entry
if entry is easy, rapid and costless, profits would attract competitors
The higher sunk costs, the less likely that entry will occur, which makes the incumbent stronger. fixed sunk costs can be either exogenous or endogenous, or both:
Exogenous sunk costs are the plants and machines, technology for producing and distributing the good.
Endogenous sunk costs refer to R&D and advertising to increase the perceived ąuality of their products, and they are a choice….thus, it is unlikely that entry will affect incumbents
When analysing market power , look at the likelihood of entry.,,, switching costs, lock-in e$ects and network externalities are obstacles to entry, as consumers lack incentive to turn to new even better suppliers
Also look at the history of the industry, in particular previous episodes of entry and the incumbent’s reactions. Suppose that the incumbent leader has consistently priced aggressively whenever new competitors have entered…. this reputation discourage them from entering( predatory behaviour by an incumbent dominant firm).
Buyers³ power
The ability of a firm to charge high prices also depends on the buyers’ degree of concentration. A firm is more free to exert map if it faces a large number of dispersed consumers or buyers, than if it faces one or a few strong buyers.
A strong buyer can make use of its bargaining power to stimulate competition among sellers, either by threatening to switch sellers, or by threatening to start upstream production itself.
when there are few buyers, entry into the sellers’ market is much easier.
for instance: there is an incumbent monopolist, and potential entrants would have to make a considerable sunk investment to operate in this market. If buyers are dispersed, and potential entrants have similar cost levels, orders are likely to be distributed across sellers. and as a result no new firm might enter the industry, even though each potential entrant is more efficient than the monopolist.
if buyers fail to coordinate their decision of which seller to select, they might end up with a seller monopolist, and hence higher bills than if entry had occurred…..When instead there is just one buyer (or all the buyers coordinate), he orders from one of the entrants, thereby making entry into the industry possible.
buyer concentration [or coordination], constrains seller’s map.
The role of buyer power in constraining sellers is typically well taken into account by antitrust agencies. In European merger cases, for instance, it has led
the EC has cleared anticomp mergers , on the basis that buyer power constrained seller’s map
In Enso/Stora, the merging firms produced Liąuid Packaging Board (LPB), used for the packaging of milk and fruit juice. The merger was expected to give them a market share between 50% and 70%. Other industry characteristics, such as high barriers to entry, also suggested an anti-competitive impact. Yet, the merger was approved on the grounds that buyer power in this industry was so large (Tetrapak alone buys 60-80% of total sales) that the merging firms would have been unlikely to exercise market power. The EC argued that the main buyer, Tetrapak, ”would have the option of developing new capacity with other existing or new suppliers, should the parties attempt to exercise market power“ (Enso/Stora: paragraph 91).
A similar argument was used by the EC in the ABB/Dasmler−Benz case, which was cleared (subject to conditions) mainly on the grounds that Deutsche Bahn (the German railways operator), the only buyer for mainline trains, would have exerted competitive pressure on the producers of trains and railways materials
lower prices obtained by buyers are passed on to consumers only if there exists enough competition among the buyers
•Econometric techniques:
A/(direct) map assessment
B/The estimation of the residual demand elasticiy [RDE]:
Instead of asking by what percentage a price rise of firm A would increase demand of firm B, C, and so on, this techniąue just asks by what percentage a price rise of firm A, would decrease its own residual demand [the demand left after all the other firms have satisfied their demands].
A low estimate of the RDE means high map [of firm A], as a many consumers would continue to buy from firm A, rather than switching (or ceasing to buy the product).
a high estimate would suggest low map.
the RDE estimation cannot tell us whether market power is low because of competition from firm B, or C, or other, since the rival firms are considered as a collective, and their specific role in constraining the market power of firm A cannot be singled out.
Collusion and horixontal agreements
types of Collusive agreements:
agreements on sales prices, allocate ąuotas among themselves, divide markets [so that some firms decide not to be present in certain markets in exchange for being the sole seller in others], or coordinate their behaviour.
types of collusive arrangements
cartel-like structure where a central office (secret, if anti-trust laws exist) takes the main decisions
firms communication to sustain the agreement.
tacit collusion: where firms never meet , or exchange sensitive information ….here, the law should not intervene
•What is collusion?
where firms set prices close to monopoly prices.
collusion can occur both when firms act through an organised cartel (explicit collusion), or when they act in a purely non-cooperative way (tacit collusion).
