Market power, competition, and welfare

Transcription

1 Market ower, cometition, and welfare. Allocative efficiency. Prodctive efficiency 3. Dynamic efficiency 4. Pblic olicies, and incentives to innovate 5. Will the market fix it all?

2 . Allocative efficiency Definition of market ower: the ability of a firm to rofitably raise rice above marginal costs A matter of degree, not of existence The deadweight loss (see Figre.) Inverse relationshi between market ower and welfare An additional loss of monooly: rent-seeking activities (see Figre.)

3 Figre.. Welfare loss from monooly O m R c T S c q m q c O q 3

4 Figre.. Possible additional loss from rent seeking O m R c T S c q m q c O q 4

5 . Prodctive efficiency Additional welfare loss if monoolist has higher costs (see Figre.3) Qiet life and managerial slack Princial-agent models: market cometition hels, bt too fierce cometition may decrease efficiency Nickell et al.: individal firms rodctivity higher in cometitive indstries Darwinian argments: cometition selects more efficient firms Olley-Pakes, Disney et al.: indstry rodctivity mostly increases throgh entry/exit 5

6 Figre.3. Additional loss from rodctive inefficiency O R m m R c V W c c T T Z S c q m q m c q q c O q 6

7 Prodctive efficiency, II Nmber of firms and welfare: trade-off between allocative and rodctive efficiency As nmber of firms increases, market ower decreases, bt also welfare Imortant: defending cometition, not cometitors! (else, inefficiencies, and fixed cost dlications) 7

8 3. Dynamic efficiency U-shaed relationshi between market ower and welfare: trade-off between aroriability and cometition in R&D investment Lower incentives to innovate of a monoolist: innovation introdced if additional rofits higher than costs Aroriability matters: no (little) innovations if no atent rotection, comlsory licensing etc... 8

9 4. Pblic olicies and incentives to innovate Ex ante (incentives) v. ex ost (diffsion): IPR rotection garantees market ower Essential facilities (EF) doctrine Necessary, non-rerodcible ints Ex.: airort slots, ort installations, local loo EC accet EF doctrine, bt ECJ: Bronner case Imortant to reserve incentives to innovate! Aly EF doctrine only when owner has not invested to create the facility 9

10 5. Will the market fix it all? Contestable market theory: does free entry eliminate all concerns abot market ower of incmbents? Persistence of dominance nder free entry Endogenos snk costs indstries: finiteness roerty Network externalities (definition, direct and indirect, coordination effects, interoerability) Switching costs (definition, natral v. artificial, cometitiveness of switching cost markets) Predatory and exclsionary ractices 0

11 Contestable markets Assme an incmbent I and a otential entrant E are eqally efficient and rodce homogenos goods. Cost of rodction is F+cq Bamol et al (98): at eqilibrim I will not set monooly rice, bt eqal AC: =c+f/q Proof (a contrario): If >AC, firm I wold make rofits; E wold be attracted into the indstry, set =AC-ε and earn ositive rofits If <AC, firm I wold make losses.

12 Contestable markets: discssion The theory of contestable markets wold have strong imlications: if entry is free, we shold not care abot monoolists, as efficient otcome is reached. Critiqe: the theory hinges on two strong assmtions: Unrealistic timing of the game (I cannot change rice as E enters the market) No fixed snk costs of entry (hit-and-rn strategy not rofitable for E if some costs are nonrecoverable) Bt the theory has the merit to stress the role of free entry in limiting market ower: crcial in merger analysis.

13 Finiteness Proerty Consider the following model (a very simlified version of Shaked-Stton s (98) There exist n firms each with a rodct of qality k (labelled so that > > > n ) and a rice k There exists a continm of consmers with identical tastes bt different incomes t. t is niformly distribted with density S (S=size of the market) on a sort [a,b], with a>0. Consmers by one nit of the good (the market is covered), and have tility U(t,k)= k (t- k ) 3

14 The game. Firms decide on entry (fixed cost ε>0). They decide on qality of the good 3. They decide rices and sell (zero marginal costs) Proosition: If b<a, only one firm will enter the indstry at eqilibrim (whatever S) (As income becomes less concentrated, more firms can enter; e.g., if a<b<4a, two firms will enter at eqilibrim. Generally, the nmber of firms which coexist at eqilibrim is finite even as S goes to infinity) 4

15 Proof of the roosition We show that two firms cannot co-exist at eqilibrim. Firms demand is derived by finding the consmer indifferent between the two qalities: From: (t- ) (t- ), we obtain: t t (,,, ) All consmers with income tt will by, all others will by. Therefore: q = = b t; q = t a. 5

16 6 Proof (cont d) Profits can be written as: By setting d i /d i =0 we obtain the best rely fnctions: Eqilibrim rices are given by: Therefore, if b<a there exists no eqilibrim with ositive, and firm will not enter the indstry. ; a b = Π = Π ( ) ( ) ; : ; : a R b R + = + = ( )( ) ( )( ) * * 3 ; 3 a b a b = =

17 Eqilibrim, when b>a R R E E 0 E 7

18 No eqilibrim, when b<a R (a >a) R When a increases, s best rely fnction shifts wards and the eqilibrim shold involve a negative rice of firm. R 0 8

19 Generalisation The finiteness roerty holds if the cost of rodcing a higher qality does not fall on variable costs It holds across a nmber of different secifications (see e.g., Shaked-Stton, 987) Stton (99) ts the reslt to an emirical test. It shows that in advertising-intensive indstries as S increases the indstry does not become fragmented (when S increases, firms have incentive to increase Ad, which in trn raises fixed costs and limit the nmber of firms in the market). 9

20 Network effects: miscoordination Assme that consmers vale a network good i as: U i =v i (n)- i, Where v i (n) is valation if n consmers by good i. v i (n) is non-decreasing and concave, with v i ()=0 and v i (z) = v i (z+j) for any j>0 (all externalities exhasted at size z) There are an incmbent I and an entrant E, with c E <c E. Networks of eqal qality. Small fixed cost of entry, ε>0. There are z old consmers, and z new consmers. 0

21 The game. E decides whether to enter or not. Active firms set (niform) rices, I and E 3. The z new byers decide btw. network I and E Assme the two networks are incomatible. This game admits two tyes of eqilibria: Entry eqilibria, where the efficient entrant enters Miscoordination eqilibria, where the inefficient incmbent remains a monoolist

22 Entry eqilibria There is an entry eqilibrim where E enters, ( I, E )=(c I,c I -ε), all z new consmers join E s network. Proof. A consmer wold have no incentive to deviate. At (candidate) eqilibrim, its srls is v(z)-c I. By deviating and bying from I, it also gets v(z)-c I. Firm I has no incentive to deviate (zero rofits also if it raises rice, negative rofits if redces it). Firm E neither: zero rofits if it raises rice, lower rofits if it redces it.

23 Miscoordination eqilibrim There is a miscoordination eqilibrim where E does not enter, I sets monooly rice I =v(z), and all z new consmers join I s network. Proof. Sose the entrant has entered and set a rice as low as c E. A consmer wold have no incentive to deviate. At (candidate) eqilibrim, its srls is 0. By deviating and bying from E, it gets v()-c E <0. Firm I has no incentive to deviate (zero rofits if it raises rice, lower rofits if redces it). Firm E neither: negative rofits if it enters. 3

24 Exclsion in network markets Incmbents can se their cstomer basis to exclde more efficient entrants. For instance: By sing rice discrimination the incmbent can exclde easily Making a rodct/network not comatible with other rodct/networks consmers may not by the latter Since coordination of consmers lay imortant role, incmbent may manilate exectations so as to deter entry 4