Zero-price markets: Updating the Analytical Toolkit

M. Botta, Presentation here.

  1. 1. COMPETITION POLICY IN ZERO-PRICE MARKETS UPDATING THE ANALYTICAL TOOLKIT MARCO BOTTA Joint meeting of the OECD Competition and Consumer Policy Committees Paris, 28th November 2018 Max Planck Institute for Innovation and Competition | Munich
  2. 2. Outline 2 • Updating the competition analytical tools: 1) Relevant market. 2) Market power. 3) Anti-competitive conducts. 4) Potential remedies. • Conclusions – questions for further debate. 
  3. 3. Limits of the SSNIP test 3 • The relevant market is traditionally defined via the SSNIP test. • Problem: what is “small, but significant price increase” in a zero-price market? • ‘Free effect’ : when the reference price is zero, consumers will automatically switch to any competing product in case of price increase ➢ excessively broad definition of the relevant market. • Multi-sided markets: SSNIP test will limit market definition to one side of the market. 
  4. 4. Alternative tools to define therelevant market 4 • Alternative tools follow a similar logic as the SSNIP test: 1) SSNIC (i.e. increase consumers’ costs): + data / attention ➢ + consumers’ costs. 2) SSNDQ (i.e. decline product quality): + data / attention ➢ - product quality. • Limits of the alternative tools in zero-price markets: 1) Quantification: +5% amount of personal data / attention? 2) Heterogeneous consumers’ preferences: what type of data/attention should we take into consideration? 3) Positive effects: + data transferred can increase the product quality. 4) SSNIC and SSDQ do NOT catch multi-sided markets. 
  5. 5. Market power in zero-price markets 5 • Market power within the relevant market: element to trigger enforcement of competition policy (e.g. unilateral conducts, merger control, vertical agreements). • Traditional definition of market power: ability of the firm to raise prices above the competitive level. • ‘Free effect’ : in zero-price markets firms can never raise prices above 0 ➢ consumers would always switch to other products = NO firm has market power. • The market share has limited relevance to assess market power in zero- price markets. 
  6. 6. Factors to estimate market power in zero-price markets 6 • A number of factors can be assessed to estimate the degree of market power in zero-price markets: 1) Attention degree: users’ attention on the Internet is a ‘scarce’ resource. 2) Direct and indirect network effects: number of users; product quality. 3) Multi-homing and switching costs. 4) Access to data ➢ possibility to purchase data from third parties. 5) Sunk investment costs. 6) Degree of innovation: a) Relevance of innovation in the market; b) Evidence of past radical innovations; c) Evidence of past entry. 
  7. 7. Updating anti-competitive conducts 7 • Assessment of anti-competitive conducts based on ‘price’ should be revised in zero-price markets. • Cartels fixing the price at zero: shift from a per se prohibition to an effect analysis. • Predatory pricing ➢ what is ‘predatory’ in a zero-price market? 1) Fallacies in accordance with the current legal standards: a) EU: presumption of predation when prices are below average marginal costs. b) USA: requirement of likely recoupment in the same market. 2) Recoupment requirement should be always required, BUT extended to other ‘sides’ of the market. 
  8. 8. Updating anti-competitive conducts 8 • Exploitative conducts in zero-price markets (EU): 1) Excessive pricing (i.e. asking ‘too much data’) ➢ NOT relevant. 2) Discriminatory pricing ➢ NOT relevant. 3) Unfair contractual clauses: relevant in zero-price markets a) Clauses ‘unilaterally’ imposed by the dominant firms (e.g. social network unilaterally modified the data protection terms). b) ’Unfair’: clauses ‘un-related’ to the product, and outside the ordinary commercial business practices (e.g. users’ data are transferred to third parties without the user’s consent). c) Relationship with consumer and data protection law: open question. 
  9. 9. Competition law remedies in zero-price markets 9 • Zero-price markets pose new challenges to the application of the traditional antitrust toolkit ➢ infringement decision + fine is NOT an effective remedy. • Structural v. behavioural remedies: 1) Structural remedies (e.g. un-bundling; divestiture of a subsidiary): NOT efficient ➢ negative effect on direct network effects and product quality. 2) Behavioural remedies: the NCA ‘guides’ the firm in terms of competition law compliance: a) Tailor-made ➢ designed in cooperation with the firm (i.e. commitments); b) Possible periodic revision ➢ adaptation to the market dynamics. c) Need of monitoring. d) Risk of market regulation ➢ overlap with data protection and consumer law. 
  10. 10. Behavioural remedies in zero-price markets 10 • Examples of behavioural remedies in zero-price markets: 1) Increase consumers’ awareness (e.g. increase transparency of the contractual terms; info about the personal data collected); 2) Setting minimum standards of data protection terms (e.g. max. duration of data storage); 3) Giving consumers the opportunity to periodically revise the consent to the processing of their personal data; 4) Right to data portability. 
  11. 11. Relationship with sector-regulation 11 • Antitrust remedies can clarify unclear aspects in data/consumer law protection. • Cooperation between NCA and data protection /consumer law authorities: 1) Exchange of information during the investigations; 2) Joint sector-inquiries; 3) Consultation in designing behavioural remedies. • Competition v. consumer / data protection remedies: 1) Advantage: antitrust remedies ensure higher degree of deterrence. 2) Disadvantage: definition relevant market and market power. 
  12. 12. Conclusions – questions for further debate 12 • Are SSNIC and SSNDQ effective tools to define the relevant market? • What aspects should be taken into consideration to assess market power in zero-price markets? • How should the assessment of anti-competitive conducts be adjusted to the peculiarities of zero-price markets? • What type of competition law remedies could be introduced in zero-price markets? • What are the possible forms of cooperation between NCAs and sector regulators when it comes to designing the remedies? 
  13. 13. 13 Thank you very much for your attention! marco.botta@ip.mpg.de 

Algorithms and competition

M. Vestager, here . Large friendly letters? "It's true that the idea of automated systems getting together and reaching a...