www.sciencemag.org SCIENCE VOL 343 28 FEBRUARY 2014 973
Our theoretical analysis (3) looks at consequences of this modified standard-setting process. Initially, it might be thought that this process, too, might lead to a variety of opportunistic behaviors. For instance, a patent-holder
might set a high price cap in order to achieve
a stronger bargaining stance when it comes to
dividing profits from the eventual patent pool
formed around the standard. Such behavior
might lead to excessively high licensing rates
and too little technology adoption. Despite
the possibility of such gaming, our analysis shows that structured price commitments
achieve the desired “competitive benchmark,” with license prices the same as would
prevail if there were no network externalities and, thus, no need for a standard, which
would free users to adopt whatever mixture of
technology they wished. This provides a reasonable balance between incentives for technology innovation and diffusion.
How can price caps selected before standard-setting work? Intuitively, when patents
are fairly interchangeable with each other
before selection of the standard, the requirement to commit to a cap forces patent owners to choose a lower price than they would
afterwards, so as to get their patent included
in the standard. Otherwise, the SSO would
choose a more affordable option. And when
patents are complementary rather than interchangeable, setting one’s price away from its
competitive level leads to a loss of profit: For
instance, if a patent owner opts for a low price
cap, other owners will be less willing to give
him or her a large share of the proceeds of any
patent pool that may form.
Such commitments lower the amount that
ultimately will need to be paid to license the
key patents to implement the standard, and
also better match the rewards to innovations
(royalties) with their contributions.
We do not suggest abandoning FRAND.
SSOs may make mistakes, and important
patents may escape attention at the time the
standard is promulgated. However vague and
limited, FRAND still imposes some ex post
constraints on licensing terms. Thus, we view
structured price commitments as a comple-
ment, not a substitute, to FRAND.
Can Regulation Work?
The idea that patent owners might commit
to prices before standard-setting is not new;
both academics and policy-makers [e.g., (4)]
have mentioned that possibility. Furthermore,
this policy has been tested by some SSOs (see
text box). Yet there has been little adoption by
SSOs in general.
One issue is that economists had no
framework for thinking about price commitments. There was no way to predict, even at a
theoretical level, what would happen if such
commitments were mandated. The theoretical analysis comforts us about the normative
impact of commitments (should we expect
a good outcome from this new institution?).
But thus far, it does not explain why SSOs
which, rightly or wrongly, tried them, failed
to appeal to IP owners and set a new “
standard” for the standard-setting process.
Consistent with historical examples, our
model highlights that patent owners prefer
to go to an SSO that does not constrain their
freedom to license patents as they see fit.
Competition between SSOs, which we term
“forum shopping,” allows IP owners to find
more complacent SSOs that do not put such
pressure on their prices. Owners of equally
important, standard-relevant patents will
always opt for a lenient (commitment-free)
SSO. Only in specific circumstances will patent owners be attracted by such requirements,
e.g., when owners of clearly essential patents
fear that an SSO will include a number of
more minor patents, with whose owners they
will have to share the pie equally.
The corollary is that without a govern-
ment mandate and/or purchasing prefer-
ences, SSOs are unlikely to opt for the “right
answer” and demand price commitments.
Nevertheless, the suggestion that price com-
mitments be mandated has problems: It
could be infeasible or, if feasible, the regula-
tion might do more harm than good.
Concerning feasibility, the rule can be
implemented where public procurement
is sizable, such as telecommunications. It
would suffice to require suppliers to conform to “compliant standards” enacted
by SSOs requiring price commitments.
Furthermore, antitrust authorities and courts,
already involved in regulating standard-setting through reviews and litigation, could
favor compliant standards and discourage
As for unintended consequences of regulation, the requirement is simply to commit to a price cap. If practice does not work
according to theory (only experimentation
can tell), the hazard is that IP owners will set
too high a cap. But it cannot be worse than
the current policy, which is formally an infinite price cap. The combination of a cap with
FRAND cannot be worse than only FRAND.
So this is rather low-risk.
In recent years, antitrust authorities
have wakened to the importance of standardization and the way in which firms can
manipulate this process (5). Now, the challenge is to create rules that will ensure that
standards can effectively promote translation of scientific and technological insights
into new products and economic growth.
Requiring price commitments is theoretically appealing, rather risk-free, and worth
References and Notes
1. Lemley and Shapiro (2) suggest that these disputes be
settled via arbitration, making adjudication more expedient, better informed, and overall more consistent.
2. M. A. Lemley, C. Shapiro, Berkeley Technol. Law J. 28,
3. J. Lerner, J. Tirole, “Standard-essential patents” (
Working paper no. 19664, National Bureau of Economic
Research, Cambridge, MA, 2013).
4. D. P. Majoras, “Recognizing the procompetitive potential
of royalty discussions in standard setting,” Standardization and the Law, Stanford University, 23 September
5. National Academy of Sciences, Patent Challenges for
Standard-Setting in the Global Economy: Lessons From
Information and Communications Technology (National
Academy Press, Washington, DC, 2013).
6. T.O. Barnett, “Business Review Letter: VITA,” Antitrust
Division, U. S. Department of Justice, 30 October 2006.
7. J. L. Contreras, Jurimetrics 53, 163 (2013).
Acknowledgment: J. T. is a member of Industrial Economics
Institute, which receives funds from Microsoft, Orange, and
The international trade association VITA, which develops standards that govern modular embedded
computer systems in such products as advanced avionics devices and complex medical instruments,
became frustrated when owners of patents that were essential to implementing one of their new standards
demanded a higher royalty rate than anticipated after being deemed standard-essential. As a result, VITA
proposed a new policy, which the U.S. Department of Justice’s Antitrust Division approved in October
2006 (6). VITA demanded that members declare the highest royalty rates and most restrictive terms under
which they would license these patents for implementation of the given standard. In response to this
controversial policy, Motorola launched a campaign to decertify VITA as an SSO recognized by the
American National Standards Institute, an effort that ultimately proved unsuccessful. Although VITA ended
up moving forward with its standardization process, it was not a spectacular success (7). Similarly, the
Institute of Electrical and Electronics Engineers and the European Telecommunications Standards Institute
encountered stiff resistance and, in the end, had to make price commitments optional. Even if price
commitment requirements lead to a good outcome, we cannot expect SSOs to mandate them by
themselves or expect patent holders to cooperate with standard bodies that mandate such commitments.
10.1126/science.1246439 P H O