67 Pages Posted: 5 Apr 2016
Date Written: April 4, 2016
This Article addresses a policy question that has challenged scholars and lawmakers since the 1850s: Do the transaction cost benefits of patent pools outweigh their potential for consumer harm? This question has special importance today. Patent pools are on the increase, due to large numbers of patents in critical industries such as software and mobile phones. In this Article, we present the first empirically-based estimate of the transaction costs savings engendered by patent pools. Drawing on interviews with administrators of prominent pools, we document the costs of assembling and administering a functioning pool. We then estimate the transaction costs that would result if the pool were never formed. This means estimating the costs of large-scale bilateral licensing of all patents included in the pool. We include an estimate, again based on empirical data, of the cost of occasional litigation when bilateral negotiations break down. Comparing the cost of running a pool with the counterfactual cost of licensing (plus probabilistic litigation) in the absence of a pool, we estimate empirically the transaction cost savings from pooling patents. The numbers are impressive: many pools save hundreds of millions of dollars in transaction costs.
Next, we tackle consumer welfare losses. Antitrust regulators and scholars identify two chief costs: lost substitutes and lowered incentives to invent improvements. Substitutes may be lost when a pool combines patents on two technologies that perform the same function. This has the potential to increase consumer prices. We present a method for estimating social welfare losses from combining substitutes. Through case studies, we apply this method to estimate the welfare losses (in dollars) caused by specific patent pools. We present a second method for estimating the consumer welfare losses represented by lowered incentives to innovate. The chief feature of pools that affects future incentives is the “grantback clause,” under which members agree to license future members into the pool. Again drawing from real-world case studies, we apply our method to estimate the potential losses in dollars that flow from grantbacks. This phase of the analysis draws on cutting-edge patent portfolio mapping techniques to estimate future lost substitutes due to grantback clauses.
When the welfare loss estimates are compared to the transaction cost savings, one arrives at a comprehensive methodology for evaluating patent pools. The systematic approach presented here allows a regulator to say, for example, that society could tolerate a certain number of lost substitutes, given the cost savings of a pool. And it allows a regulator to estimate the future substitutes, lost due to grantback provisions, that can be tolerated given the pool’s cost savings. Thus we present a comprehensive, reproducible, and rigorous framework for evaluating the net effects of any proposed patent pool.
This Article contributes two important “firsts” to the patent pooling debate: (1) We quantify the benefits (transaction cost savings) of patent pools, and (2) we quantify the consumer welfare costs from (a) lost substitutes, and (b) pool grantback clauses. The bottom line is a rigorous empirical approach to a policy question that has, until now, been carried along solely by theory and conjecture.
Keywords: patent pools, transaction costs, collective licensing, patent licensing, patents, licensing, standard-setting, innovation, antitrust
JEL Classification: K00, O34, D45, O31
Suggested Citation: Suggested Citation
Merges, Robert P. and Mattioli, Michael, Measuring the Costs and Benefits of Patent Pools (April 4, 2016). Ohio State Law Journal, Forthcoming; UC Berkeley Public Law Research Paper No. 2759027. Available at SSRN: https://ssrn.com/abstract=2759027