Oft Lectures On Network Effects And Two Sided Markets

Network effects and two sided markets OFT lecture - 16th of January, 2007 ! Alexia Gaudeul University of East Anglia and ESRC Centre for Competition Policy

Plan of the presentation ! Two parts: Network effects and Two-sided markets ! In each part Definition Motivation / Application Models with basic concepts and logic Policy implications Review of required reading

Method of presentation ! Theoretical models to understand the mechanisms of competition in the specific setting under study. ! Application of theory to draw out policy implications ! Work required: Work out models on your own to assimilate their logic. Read the required readings!

Network effects ! Def: A change in the benefit, or surplus, that an agent derives from a good when the number of other agents consuming the same kind of good changes. ! Spreadsheets (Brynjolfsson and Kemerer 96): 1% increase in installed base means a 0.75% increase in price Products that adhere to the standard get 45% higher price ! Network equipments (Forman 01 and Forman and Chen 03) Firms are able to exploit cross-product and cross-firm switching costs to induce strong legacy effects and lock-in. Compatibility is a strategic variable in firms’ product design ! DVD vs DIVX war (Dranove and Gandal 03)

Rational expectations and multiple equilibria: Katz and Shapiro (85) Duopoly with two incompatible products. Consumer of type z gets utility v(qi)+z from network i=1,2 of size qi. Net utility: u(z)=v(qi)+z-pi Net price: Pi=pi-v(E(qi)) All consumers choose the network which net price is the lowest. For both firms to survive, I must thus have P1=P2=P ! The choice thus depends on the expectations of the consumer. ! If z is uniform over [0,1], then z>P buy one of the goods and Total demand is q=1-P= q1+q2=1-pi-v(E(qi)), i=1,2 ! Profit is then qi(pi-c). One obtains reaction functions and a Nash equilibrium, with the requirement that E(qi)=qi in equilibrium (rational expectations). ! ! ! ! !

Rational expectations and multiple equilibria (suite) ! Competing for the market, or competing in the market (standardization)? ! What drives expectations? Chicken and egg problem: low price to encourage adoption, high price to exploit adoption...

! Pricing for progressive adoption (subsidy then exploitation of the user base) Do consumers anticipate this potential lock-in problem?

Coordination and lock-in (Farrell and Klemperer, 2006) ! Consider two users who consider two technologies, old or new. The old technology provides utility u(q) if q=0,1,2 join, while the new technology provides utility v(q) if q=0,1,2 join. ! Assume u(2)>u(1) and v(2)>v(1): there are positive network externalities. ! Assume also that u(2)>v(1) and v(2)>v(1) (it is preferable that both users coordinate on the same network. ! There will be excess inertia if v(2)>u(2) and both users stick to the old network for fear the other may not follow. ! There will be excess momentum if u(2)>v(2) and both users switch to the new network for fear not to have followed.

Coordination and lock-in (suite) ! Path dependence: Initial conditions in the market determine its future. Initial conditions are greatly influenced by random factors

! Patterns of adoption depending on strength of preferences for one or the other network Early adopters and followers (bandwagon)

Compatibility problems on the supply side: Besen and Farrell (94) ! Two firms have the choice to make their product compatible or incompatible. If products are incompatible, then each firm has an equal probability to win the whole market and make monopoly profit M. If firms choose compatibility, then they will both make duopoly profit D. Typically, M>2D , as competition reallocates some social surplus from the firms to the consumers.

Compatibility problems on the supply side (suite) ! Consider first a compatibility game that occurs over two periods. In the first period, firms choose compatibility or no compatibility. In the second, profits are made. Since M>2D , firms prefer no standard and probability _ to win in the second period. ! Consider now a modified game whereby it is the firm that spends the most in the first period that wins the standards war. Then, if incompatibility is chosen in the first period, one firm preempts the other and spends M so the other spends 0 and the expected profit is 0 for both. Therefore, firms will choose compatibility.

Compatibility problems on the supply side (suite) ! Shapiro and Varian (1999) “The Art of Standards Wars” "Evolution" strategy (new technology compatible with old, eg Color TV) "Revolution" strategy (incompatibility, eg DTV) "Rival Evolutions" (compatible with old, incompatible with other new, eg DVD vs DIVX)

! See also Besen and Farrell (1994), “Strategies and tactics in standardization”.

Policy issues ! Fragmented vs. dominant network Ex-post and ex-ante efficiency When to intervene? (DTV, GSM)

! Network as essential facility The problem of bottlenecks (eBay, Microsoft) How far can one prevent access to one’s network?

! Network tipping Abuse of dominance, eg Microsoft Problem if consumers are myopic

Required reading: summary ! Katz and Shapiro (1994): “Systems competition and network effects” ! Divergence from social optimum is possible (monopolies, externalities) ! However: Markets may self regulate. Government may have wrong incentives, biased in favor of existing users. Lack of information to impact the outcome favorably.

Two sided markets ! “(M)arkets in which one or several platforms enable interactions between end-users, and try to get the two (or multiple) sides “on board” by appropriately charging each side” (Rochet and Tirole, 2004)

Two sided markets Auctions: eBay " Buyers and sellers benefit from variety and competition on the other side.

Payment systems: credit card networks "

If merchants multi-home, credit card holders need not do so.

Shopping malls "

Ability to select participants on one’s own side (upscale, downscale)

Video game consoles: Sega vs Nintendo "

Platform for development, attract investment.

Media "

Readers dislike advertisement, but like lower prices

Match-making "

Who does the choice?

Ownership in two sided markets (Gaudeul and Jullien 2001) ! Consider two sides, A and B, of a market, who can contact each other only through an intermediary. Side A receives surplus fN when N members of side B join, while side B receives surplus Fn when n members of side A join. Prices for joining are set by the intermediary at p for side A and P for side B. ! Net profit for the intermediary is (p-c)n+(P-c)N, assuming marginal cost of operation is a constant c, the same for both sides. ! Total welfare is W=(f+F)Nn-a(n)-A(N)-c(N+n) a(n) and A(N) are the sum of access costs for all participants in the market Participants may be more or less eager to join, and this may depend on how many joined already (gregariousness and then congestion)

Ownership in two sided markets (suite) ! State-owned subsidized monopoly intermediary ! Not-for profit monopoly intermediary ! For-profit monopoly intermediary ! Competing intermediaries with global database of clients ! Competing intermediaries with proprietary database of clients Divide and conquer strategy

Policy issues ! Below-cost pricing must be judged by considering both sides of the market. Not necessarily predatory. ! Ownership of intermediaries, of databases, and ability to multi-home must be considered. ! Buyer-side and seller-side intermediaries Possible limitations in choice Market design becomes a private matter

Required reading: summary ! Wright (2004) “One sided logic in two-sided markets” Prices may not reflect costs Above cost prices may be sustained in a competitive equilibrium Below cost prices do not indicate predation Competition may decrease efficiency Imbalanced prices for the two sides does not indicate crosssubsidy (i.e. exploiting strength in one market to obtain dominance in the other). In fact, each side may benefit from an imbalanced pricing structure.

Conclusion ! Interdependence of the participants in a market. within one’s own side (network effects) vis-à-vis the other side (two-sided markets)

! Consumers are limited by the choice made by other consumers, or by the intermediary. ! Principles for the regulation of intermediaries are lacking. ! The effect of subtle changes in market rules is not well understood Need for empirical work, focus on Internet markets.

Questions, suggestions, comments? Contact: [email protected]