Following "Akzo 1985" European Competition Commission antitrust duty, (Official Journal L 374, 31/12/1985, p 1-27), explain the circumstances under which the elimination of a competitor by under-cutting its price is considered as unfair.
Industrial Economics
7th February 2004
Russell Manley
Following "Akzo 1985" European Competition Commission antitrust duty, (Official Journal L 374, 31/12/1985, p 1-27), explain the circumstances under which the elimination of a competitor by under-cutting its price is considered as unfair.
"Predatory pricing schemes are rarely tried and are even more rarely successful."
There is still much debate on the subject of Predation or Predatory Pricing. Economic analysis and antitrust laws have largely developed separately from one another. In more recent times antitrust laws have been following the more extensive research into this subject and hopefully a more cohesive unambiguous definition of predatory pricing and the associated legalities can be presented. It is the most difficult area of competition and antitrust policy and as highlighted later there are disagreements between decisions on different cases and about general principle. The first issue to assess is whether predation exists. Often it can be confused with a competitive response to the threat of new entrants and therefore analysis will take place of its existence in theory and practice. Then analysis will take place on the welfare effects of predation and the legal ramifications to see why exactly it should be unlawful. Finally, we will look at if it is possible to differentiate between predation and competitive behaviour and throughout we will look at the previous and current legal cases and outcomes to identify when predation is actually illegal, and more precisely when a successful case can be brought against a predator.
Predation is in essence a simple concept that unfortunately is easy to show in theory and nigh on impossible to prove in practice. If an incumbent monopolist (I) cannot prevent entry into the market then it has the alternative of inciting or forcing the exit of rivals or new entrants (E) to gain monopoly power. This process is known as predatory pricing. This type of pricing involves pricing below cost to force out competitors and once this has been done, the monopolist will then increase the price to gain the monopoly profit ( PM ). The best way to illustrate predation is through examples. KLM was the market leader in flights between London and Amsterdam. EasyJet (EJ) then started flying on the same route and as their business plan/policy dictated, they ran it on very low cost and therefore very low price. KLM as market leader (40%) responded by cutting their prices that led to losses for EJ. It can be assumed at this point that KLM had the explicit desire to induce exit of EJ from the London to Amsterdam market. This could be done in one of two ways, dependant on the costs of KLM and EJ. "Limit" pricing would be KLM pricing below EJ cost but above their own costs. This is legal and just results from greater efficiencies and economies of scope and scale. If however, they priced below EJ cost and below their own cost then that would be predatory pricing and is considered illegal under Article 86, Treaty of Rome1. EJ having faced this price undercutting has choices that they can make immediately and in subsequent periods but this is based entirely on the information that it has regarding KLM's costs.
The first question that EJ must attempt to answer is whether the threat of predatory pricing is credible, as this will determine the outcome and profits/losses in subsequent periods of the game. The game has 2 periods and there is an incumbent monopolist (I) and a new entrant (E).
In t = 1 I decides to set price (p) ??low, high?
If p = low I and E make a loss = -L
If p = high I and E make duopoly profit = PD
Therefore, E has the decision to stay or exit the market d ??stay, exit? at end of the period.
In t = 2 If d = exit then I earns monopoly profit = PM
If d = stay then same decision as period 1
The Nash perfect equilibrium is then for I to set the price high at t = 1, E then stays and at t = 2, I sets the price high again.
Extensive form of the game of perfect information:
I
High Low
E E
High Low High Low
I I I I
High Low High High Low High
(2PD , 2PD ...
This is a preview of the whole essay
In t = 2 If d = exit then I earns monopoly profit = PM
If d = stay then same decision as period 1
The Nash perfect equilibrium is then for I to set the price high at t = 1, E then stays and at t = 2, I sets the price high again.
Extensive form of the game of perfect information:
I
High Low
E E
High Low High Low
I I I I
High Low High High Low High
(2PD , 2PD ) (PD - L, PD - L) (PD + PM , PD ) (-L + PD , -L + PD ) (-2L , -2L) (-L + PM , -L)
Assuming PD > L.
The main problem when facing a situation such as this is how do we know, or more specifically how can it be proved that KLM lowered its prices with the explicit aim of forcing EJ out of the market?
As classical economic theory tells us the more firms in a market the lower the price. As we see from a Bertrand Duopoly in an homogenous product market with no collusion, simply going from one monopolist to two firms implies a price drop from monopoly price to a price equal to marginal cost. Therefore, it is possible that KLM was just responding to a new entry threat with a shift in equilibrium, i.e. competitive behaviour. It can be seen that there is a very fine line between competitive (low margins) and anticompetitive (pricing below cost) behaviour and this blurredness of boundaries is central to the problem of establishing the existence of and ruling against predatory pricing.