Consider two fruit sellers in a street market. both sell pears of identical ąuality, and each pay 18 per kilo to their suppliers. Imagine also that each seller thinks that 28 per kilo is the monopoly price, and believes the other thinks in the same way. When a seller arrives at his stall, he has to decide the sales price. Suppose that he thinks the rival is setting a price of 28. If he charges 28 for his pears, he will get roughly half of the buyers. But he will have a strong temptation to charge a lower price than his rival: if he sets a price of, say, 1.98, consumers will all buy from him. he will make more profits than if he sold at the ”collusive“ price of 28.8
Thus, for collusion to arise, we need:
1- its participants must be able to deteGt in a timely way, any deviation (a firm setting a lower price or producing a higher output than the collusive levels agreed upon)
2-there must also be a punsihment, which might take the form of rivals producing much higher ąuantities (or selling at much lower prices), after the deviation, thus depressing the profit of the deviator.
but… under which conditions he will deviate?
If the street market is small enough, and if the sellers post the prices of the fruit they sell, detection of the price cut will be immediate. After the price cut war starts…and both lose.
collusive prices arise through purely non-cooperative seller behaviour. In other words, if detection of deviations is rapid, and if (market) punishments of deviations are likely, then tacit collusion arises: the sole awareness that a deviation will be identified, and that a ”punishment“ will follow.
Coordination: Tke difference between tacit and overt collusion
in tacit collusion, If firms cannot communicate, they can select a price (or a ąuantity) which is not jointly optimal for the firms, and which might be difficult to change.
Using the market to signal intentions to coordinate on a different price might be very costly. If a firm believes the ”right“ price for the industry is higher and increases its own price to signal it, it will lose market share in the adjustment period. If a firm decreases its own price to try and coordinate on a lower eąuilibrium price, this might be understood as a deviation and trigger a costly price war.
Under explicit collusson, instead, firms can coord a jointly preferred eąuilibrium…Furthermore, if there are market shocks, explicit collusion allow the firms to change to a new collusive price without the risk of triggering a period of punishment.
if one seller knows that demand for pears has decreased, so that he thinks the optimal price is now lower, say 1.88. Absent communication with the other vendor, our seller faces a problem: if he reduces the price to 1.88, collusion might break if the rival vendor misinterpret the new low price as a ”deviation“, and start a price war as a punishment. However, if he sticks instead to the usual price of 28, he will make lower profits, because demand is lower…..Explicit collusion avoids this problem
Market allocation (or market-sharing) schemes: MAS:
where a firm sells in a certain region (or serves customers of a certain type), whereas the rivals sell in other regions (or serve customers of a different type) –
MAS, whether achieved by explicit collusion or by historical accidents, allows prices to adjust to new demand or cost, without possible price wars….because, as long as each firm does not serve rivals’ allocated segments of demand, prices can change without the collusive outcome being disrupted. This is why MAS, as collusive schemes, are preferred.
cp should focus on explicit collusive practices
collusion factors:
They are identified by using the icco = the incentive constraint for collusion: ICCO : only if the immediate gain [from a deviation], is lower than the future lost profit [when rivals react] , the firm choose the collusive strategy.
Studying the industry to find the factors likely to lead to collusion, is crucial.
types of collusion factors:
A/Structural factors:
Low Concentration of firms: collusion is more likely. the less the firms in the industry.
If firms are symmetric, a lower number of firms is eąuivalent to a higher degree of concentration, which means more likelyhood of (tacit or explicit) collusion.
if firms are asymmetric (in capacities, market shares, costs, or product range), the less likely collusion will be.
concentration facilitates collusion
B/ Entry factors: The easier the entry, the more difficult to sustain collusive prices.
C/ Cross-ownerskip and other links among competitors: If a firm has a participation in a competitor, even without controlling it, the scope for col- lusion will be enhanced….Thus, is best not to allow a firm, to have minority shareholding in a competitor.
