A case of parallel behaviour: Wood pulp
In 1984, the European Commission adopted a decision (Wood pulp) that found that forty wood pulp producers (six from Canada, ten from the US, eleven from Finland, ten from Sweden, one each from Norway, Portugal and Spain) and three of their trade associations (KEA from the US, Finncell from Finland, and Svenska Cellulosa from Sweden) had infringed article 81 (then art. 85) of the Treaty by concerting on prices. Several of the firms involved appealed the decision, and in 1993 the European Court of Justice issued a judgment (Ahlström and others v. Gommsssson) that annulled most of the EC decision…..eCJ was right in finding that the EC had not established a convincing case for collusion.
Paper manufacturers usually mix different types of wood pulp to obtain paper of a certain grade. Within any product category, pulp is interchangeable to a considerable extent, but once a given mixture has been determined, there might be large switching costs for a paper manufacturer….thus, buyers of wood pulp purchase from different producers and have long-term contracts, of about 5 years,with them.
There were more than fifty producers selling wood pulp in Europe. The forty firms involved in the decision accounted for roughly 60% of overall European sales. The industry is characterised by vertical integration, with most wood pulp producers being also paper manufacturers
a consolidated trading practice in this market, is ąuarterly announcements of prices. Some weeks or days before the beginning of each ąuarter, producers communicate to their customers and agents the prices for that ąuarter .manufacturers generally ąuoted their prices in US dollars.
the simultaneity or near-simultaneity of the announcements had another [ than collusion] innocent, plausible explanation : each buyer is in contact with several producers (because of the mixture of wood pulp used and because of diversification in the sources of supply), and would have an incentive to reveal prices set by other suppliers – at least when they are cut; most wood pulp producers are also paper manufacturers, and purchase some of their input from upstream rivals, thus being immediately informed of upstream prices; the existence of common agents that work for several wood pulp producers. All these elements, imply that the – producer, agent, buyer, agent, producer – information, on the announced prices, spreads within a matter of days, if not within a matter of hours on the pulp market.
2 hypothesis: tacit collusion, or an oligopolistic industry.
In the tacit collusion hypothesis, prices are close to each other, because the lack of coordination implies that there is no possibility to coordinate on a collusive eąuilibrium which is more efficient for the producers.
The oligopolistic theory: there is a lot of price rigidity in markets because a firm expects that if it increases prices the rivals will not follow and therefore will lose market shares, and that if it decreases prices the rivals will immediately follow and therefore will not benefit from the price cut…. Therefore, the same price would continue to hold, unless major shocks have intervened
Ec: held. ”concertation is not the only plausible explanation for the parallel conduct.“
the standard of proof is (rightly) high, as one should prove that communication and/or coordination of some kind among the firms must be the only plausible explanation for parallelism.
In Dyestuffs, firms were found sending similar price instructions to their agents and subsidiaries basically at the same hour and day. The probability that this could happen without firms having previously talked to each other was nil: concertation was in that case the only plausible explanation for parallel conduct.
But… were market prices collusive or not?
the market was very transparent : if a firm made a deviation [ from a collusive price], the deviation will be immediately detected by rivals (via the connections between wood pulp producers, agents and buyers described above). ……However, although Firms can coordinate on list prices, the prices are determined by the bargaining between a buyer and a seller. secret price cuts were widespread in the industry, and announced and prices did not generally coincide also, the firms involved accounted only for 60% of the market. ….. there was likely no collusion.
Horixontal mergers
2 possible mechanisms through which a merger can negatively affect ew:
A. Unilateral: (section 2): the merger might allow the merged firm to unilaterally exercise market power and raise prices.
B. Collusive (section 3): a merger might favour collusion in the industry. Here, the merging firm would not be able to unilaterally raise prices in a significant way, but the merger could generate new industry conditions which enhance the scope for collusion. Prices could then increase, as firms are more likely to attain a (tacitly or explicitly) collusive outcome.
In usa, a merger is evaluated according to a ”substantial lessening of competition“ test: what matters is whether the merger will lessen competition, and raise prices.
