Prof. Doug Lichtman, UCLA School of Law
February 6, 2012

A few weeks ago, I wrote a post about the RAND commitment, explaining how technology firms often license patents not by explicitly agreeing as to price, but instead by more vaguely agreeing to license their patents on “reasonable” and “non-discriminatory” terms.  As I explained in that post, this practice typically comes up in the context of standard-setting, where thousands of patents might be at issue, and hundreds of firms might be involved in the technical and economic negotiation.

I focused last time on a discussion of why firms might choose RAND over more explicit forms of pricing.  This time, I want to think about how the RAND commitment should be interpreted in the event of litigation.

From an economic perspective, the purpose of the RAND commitment is to ensure that patent holders are not able to earn exaggerated royalties merely because price negotiations have been delayed.  Without some sort of pricing commitment, this is exactly what would happen.  The price for any technology included in the standard would go up simply because it was chosen.  And that is emphatically not the point of RAND.  Standard-setting participants defer pricing negotiations because they want more information, or because they want to implement the relevant standard more quickly, or because they want to minimize upfront costs.  But it seems implausible to think that standard-setting participants opt for RAND in order to randomly and artificially increase each patent holder’s ultimate leverage.

To see this point more fully, consider a situation in which two comparable technologies are vying for inclusion in a given standard: Dolby’s high-fidelity audio compression codec on the one hand and DTS’s rival audio compression technology on the other.  Were prices being negotiated at the time of the selection, participants in the standard-setting process would compare the Dolby and DTS approaches.  They would identify advantages and disadvantages, and they would ultimately offer the winner a price that reflected its marginal value as compared to the unsuccessful alternative.  If the winning patent holder were to hold out for more, standard-setting participants would presumably threaten to switch to the second-best technology.  Ultimately, a competitive bidding process would typically yield something close to the efficient price.

Now consider what would happen if, instead of negotiating at the time of selection, standard-setting participants were to wait and negotiate a few years later.  Two important considerations would by then have changed.  First, a given licensee would by that point likely have made some technology-specific investments.  The firm will have designed its products.  It will have built manufacturing facilities.  It will have made commitments to buy components from its suppliers.  And it will have promised relevant functionality to downstream customers.  A patent holder would be able to take advantage of all of those commitments, demanding a royalty that reflected not only the value of the patented technology as compared to next-best alternatives, but also the value that this licensee would place on avoiding disruptions to its already-made investments.

Second, even if a particular would-be licensee has not made patent-specific investments, its peers will have – and that triggers a similar dynamic.  Consider standards with respect to driving.  Before driving norms were established, the value of “driving on the left” was roughly equal to the value of “driving on the right.”  Everyone surely agreed that all the drivers in any particular region ought to adopt the same default rule; but the choice between the two was likely a draw, and thus patents related to either one would have been of similar value.  Once a great deal of traffic had opted for the right, however, the economics changed.  A patent related to the idea of driving on the left was worth very little.  A patent related to the idea of driving on the right was worth a fortune.  The change had nothing to do with the relative merits of these two technologies.  It was just an example of a more general phenomenon associated with standardization: Patents related to a chosen standard increase in value as more people adopt the standard.  This is then another problem with delayed patent negotiations.  Patent holders who negotiate after standardization are able to charge prices that reflect the now-realized network value, in essence charging licensees for the fact that other licensees are already committed.

From an economic perspective, then, the important work of the RAND commitment is to minimize these economic distortions.  Participants in the standard-setting process might well intend to pay patent holders the royalties they would have gotten had sufficient time, cash, and predictive information been available so as to enable complete negotiation ex ante.  And participants in the standard-setting process might also intend to pay each patent holder a bit extra as thanks for that patent holder’s willingness to delay the price negotiation and in that way reduce the costs associated with the standard-setting process.  But there is no reason to believe that standard-setting participants also mean to allow patent holders to hold hostage each participant’s standard-specific investments, or to allow patent holders to arrogate to themselves the value created from the group’s standardization effort.