D/ Regularity and frequency of the orders: Regular orders facilitate collusion : an unusually large order would give a very strong temptation to deviate
E/ Buyer [power=concentration]: A strong buyer can use its bargaining power to stimulate competition among the sellers, either by threatening to redirect orders, or by threatening to start upstream production itself.
also, by grouping small orders into one or few, a powerful buyer can break collusion, as such groupings would induce suppliers to deviate (from the collusive strategy ).
Finally, strong buyers might design procurement auctions, to minimise the risk of collusive behaviour among suppliers
F/ demand shocks: when current demand conveys no information about future demand – that is, demand shocks are identically distributed. this is called a ‘positive shock’…. Then, it is as if a large order suddenly arrived, and the analysis would be as above: firms would deviate ….(Conversely, collusion will be more likely if a negative shock occurs.)
Demand stability might help collusion, if it increases the degree of observability in the market
G/ Symmetry:
Symmetry can concern different dimensions (such as market shares, number of varieties in the product portfolio, costs and technological knowledge, capacities)…..people who are in a similar position find it easier to arrive at an agreement which suits all of them.
H/ Multi-market contacts: [the same firms meeting in more than one market]
why the help collusion? because it is more costly to deviate, since they would be punished in all the markets at the same time.
eg. airline fares are signifi- cantly higher in routes where there exist carriers that have contacts on several routes;
eg.prices tend to be higher in US mobile telephone markets characterised by multi-market contacts.
E/Price transparency and exckange of information [agreements to exchange past and current data]
J/Observability of firms actions: Detection of deviations:
Secret price cuts: if prices (or price discounts) charged by competitors, and market demand, are not observable [by sellers], collusion would be more difficult to sustain, but it could still arise at eąuilibrium.
observability of prices and ąuantities helps firms reach the most collusive outcomes (under perfect observability, price wars that are costly for the firms would not occur)…thus, cp should pay special attention to practices that help firms monitor each other’s behaviour. One example are information exchange agreements IEAs. resale price maintenance and best price clauses, also increase observability.
K/info exchange [via trade associations or in other ways, firms exchange data on prices, ąuantities, capacities, customer demand, cost and so on], facilitates collusion, as it allows to identify deviators and better target punishments….BUT….there might also be efficiency effects behind infoexchange. For instance, better information about demand might allow firms to increase production in markets, times, and areas where demand is higher….thus helping ew…. However, there is no need for infoexch to achieve such ew anyway.
L/ Coordination issues: The role of communication
Consider for instance two firms are told by a regulator that their prices cannot be higher than a certain level, say 100. [a clear benchmark (the focal point)], and one can bet that 100 will be the price that they will set.
History might also provide hints. Many European markets have been protected from foreign competition for a long time, resulting in national monopolies in many industries.
if, once barriers started to fall, if each firm had stayed in its national market [without entering foreign ones] it would provide a good collusive eąuilibrium….Instead, starting to export might be considered a deviation and might trigger a national price war, with rivals ex- porting in turn. Therefore, remaining in its national market, is a focal point [like a cma/ec benchmark]…. and only when demand and technology substantially change, might firms want to enter foreign markets.
If firms were colluding explicitly they would communicate to achieve higher collusive prices (provided that firms are symmetric enough), and/or market sharing rules…. But even if they did not explicitly collude, they could still transmit information to each other:
a. Announcements of future prices (or production plans) ,(known as ”cheap talk“), help collusion. 2 types of announcements:
a.1 / “Private” announcements: are directed to competitors only. eg. a firm sending a fax to rivals where it is stated that from next month it intends to set a certain price….this helps them collude, by avoiding price wars and price instability…A firm might announce a price increase effective, say, in 60 days, but then revert to the current price if the other firms did not follow similar announcements ofprice changes…. This way, firms arrive at a commonly agreed price without incurring the risk of losing market shares or triggering price wars during the period of price adjustments
The usa (ATP) case: ATP is a company owned by the major US airlines whose main purpose is to disseminate price information to airlines and operators, like travel agents, using computer reservation systems. Such information is fed to the ATP by each company, and it contains the fare and the route to which the fare is applied, the possible restrictions to this fare (for instance, which type of consumer can buy it, if advance payment is reąuired, if a minimum number of days of stay are reąuired; first and last tsGket dates (that indicate the period during which the fare can be sold), and first and last travel dates (that indicate when the travel should take place).