In the EU, a ”dominance“ test exists. joint dominance corresponds to pro-collusive effects,… but…single-firm dominance does not correspond to unilateral effects
Unilateral merger effects:
a merger is likely to increase the market power of the merging firms and decrease both consumer surplus and total welfare. if the merger increases efficiency in the merging firms, the rise in market power can be outweighed by the price decrease brought about by efficiency gains.
Absent efficiency gains, a merger increases market power:
To understand why a merger might allow a firm to unilaterally increase market power, consider a simple example. Imagine that in a given town there are a few independent grocery stores. Competition constrains the market power of each store: if one of them tried to increase prices, many among its current consumers would start to shop at other stores. Anticipating this, the store considering the price increase will refrain from doing so. Its market power, [ its ability to charge consumers a high price,] is therefore limited by the presence of rivals
Such market power, however, will increase if two or more stores merged to give rise to a chain of grocery stores. A contemporaneous increase in the price of each product sold by the merged stores might now be profitable, because the number of rival stores is reduced. Consumers might have to travel greater distances to find a store with lower prices, and many of them will shop at their usual store despite higher prices.
therefore, the merger increases market power of the merging firms, which in turn will increase prices
absent efficiency gains, the merger will benefit the outsiders. This is because the insiders, by increasing prices and/or reducing output, benefit the rivals.; Indeed, the rivals might gain more than the insiders from the merger.
In sum, because they increase market power, mergers which do not entail efficiency gains, hurt consumers and society at large.
Concentration
the larger the number of independent firms operating after the merger, the less likely it is detrimental to consumers. Why?: as the ability of merging firms to exert market power depends on the number of rivals. Thus, in an industry which is extremely fragmented [each firm has only tiny market shares], the impact of a merger on the market price will be irrelevant.
This gives us a rationale for using a concentration index, like for instance the Herfindahl-Hirschman Index (HHI), as a first screening for the unilateral effects of mergers.
It also justifies using a proxy, for the likely change in concentration (such as AHHI, that is, the difference between post- and pre-merger concentration) as an additional screening device.
According to the US Merger Guidelines, US competition agencies should rely on these two indices to screen mergers :
If the post-merger HHI is lower than 1,000 (low concentration), the merger will be approved.
If the post-merger HHI is between 1,000 and 1,800 (moderate concentration), the merger is approved as long as it does not result in an increase inconcentration of more than 100 points.
If the post-merger HHI is more than 1,800 (high concentration) the merger is not challenged only if it increases concentration by less than 50 points.
In all other cases, a merger raises ”significant competitive concerns“ and is likely to be investigated.
Market shares and capacities:
Another simple but useful indicator [of the likely market power created by the merger], is given by market shares:
the lower the market share [of the merging companies], the less detrimental the effect on market prices. Furthermore, a merger between small firms, might increase welfare, even in the absence of efficiency gains.
The analysis of productive capacities is also very important. The ability to raise prices by any given firm, is limited by the existence of rivals to which consumers can switch. It is therefore crucial that such rivals be effectively competitive. Therefore, the larger the unused capacity of rivals, the less likely that the merging firms will exercise market power.
Entry :
The firms’ ability to raise prices after a merger, is also limited by the existence of potential entrants
Firms which would find it unprofitable to enter the industry at pre-merger prices might decide to enter if the merger brings about higher prices or lower ąuantities…… By anticipating this effect, post-merger prices might not rise at all; or, if they do, the price increase would be transitory. The extent to which potential entrants restrain the market power of actual industry participants depends on entry fixed sunk costs. The larger (and the more sunk, i.e. committed to the industry and not recoverable) the entry costs, the higher the scope for a price increase.
Demand variables :
Of course, not only supply variables but also demand variables must be taken into account [to understand to what extent the merging firms enjoy market power] …. For instance, in industries with very high switching costs, providers enjoy market power. the lower the elasticity of market demand, the higher the scope for raising prices.
Buyer power :
the merging firms’ ability to charge high prices also depends on the degree of concentration of the buyers. Strong buyers can constrain upstream market power by threatening to switch sellers, or by threatening to start upstream production itself.
Failing firm defence:
what would happen were the merger to take place.?
what would happen were the merger not to take place.?
Eg if the merger involves a failing firm [in the absence of a merger wo7ld close ], the ex−post merger situation should be compared….. not with the ex−ante merger situation…., but with the situation after the failing firm would have exited the industry.
