differently in different patents that’s not such a successful enterprise. Network effects are associated w/standard terms of contracts and might arise here.
Reasons they won’t embrace standard language: if I want to practice exclusively and keep tech as secret as possible, I’d like the advantage in the race to generate follow-on innovation; we know many inventors get the benefit of both patent and trade secret. Notice: lots of patent applicants are looking for opportunistic profit, and they want to hide/obscure for ex post bargaining advantage. In re Kubin case: protein in a patent; examiner found prior art identifying the protein under another name; applicant will lose novelty once the link is found.
Private associations generating public goods, like standards: there is reason to be optimistic that private parties can do this. Caution: setup costs of codifying tacit knowledge—depends on market conditions and pace of technological change (if too fast, don’t bother codifying).
PTO maintains the acceptable ID of goods and services manual to classify trademarks, and the US Patent Classification to classify patents. Improving due to global cooperation; agents are capable of standardizing.
Why not private/public cooperation, as b/t PTO and Google; PTO could subsidize development of IP registry search tools. Probably the right result is a mix.
Kappos: Nautilus case provides impetus for standardization, because functional claiming is now a quick and dirty way to get an invalid patent.