The DOJ alleged that airlines used this information to tacitly collude to coordinate price increases. For instance, airline A could announce today a price increase on the route from city 1 to city 2 and put a first ticket date in thirty days time, so that nobody could actually sell a ticket for that route at the new fare. Since this information was public to all other airlines, airline A could then see the reaction to this price announcement. If its competitor on the same route, say airline B, matched the price increase, then it would be left unchanged and later become effective. But if airline B did not match the price increase, then airline A could still revise the fare. The process could then continue, with airlines adjusting their fares until a convergence is reached, but without consumers ever having the possibility to buy tickets at the announced future fares.
eg. Communication in auctions : in first-price auctions [where several objects are for sale at the same time]…bidders might use their bids to signal a way to share the objects…For instance, if the government was auctioning an asset in Milan (whose area code is 02) and other assets elsewhere, I might indicate my interest in the Milan asset by submitting one or more bids for, say, 1,002 euro each. Other bidders will then understand that I am proposing to share this asset, so they let me buy it.
-Another case is given by a 1999 auction of ten identical blocks of spectrum in Germany.: By making different bids for identical objects, Mannesmann was signaling the intention to share the market. Further, by choosing 18.18 and 20 it was signaling the fact that it would have been happy with its rival increasing its bid on blocks 1-† by 10 per cent (the minimum admissible bid increase), so that both would have ended up with five blocks each paid at 20 million. Sure enough, T-Mobil’s managers got the message (as they admitted in public declarations), and in the second round they bid 20 million on blocks 1-†, after which the auction ended.
B/Public announcements
public statements about bidding intentions should not be allowed, and measures might be taken to avoid signaling through bids: ”For example, bidders can be forced to bid ”round“ numbers, the exact increments can be pre-specified, and bids can be made anonymous.“
eg. a firm advertising the prices of its products in newspapers.
when prices are ”transparent“ for both consumers and firms, this should not be considered as an anti-competitive practice.
In some cases, prices are made public to increase transparency among rival firms, rather than to the benefit of customers.
the ethyl case:
the four producers of anti- knock gasoline additives allegedly used facilitating practices to decrease competition. One such practice was that firms issued press notices on price increases (the other practices were advance notice of price changes, uniform delivered pric- ing and most-favoured nation clauses).
announcements directed to rivals only, should be forbidden.
announcements about current and future prices which carry commitment value vis-a-vis consumers, are ew enhancing.
Pricing rules and contracts:
Some clauses in long-term contracts between a seller and a buyer might condition the price paid by the buyer either:
a.to the price offered by the seller to its other buyers, or
b. to the price offered by other sellers to the same buyer. this one is more likely to be anti-competitive
A most−favoured nation (or most-favoured customer) clause (MFN) :
engages a seller to apply to a buyer the same conditions offered (by the same seller) to other buyers.
The clause can be of two types:
a. retroactive MFNC states that the buyer will be offered a price reduction, if future buyers will get a lower price for the same good
b. contemporaneous MFNC engages a seller to offer a buyer the same price as its other buyers . is an agreement not to price-discriminate.
These clauses are often said to be anti-competitive, because they make it more costly for firms to give price discounts: a firm that wants to attract new customers must also reduce its price and margins on existing customers.
However, MFNC does not facilitate (tacit or overt) collusion, but just makes it harder to deviate from a collusive outcome…but it also makes it more costly to carry out a punishment, for the same reason (the punishment cannot be selective, and target only rivals’ customers).
Meeting−competition clauses : MCC
if the buyer receives a better price offer from another seller, the current seller will match that price. In this case, the potential for collusion is higher, and twofold:
1- the clause works as a device to exchange information: whenever a buyer is offered a better price, it will have an incentive to report that information to the current seller. This will make firms immediately aware of a deviation…and timely detection of deviations is a crucial element for collusion.
2-the clause reduces the incentives to deviate in the first place: if rivals can retain their current customers due to a meeting-competition clause, the price decrease can only attract new buyers, but cannot steal existing buyers from other firms.
the pro- collusive impact of mcc isso strong that cma.ec should put them under a per se prohibition rule.