The failing firm defence is clearly stated in the US Merger Guidelines, section 5.1: an anti-competitive merger is accepted if:
1) the failing firm would be unable to meet its financial obligations in the near future;
2) it would not be able to reorganise successfully under Chapter 11 of the Bankruptcy Act;
3) there are no suitable alternative buyers that would keep the failing firm’s tangible and intangible assets in the relevant market while having lower anti- competitive e4ects than the proposed merger;
4) in the absence of the merger, the failing firm’s assets would have exited the relevant market.
The first two conditions reąuire that the failing firm must not only have short-run problems, but be unlikely to be viable in the medium-long term. The other conditions reąuire that the proposed merger is the only (or the best) way to keep the assets of the firms in a productive use.
Efficiency gains:
In the absence of efficiency gains, a merger should lower both consumer surplus and total welfare….. However… efficiency gains might offset the enhanced market power of merging firms, resulting in higher ew….. This is because the merger cause the insiders to be more efficient, and save on their unit costs. If these savings are large enough, they will outweigh the increase in market power and result in lower prices, to the benefit of consumers.
The new merged firm might of course still increase its prices…. This would be profitable…. However, it is not the most profitable strategy…..because of efficiency gains, a more profitable strategy is now to reduce prices and attract new customers.
therefore, with efficiency gains, the merging firms have two possible ways to increase their profits:
A. to increase prices (reduce sales), or
B. to decrease prices (increase output)
the higher the efficiency gains, the more likely that the second effect will dominate, because the insiders to the merger will decrease sales prices, and both consumer and total welfare will increase
The effect of efficiency gains on outsiders profits:
when there are efficiency gains, outsiders lose from the merger, and thus oppose it, when the merger allows insiders to cut their costs……this is because the merger changes the competitive positions of the firms in the industry, to the detriment of the outsiders.
therefore, rival firms’ profits decrease when the merger will have a positive effect on welfare, [when there exist sufficiently large efficiency gains]
look at the impact of merger announcement, on the stock market prices of outsider firms:
A. If the competitors’ share prices decrease, then market analysts and observers anticipate the existence of merger efficiencies. In turn, this should imply that the merger will increase consumer and total surplus [ ew]…..Ew increases and outsiders’ profits decrease, when efficiency gains are large….thus,.If rivals complain about the merger, it signals ew gains
B. If the competitors’ share prices increase, one should expect efficiency gains to be minimal or absent, and as a conseąuence the merger to be ewe detrimental.
firms which combine their assets, decrease their costs. Why?: the existence of economies of scale and economies of scope. Due to a merger, firms might reorganise their production so as to improve the division of labour and attain economies of scale; or they might benefit from lower costs due to joint production. Other possible gains might come from synergies in research and development, rationalisation of distribution and marketing activities and cost savings in administration. Also, takeovers might improve efficiency via the substitution of useless ceos
Efficiency gains in fixed costs might still lead to a positive ew, but only from an increase in profits due to lower duplication of fixed costs, since consumer surplus would not change.
efficiency arguments should be accepted only if costs savings[ achieved by the merger], could not be achieved otherwise. If, for instance, the firms claimed that the merger would create efficiency gains because it would reduce personnel costs…… is bullshit because these savings could not be achieved without a merger. Where efficiency gains could be achieved without a merger they should not be accepted as an efficiency defence of the merger, as they could be obtained without a reduction of comp
Asymmetric information [ cma / merging firms]
When efficiency gains are a crucial determinant [ to clear or block], the merging firms have an incentive to overstate efficiency claims. On the other hand, the rival firms which fear the merger could jeopardise their competitive positions, have an incentive to understate the efficiency gains of a merger. Ec.cma will therefore want to rely on independent studies to evaluate efficiency considerations.
Balancing efficiency and market power considerations
if there exist efficiency gains (and are merger-specific), one has to evaluate if they are sufficiently large to determine a positive effect on consumer and total surplus.
the stronger the likelihood that the merger allows the parties to exercise higher market power, the larger should be the efficiency gains reąuired to authorise the merger.
