Some Economics of Refusal to Supply William Bishop June 2005 The Modernization of Article 28 Conference. River Associates Charles Brussels, 17 GCLC
Some Economics of Refusal to Supply Critical issues in antitrust 1) In what circumstances will we compel one firm to deal with rivals? – What exactly are the criteria? 2) When we do compel dealing, how do we set the price and terms on which they deal? 3) Should special rules apply where IP is involved? – How do we set the royalty? 2
Some Economics of Refusal to Supply The Microsoft Cases n Three phenomena are joined: abuse of dominance, economics of regulation and economics of IP rights 1) The cases have led to advances in economic theory – Barriers in network industries – the “applications barrier to entry” – Clarification of leveraging (traditional) to extend monopoly – “Defensive leveraging” concept: Carlton-Waldman model – Bundling theory – especially technical bundling – IP rights and R&D in network industries 2) The cases have compelled hard-thinking about essential facility 3) Future stages will lead to clarification of IP issues - whether IP is special 3 - and how compulsory IP licences are priced
Some Economics of Refusal to Supply Contrast US v EU n EU: – compels MSFT to disclose information on servers to rivals – prohibits MSFT from compulsory bundling of WMP with Windows (i. e. compels MSFT to supply Windows without WMP on it – if anyone wants it) n US: – Verizon v Trinko: Scalia criticized “essential facility” doctrine – Explains away Otter Tail and Aspen Ski – Suggests it will apply only to protect existing arrangement (i. e. a “shield not a sword”) 4 – MSFT (N° 2) in DC Circuit: Ginsburg expresses hostility to interference with “product design” (implicit criticism of media player part of EC decision? )
Some Economics of Refusal to Supply Patent, Copyright n Patent + copyright life are the same for all patents & copyrights n Yet the incentives needed to induce the R+D differ greatly n With network effects the incentive can be perversely large – In 1980 all telephone switches in the UK were BT’s IP, mutatis mutandis for France and Germany – BT, DT + FT would rather build their own than buy someone else’s product, almost regardless of quality – In a network, “owning” the complementary asset can confer major monopoly profits. – MSFT’s browser (and search engine? ) may be similar examples 5 n Few worries about blunting incentive in cases of “me-too” inventions
Some Economics of Refusal to Supply When will we compel dealing? (1) n When ought we to declare something to be an “essential facility”? – Old Chicago solution: “never” – “There is no leverage – there is only one monopoly profit” – Abandoned economics today – But applies in some circumstances 6 – e. g. Nalebuff on Hilti
Some Economics of Refusal to Supply When will we compel dealing? (2) n Clearly policy should rely on non-intervention unless intervention increases economic welfare substantially. n More information ex post than when IP right was created – So we can improve economic welfare n But problem of upsetting regulatory bargain (“reneging”) – So should be rare – i. e. when use/abuse is particularly harmful 7
Some Economics of Refusal to Supply When will we compel dealing? (3) n Generally appropriate when: – The value of the activity is large – The welfare loss flowing from the monopoly is large – There is a remedy available that will probably work – The costs of that remedy (admin cost, error costs, avoidance activity, etc) are outweighed by the welfare improvement – CC/MMC in UK sometimes frank like this 8
Some Economics of Refusal to Supply When will we compel dealing? (4) n Two inadequate answers n In Europe – Bronner “impossible to duplicate” test is too simple – Both under inclusive and over inclusive n In USA – Scalia: “never except as a shield” / “leave regulation to Congress” – Makes Otter Tail et al depend on accident of contracting – Implies US v AT&T was an intellectual mistake 9 – In Europe Art. 82 is a fundamental basis for regulatory policy
Some Economics of Refusal to Supply Examples n n Electricity grids: Otter Tail case, et al n Software: MLP/NMPP, Microsoft (interoperability information) n Information: Magill, IMS Health, Bell South/Donnelly n Delivery system: Oscar Bronner (value and welfare loss both small) n 10 Telecommunication interconnection: intervention is ubiquitous Marketing Arrangement: Aspen Ski (questionable on most counts)
Some Economics of Refusal to Supply How to price “access” to an essential facility? n In some cases there may be a market price to guide us n General: if access needed, then access fee/price should protect investment incentives judged ex ante. That is: sufficiently high that the “investor” would have done the R+D, viewed ex ante with all risks considered. n Other classic regulatory issues arise in particular contexts: – Risks of inefficient entry – Difficulty of setting Ramsey prices 11
Some Economics of Refusal to Supply “Exceptional circumstances” in economic terms Condition Dominance Otherwise conduct can’t injure Yes, acknowledged competition dominance in PC OSes Adjacent market affected Protection against unauthorised entry in original market is part of the standard “IP bargain” Yes, market for work group server OSes Incentive to extend dominance to adjacent markets Not always present (“one monopoly profit”) Yes, to help maintain PC OS market power Refusal to license has significant foreclosing effects 12 Reason for condition Application to server case Otherwise no case for compulsory licensing Yes, ability to interoperate with Windows environment essential
Some Economics of Refusal to Supply General lessons for bundling Necessary conditions Strong dominance Yes, in PC operating systems Clear incentive to extend dominance Yes, “defensive leveraging” One monopoly profit situations Clear mechanism by which bundling today can affect future competition 13 Media player case Counter example Yes, through indirect network effects Tetra Sidel
Скачать презентацию Some Economics of Refusal to Supply William Bishop
142706904a0c80efb52c7a615d4a5382.ppt