Resale price maintenance RPM :
Resale (or retail) price maintenance (RPM) is a vertical agreement whereby a manufacturer imposes upon its retailer(s) the price at which the good should be sold in the final market. As I will show in chapter 6, there are a number of reasons why RPM can be efficient, and therefore pro-competitive. However, RPM might also facilitate collusion among manufacturers.
The above analysis has identified the main factors which affect the likelihood of collusion. For cp, this is useful in two respects:
- it indicates which factors an anti-trust authority should pay attention to, if it wants to try and prevent (explicit or tacit) collusion.
- in a merger case which raises doubts of coordinated effects (joint dominance), it is important to correctly assess the role played by each variable, in determining the likelihood of future collusion in the industry.
Conditions for collusion
Collusion can arise only if each firm prefer to play the collusive action rather than deviate from it (and be punished). the lower the deviation profit, relative to the collusive profit, and the lower the profit in the punishment phase, the more likely that collusion will be sustained. (The harsher the punishment the stronger the deterrent to cheating on the collusive agreement.)
Credibility of punishment, once a deviation takes place:
If firms were not be willing to carry out the punishment,[lowering prices], once a deviation is observed, the threat of the punishment would not exist, and collusion could not be sustained. thus, for collusion to be sustained, firm must prefer to engage in the punishment action (i.e., retaliate), rather than to deviate from it (i.e., not retaliate).
In ”trigger strategies“ models, after a deviation from the collusive action is observed, the punishment consists in all firms reverting to the one- shot Nash eąuilibrium forever. Clearly, here participating in the punishment is rational: given that all other firms play the Nash eąuilibrium actions, a firm has no incentive to deviate from its own Nash eąuilibrium action.
In another class of models, firms carry out optimal penal codes that involve stick and carrot strategies….the punishment (the stick ) for deviation, resulting in firms making negative profits during the punishment period.
the penal code also establishes that firms will immediately revert to the collusive actions if they do take part in the punishment (the carrot), whereas the punishment would continue unless all firms take part in it.
Cartels and renegotiation:
a collusive outcome can be reached without actual meetings. But in explicit collusion (rather than tacit collusion), firms meet to coordinate on prices and punishment strategies.
the possibility that firms meet again and renegotiate the agreement might actually break the cartel
since firms anticipate that the punishment would be renegotiated and therefore would not take place, there is nothing that will prevent them from cheating in the first place. Only if firms could commit not to meet again or they found it very costly to meet, would they be able to sustain collusion at eąuilibrium.
what is legal ?
Standards of proof: market data v. hard evidence
Consider a literal use of the definition of collusion: a collusive outcome is where prices are ”high enough“
However, it would be very difficult to look at market outcomes to decide whether there has been an antimcomp. why:?
a.price data might not be available, and when they are, they might refer to list prices, rather than effective prices
b.even if reliable data existed, there would probably be disagreement about what the monopoly price in an industry is. Sellers might have very different views of what that price would be, and an outside observer could have yet different perceptions. also, cost estimates differ widely
c. how close to the monopoly price should sales prices be, for them to be ”too high“ and therefore collusive?
d. the very principle that firms could be convicted solely on the grounds that they charge ”too high“ prices is wrong, because it punishes with cma.ec interventions, firms that are successful enough to find consumers willing to pay high prices for their products. cl should not punish map per se.
to observe that prices move in a similar way over time [price paralelism], is not enough to establish that firms are guilty of collusion. Common exogenous shocks such as the increase in input prices of all the suppliers, or an increase in inflation, or an increase in property prices would lead all the sellers to increase prices proportionally, without colluding.
ec: tacit collusion is an infringement of article 81 (ex-85)…. the term ⁵concerted practice’ mentioned in article 81 [among the prohibited practices] cover also tacit collusion’
Note, however, that in some cases price parallelism can only be credibly explained by coordination among firms, even if there is no proof of the latter. For instance, in Dyestuffs, price rises were so simultaneous that it is impossible that they had not been previously agreed upon….this is The ”parallelism plus“ rule :
price parallelism accompanied by a facilitating factor (eg rpm, best price clauses, or exchange of information), is only ilegal, if firms have coordinated to introduce or keep the practice…..iow, if there is no proof that they have agreed on a particular practice, the very fact that they have followed that practice should not be proof of collusion. For instance, in the Wood pulp decision, the EC saw a facilitating practice in the fact that producers of pulp for paper production announced their price changes with the same advance. Together with parallel price movements, this was seen as evidence of collusion. However, the ECJ found that the wood pulp sellers introduced this practice at the reąuest of their customers, the paper producers, who wanted to have more price transparency and be informed, in good time, about input price changes, so as to face them. Accordingly, the ECJ ąuashed the EC decision
The presence of periods of price wars is no proof of collusion either.
even if collusion might be proved on the basis of market outcomes alone, and if good and reliable data existed, econometric techniąues are always ambiguous….Thus, econometrics is only complementary evidence, never conclusive proof, of collusion.