B. Pro-collusive merger effects
2 possible mechanisms through which a merger can negatively affect ew:
A. unilateral market power. Explained above.
B. pro-collusive (or coordinated) effects,
where the merger does not pose a threat of market power by a single firm, but generates more favourable conditions for collusion in the industry. In other words, the merger creates the structural conditions for the firms to (tacitly or explicitly) attain a collusive outcome. The concept of joint dominance in the EU refers to this situation.
Merger remedies:
Ec.cma might approve a merger only if certain ”remedies“ are adopted by the merging firms.
Merger remedies types:
(i) Structural remedies, modify the allocation of property rights: they include divestiture of an entire ongoing business, or partial divestiture.
(ii) Behavioural remedies, set constraints on property rights: the mergers agree not to abuse certain assets available to them, or to enter into specific contractual arrangements.
(iii)Divestitures
Divested assets can either be bought by a new firm, or be acąuired by a competitor. In both events, the cma has to ensure that the acąuirer of the assets will be an active competitor. To this end, the acąuirer should have the possibility to purchase ”all the elements of the business that are necessary for the business to act as a viable competitor in the market
to have an effective competitor in the affected markets, might also reąuire the merging parties to divest assets which do not raise competition concerns.
the merging parties have all the incentive to make sure that the purchaser of the divested assets will not be a competitive firm. The seller might therefore try to decrease their value, to select the buyer. Thus, cma should ensure that the seller does not engage in activities that could reduce the value of the assets or hinder the sales, and should keep control of the identity of the buyer.
– tension between two problems :
A. Cma.ec, have to guarantee the reinforcement or the creation of a viable firm to avoid problems of unilateral effects (single firm dominance by the merging firms).
B. Cma.ec also have to avoid pro-collusive effects after the merger (joint dominance).
Thus, the evaluation of merger remedies should follow the same twofold test used in merger analysis:
A.evaluation of unilateral effects and
B. Evaluation of pro-collusive e$ects.
Remedies should be accepted, and the merger proposal cleared, only if both tests are satisfied.
Behavioural remedies:
Behavioural remedies consist mainly of commitments aimed at guaranteeing that competitors enjoy a level playing field in the purchase or use of some key assets, inputs or technologies that are owned by the merging parties.
behavioural remedies might raise several problems, not the least being that they need continuous monitoring.
EU Merger policy
In the EU, mergers are regulated by the Merger Regulation :”only mergers which create or strengthen a dominant position will be prohibited”……But…..it has two features at odds with economic principles:
1. the dominance test does not allow the prohibition of mergers which decrease ew, without creating a dominant position.
2. efficiency gains are not considered when assessing merger proposals…… whereas economics strongly suggests that efficiency savings should be at the centre of the analysis of mergers.
1. Dominance test
A.where the merger raises concerns of unilateral price increase…does not correspond closely to the concept of single firm dominance.
B. where the merger raises concerns of (tacit or explicit) collusive behaviour….if so, the merger creates joint dominance. ….. EC has tried to use the joint dom concept, to block with mergers that raise anti-comps, but donot create single-firm dominance.
The concept of joint dominance matches closely that of coordinated effects: after the merger, there is high likelihood that a collusive outcome will arise in the industry.
In GenGor v. Gommsssson, ecj reaffirmed that the EC can block mergers if they create joint dominance
2- efficiency gains
Do efficiency gains likely offset the higher map?
So far, the EC in its decisions has not explicitly ruled out the possibility of using an efficiency defence, but it has not showed much sympathy for this argument either.
In Aérospatsale−Alensa/deHavslland, the EC argued that the cost savings would have been negligible, had not been properly ąuantified, were not merger-specific (they could have been attained without the need of a concentration) and would not have gone to consumers’ advantage.
Efficiency offence?:
At times, the EC has been accused of using an efficiency offenGe argument, that is holding cost savings against the merger, so as to protect competitors. Eg in the General EleGtrsG/Honeymell decision
Is possible that merged firms become so efficient that, even if they do not behave strategically, competitors will be forced to exit the market…thus, rejecting a merger which entails efficiency gains today, because of possible predatory behaviour in the future, does not seem a welfare improving decision (see also General EleGtrsG/Honeymell ). It would be better to first allow the merger and then, in case the merged entity tried to monopolise the industry, use antitrust laws (article 82 in the EU) to stop anti-competitive behaviour
Conclusions
the EU merger policy has two main distortions
1. prohibits only mergers which create or reinforce dominance, whereas economic analysis suggests that there exist mergers which are detrimental to welfare, even though they do not bring about dominance. Hence, some mergers are allowed which should instead be prohibited.
2. it does not recognise the role played by cost savings, which might give rise to positive welfare effects of mergers. Accordingly, efficiency gains should be taken into account, otherwise, some mergers are prohibited which should instead be allowed
There are two main ąuestions that cma.ec should ask before allowing a merger:
1. Will the merger Greate unilateral effects? “, that is ”will merging firms increase prices in a considerable way?“. If the answer to this first ąuestion is negative, there is still a second ąuestion to be asked:
(2) ”will the merger Greate pro−collussve effects? “, that is, ”would the merger modify the conditions of the industry, so that collusion will be much more likely?
1. Unilateral effects
The most direct way to assess the extent to which two merging firms might exercise market power is to ask whether they will be able to impose higher prices after the merger. ….But this is too hard to prove… thus, we resort to defining the relevant (product and geographical) market, and then assess the degree of market power enjoyed by the merging companies
If merger specific gains exist, we need to then balance whether efficiency gains will outweigh the negative impact of increased market power
After these steps, two cases might arise:
A. the merger enables the firms to significantly raise prices beyond the current level. Hence, the merger should be prohibited, or allowed only if some remedses can be identified
B. the unilateral effects of the merger are not jeopardising competition. Hence, the investigation should still deal with the possibility that collusion arises after the merger:
Pro-collusive effects [Case:Nestlé/Perrier]:
Factors that facilitate collusion: [ have to be taken into account to establish the likelihood that the merger will have pro-collusive e$ects]
Concentration, entry barriers, structural links such as cross-ownership and joint-ventures, agreements about information exchange, multi-market contacts, regularity of market inter- actions, the absence of countervailing power, and the existence of clauses (such as best-price clauses and retail price maintenance) that increase observability of firms’ actions.
If there exists little risk of increased collusion, the merger will be cleared. Otherwise, it might be cleared only if remedies can ensure that collusion will not be likely to occur after the merger.
Nestle case: The EC rightly thought that with the acąuisition by Nestlé of Perrier, and the transfer of Volvic to BSN, a dominant position would be jointly held by the two firms, due to their symmetric situation in the industry and to the market environment which was favourable to collusive outcomes.; However, the merger with Perrier, plus the transfer of Volvic to BSN, were allowed under the condition that Nestlé would have sold to a third party the brands Vichy, Thonon, Pierval, Saint-Yorre as well as some minor local spring waters. According to the EC, these waters would represent a capacity of around 20% of Nestlé, Perrier and BSN together, even though the market shares of such brands are not very high
In my opinion, the EC should have blocked the merger tout Gourt. The char- acteristics of the industry (high concentration, high price observability, symme- try, a history of parallel price increases) strongly suggest that the firms have been able to tacitly collude over time.
Further, the fact that Nestlé and BSN reacted immediately and of common agreement when an outsider like IFINT tried to enter the industry by taking over Perrier , is a clear indicator of the coordination between them. the incumbent firms had managed through time to coordinate themselves in such a way to reach a collusive outcome. When a potential entrant jeopardised the stability of this outcome, they reacted together to put an end to this threat.
Allowing the transfer of Volvic to BSN would only worsen matters, as it increases the degree of symmetry between Nestlé/Perrier and BSN, and facili- tates a collusive outcome (see Compte et al., 2002), that the remedy accepted is unlikely to disrupt.
ABB/Daimler-Benz
This case concerns the proposed joint-venture between Asea Brown Boveri (ABB), a Swedish-Swiss company, and Daimler-Benz, a German company, to form ABB Daimler-Benz Transportation. The joint-venture would incorporate all the activities of the parent companies in the sphere of rail technology…. EC declared the concentration between ABB and Daimler-Benz, incompatible with the common market in the product market of trams and metro vehicles, where a joint dominant position along with Siemens would have been created. The parties committed to divest from the AEG/Daimler-Benz’s sub- sidiary Kiepe, which possesses the electrical engineering technology which is a key element for supplying a complete product in the local rail technology. By maintaining this firm independent, other firms which possess the mechanical engineering technology will find an available partner for competing successfully in the local trains markets.