Thus, inferring illegal collusive behaviour – i.e., inferring conspiracy [in the US], or infringement of article 81 ,in the EU – from market data (that is, only looking at the outcomes) is not desirable. …instead, the legal approach which reąuests some hard evidence as proof of collusion, is best…..iow, Firms should be convicted for anti-competitive behaviour only where there is proof that they have communicated, to sustain collusion. Such communication can be of various kinds:
- minutes of meetings, e-mail messages, memos and other written (or recorded) evidence concerning agreements on prices and ąuantities
- firms coordinating, to establish the environment that facilitates collusion. For instance, they might decide to exchange detailed price and ąuantity information via their trade association, or they might set up a forum where they can announce future prices to each other (as in the Asrlsne farsff ªublsshsng case), or agree on a rpm scheme, to make more uniform or transparent their prices…..In all such cases, if there is evidence that firms have not acted unilaterally, firms should be found guilty of collusion.
BUT…..if this approach does not do enough to deter and punish collusion. What else can be done, then?
There are two types of cp interventions which help deter tacit or explicit collusion or break explicit collusion (i.e., cartels). I divide them into ex-ante and ex-post measures.
ex-ante cp v collusion
penalties if they are found guilty of collusion:
- pay a fine (to the country’s general budget).
- pay damages to private parties (in the US, treble damages).
- the firms’ managers might be given prison sentences. Recently, in the UK and in the EU there have been discussions about making collusion a criminal offence. If this was the case, executives found guilty of collusive agreements could be imprisoned, as in the US.
Merger analysis:
Another ex-ante instrument to prevent collusion from arising is given by merger control. Indeed, we know that a reduction in the number of firms in the industry, or a more symmetric distribution of their assets would favour collusion. Therefore, by increasing concentration, and to the extent that they increase symmetry, mergers create favourable conditions for sustainable collusion
ex-post cp v collusion
cma.ec should also intervene to try and break existing cartels.
1.Surprise inspections (“Dawn raids”): by police
2.Leniency programmes: saves resources of the authority: building up a convincing enough case to be defendable in courts is very costly, but the cost of this prosecution stage can be avoided or greatly reduced by leniency, since the firms would bring themselves enough evidence to the authority.
3. Joint-Ventures and other horizontal, non collusive, agreements
a. Joint-ventures: jvs.
agreements between competitors that create a new entity to do some activities, instead of the partners. An example of JVs are research joint−ventures: where partners cease doing independent research and development, to do it together, within the new entity jointly set up.
jvs are a hybrid inbetween cartels and mergers. Consider for instance a sales JV between two competitors, whose only purpose is to set prices or ąuantities in the final market. This is nothing other than a cartel between the two competitors
At the other extreme, consider two competitors that give all their research, production and sales assets, to a newly created firm whose ownership they share. Since the partners cease any independent business in the sector, the JV is akin to a merger, and should be treated accordingly by the law.
Apart from the extreme cases where the JV is nothing other than a cartel (this happens when partners simply agree on some decisions rather than carrying out some activities together), jvs are mergers….so, there might be a trade-o between map and ew.
For instance, two competitors delegate their marketing and sales, to a jointly-owned entity might have anticomps, because the jv would have larger map, than if they operated dependently. However, there are procomps, if the coordination via the jv, allows them to rationalise their distribution and marketing efforts.
As for mergers, first look at the initial market power of the partners: if they have only small market shares, the jv should be cleared.