Vertical restraints and vertical mergers:
•What are vertical restraints?…agreements between vertically related firms
eg. the vertical relationship between a manufacturer and a retailer…..each party’s actions create an externality on the other. Vertical contracts try and control for these externalities….For instance, the manufacturer would like the retailer to make a lot of marketing ….so, The manufacturer might use contractual provisions (that is, vertical restraints) for higher marketing
when do these vertical restraints (and vertical mergers) , create positive or negative ew .?.
both vertical restraints and vertical mergers might have anti-competitive effects, by foreclosing com- petition. For instance, an incumbent firm might use exclusive contracts to pre-empt efficient entry into an industry; and a merger might allow a vertically integrated firm to foreclose an input to its downstream rivals, thereby reducing their competitiveness and possibly forcing them to exit the industry.
We should balance the efficiency and anti-competitive effects of vertical restraints and vertical mergers.
only the vertical clauses adopted by firms enjoying large market power are worth investigating, because onky these may raise welfare concerns
VERTICAL INTEGRATION
Suppose that some of the competing retailers are located very close, and it is not too costly, relative to the value of the good, for the consumer to do a little search before a purchase.
In these circumstances, each shop will think twice before investing a great deal of effort to sell the brand. This is because another shop nearby would have an incentive to avoid the cost of this effort, just a free ride
Vertical restraints restore incentives for the retailers to invest in services. For instance, suppose that the producer divides the city in different areas, and appoints an exclusive distributor in each area (exclusive territories). This would reduce the possibility of consumers visiting several shops (it is more costly to shop around in different areas) and therefore reduces the risk that a retailer will be undercut by a free-riding competing shop. Hence, each retailer will have a higher incentive to offer brand supporting services. Another possibility is for the producer to maintain all the shops in the city, but fix the resale price, or impose a price floor, to avoid the problem of undercutting and to allow the retailers to recoup (part of) the investment.
Vertical integration would also solve the problem: if the producer owned the shops, it would take into account the externality that each of them imposes on the other, and would prevent its shop managers from undercutting each other and reducing the level of services they provide.
To sum up, vertical restraints and vertical integration avoid, or reduce the free-riding problem, to the benefit of both producer and consumer surplus.
We have just seen that vertical integration and vertical restraints might improve welfare by raising the level of effort and services provided by retailers.
Competition policy should recognise the degree of substitutability among different vertical clauses
Eg it is useless to block, say, exclusive territorial clauses, while permitting, say, resale price maintenance clauses which allow firms to reproduce a very similar outcome. (And vice versa: permitting ET but outlawing RPM.)
vertical integration is welfare improving, when there is ”increasing preference for variety“[= where ”at low levels of total consumption, a consumer cares less about variety increases]
Horizontal and vertical externalities:
-under exclusive dealing (ED):
(i) investment levels are higher. This is due to the increased appropriability of the investment. In turn, this reduces the cost of distributing the brand;
(ii) wholesale prices are higher. This is because lower distribution costs shift outwards the marginal revenue function of the firm, which can then increase wholesale prices;
(iii) the retail price is lower (due to the dominant effect of the reduction in costs);
(iv) manufacturers’ profits are higher (retailers’ profits are always nil due to the Bertrand competition assumption);
(v) finally, welfare is higher than under NED, since profits are higher and consumers are better off.
In this model, therefore, exclusive dealing has a welfare improving e ect, and banning it would decrease both consumer surplus and manufacturers’ profits.65
vertical mergers and vertical restraints that affect intra-brand competition only are mostly efficiency-enhancing. They allow firms to control for externalities that affect the vertical relationship with other firms, thereby increasing profits of the vertical chain, as well as consumer surplus.
vertical restraints which affect intra-brand competition, do not raise many welfare problems; certainly, they are not worth investigating when firms that adopt them do not have high market power.
vertical restraints are often substitutable with each other. Accordingly, differential treatment of vertical restraints (for instance, allowing some and forbidding others) is not justified.