Additional care should however be devoted to ancillary restraints. Unlike pure mergers, where the partners disappear and their assets are entirely transmitted to the merged entity, in a jv, partners might continue activity in the sector. Therefore, their decisions – and welfare as a conseąuence – might be affected by restraints to their behaviour. cma.ec should avoid restraints unnecessary for the operations of the JV, such as output and price restrictions in markets unaffected by the JV.
b.Research joint-ventures [RJVs]: Collaborative agreements between firms on (R&D)
R&D is non-rival, since it can be used by other parties, without its value being diminished. Knowledge might be costly to create the first time, but once it is there, its diffusion does not modify its nature. we would like R&D to spread as much as possible in society.
Further, one would like to avoid that firms repeat large investments to get knowledge which exists already: diffusion would avoid duplications of costs.
As a result of these two features, the market alone is unlikely to give rise to socially optimal levels of research. Research jvs cope with these problems and create appropriate levels of R&D. If firms collaborate in research, they will share the cost of R&D, thereby increasing their incentives to invest; they will also have immediate access to R&D output, thereby increasing di usion; finally, they can coordinate their effort, thereby avoiding duplication of investments.….However, allowing competitors to carry out R&D together might also have costs. For instance, the main motive behind doing R&D might be the hope to get a lead over rivals: if all the main firms in the industry take part in a co-operative venture, the incentive to carry out R&D falls; if the jv does not only concern the research stage, but also the production and marketing stages, co-operation might extend to the product market and result in collusive behaviour. Therefore, while there are good reasons to allow for research jvs among competitors, such horizontal agreements should be limited:
a. such a policy is beneficial only insofar as there exist high enough spillovers. Indeed, if spillovers are high, firms are unable to appropriate their R&D effort, and co-operation increases R&D. Unfortunately, the need for clear rules and for avoiding arbitrariness and influence, on the one hand, and the difficulty of measuring spillovers on the other, make it impossible
b. authorities should make sure that the R&D co-operation does not ”spill over“ to the product market competition stage. It is conceivable that firms that start to work very closely on R&D projects might start to extend the coordination of their behaviour onto other spheres of the life of the firms…..a market share criterion is a proxy, to screen potentially harmful ventures
c. cma.ec should scrutinise ancillary restrasnts to the agreement for possible anti-competitive clauses. This might be the case, for instance, if the agreement setting up the research jv, divides (geographical or product) markets, thus relaxing competition artificially; or if it calls for the payment of per-unit royalties to the jv: this would increase the unit costs of output, and would rise prices
In the US and in the EU, rjvs are seen with favour…In the EU, the R3D BloGk Exemption issued in 2000 , exempts from article 81 of the Treaty certain forms of co-operative agreements in R&D, subject to certain conditions: partners should be free to use the results of R&D independently in production and distribution (to avoid restrictions of competition), and when these partners are rivals, their combined market share should not exceed 25%….. also, where the joint-venture does not ąualify for the R3D BloGk Exemptson, it is still possible for the parties to apply for an individual exemption to article 81.
c. Cooperative standard-setting:
There are many examples (compact discs, television, encripting of digital mu- sical, digital TV, Web protocols, videocassettes) where firms that compete in the development of a new technology decide to set common standards. The ąuestion is whether such co-operation is ew improving or anti-competitive.
the main benefit of common standards , is that consumers belong to the same single network. Therefore, they will be able to communicate directly in physical networks (for instance, each consumer can exchange files with all computer users who have the same operative system), and they will enjoy a larger variety in indirect networks (for instance, software programmers will face a larger market and will develop a larger number of applications). …..imagine the inconveniences if you could call only half of the population because the other half uses a di erent telephone standard!
An additional benefit is that consumers do not risk being stranded with a product that uses a standard that turns out to be abandoned later.
A common standard also implies fiercer competition, since consumers will face greater choice within the same standard
However, setting a common standard means that there is no competition for the dominant standard. Thus:
- the industry may not pick up the best standard, or the one that would prevail if there was market competition for the dominant standard
- in absence of a co-operative standard, the market will end up with one standard, with incompatible products disappearing altogether.
Therefore, co-operative standard-setting eliminates exante comp, but it intensifies expost comp.
Overall, it makes sense to allow co-operative standard-setting, but, ancillary agreements should be carefully scrutinised. For instance, side clauses which call for per-unit of output royalty cross payments, should be avoided, because they relax competition