As can be seen from the extensive form diagram a rational entrant, E, will never exit the market if preyed upon and a rational incumbent, I, should never engage in predation (Chicago School Theory). E will make a loss in period 1, if I acts aggressively, but if it stays then I has little choice between -L and PD in period 2. So optimal solution for I in period 2 is to not be aggressive and set the high price. If PD > L then both E and I are happy. However, E must be able to cover the loss in the short term, if it cannot it will have to borrow. According to the Chicago argument, no rational predator would engage in predation and therefore it never happens in practice and any price-cutting is competitive responses to decrease in concentration. This is a little extreme and relies heavily on perfect information.
The main problem for I is to make the threat of predation credible. This is particularly important in cases where E has post-entry Nash Equilibrium positive profits. If this is the case, then, again, passive responses from I are better then aggressive ones. This means that to make the threat credible I must commit resources so that is entry does occur it can make sure that E gets post entry negative profits. This use of resources instead of price is predatory "behaviour" not pricing. For predatory pricing, to be rational there must be additional factors. If there is imperfect information then a firm may engage in aggressive price-cutting as a response to a new entrant in four possible ways/reasons. These prove that theoretically and in practice, predation does occur and can be a sensible strategy if undetected.
DEEP POCKET THEORY: Returning to the extensive form model, another issue that must be tackled is, what if the bank won't lend E the money to cover - L in the short term or more specifically that E expects that with probability p, the bank will refuse the loan. The rational thing for E to do is still to stay if it knows that profits at t = 2, PD are greater than the loss - L it will experience at t = 1 and the probability of the loan is taken into account. So E would stay if (1-p) PD > L. You would expect I to want to act passively and accept E as it will receive 2 PD and no losses. However, as there is now doubt over E being able to cover - L, it may be optimal to be aggressive and price low at t = 1. If it does it will I incur a loss of - L in period 1 but in period 2 it has probability p that it will get PM and 1-p probability of getting PD . So I will be predatory if PM > L + (1-p) PD . If both those conditions are satisfied then we can see how predation could occur in practice with I wanting to act aggressively and E wanting to resist such predatory behaviour. In addition, in p% of the time predation is a successful way of forcing a new entrant out of the market. Of course, this is entirely dependant on the financial constraints/benefits of each of the firms, but does show a highly probable situation.
LOW COST SIGNALLING: This is when I would lower its price to signal to E that its costs are low and therefore this no room for any additional firms to make any money in this market. The best example of this is the American Tobacco Company. American Tobacco Company: 1891 - 1906 American Tobacco acquired 43 small competitors, thus establishing a quasi-monopoly. By imposing predatory prices on the market before acquisition, this imposed losses on target firms and AT could buy tem at a greatly reduced price. Some reports suggest for anything up to a 60% saving. This was because the targeted firms saw AT's apparent low prices and assumed they had low costs and therefore did not bother to compete and accepted a low bid for takeover.
REPUTATION FOR TOUGHNESS: Aggressive pricing is a possibility for non-homogenous products or for pre-entry tactics. I will set prices low and therefore gain a reputation for being tough this will deter E's from staying in the market, also it will deter new entrants from entering in the first place and also will have a knock on effect for the same firm in the different markets in which it is a member. . Milgrom and Roberts show that for predatory pricing to be rational, E has to be uncertain about playing a post entry game, usually due to imperfect information on I. If this is the case then it is rational for I to use aggressive pricing policies against E and thus gain the reputation.
GROWING MARKETS: Very often growing markets require significant market share early on to ensure long-term survival. Predatory pricing can be used to ensure this market share. A good example of this was in the 1980's the battle for operating systems.
As can be seen there are many situations theoretically and in practice that show Predation is a real concept and one that merits investigation both economically and with a view to competition and antitrust laws. Therefore, if it does exist the next question is what are the welfare costs and why would it be deemed illegal?
Predatory pricing implies that E will exit the market and leave I with monopoly power. So against future high prices we must weigh short run lower prices. However, these are only relevant when I's price cut is selective. This was highlighted in the case of Sacramento Cable Television (SCT): In 1983, SCT was given a franchise in the Sacramento area. In 1987, another franchise was awarded to Cable America (CA). CA decided to only set up in 700 homes hoping to expand at a future date. CA's price reflected favourably against the larger SCT.SCT decided to cut their price below that of CA, but only in the 700 homes that CA were already based in, leaving the prices in other areas at the previous higher level. CA only lasted 7 months before they finally exited the market unable to compete with SCT. If we go back to the KLM v EJ case, despite KLM dropping their prices EJ stayed in the market therefore both lost the price war and the only winner was the customer. Another market where the effects of predatory pricing can be mixed is ones with Network Externalities. If we come back to the battle in the 1980's for operating systems. Apple and Microsoft fought for control of the market, if predatory pricing was stopped then it may save a competitor from exiting the market but it leads to mush less standardisation which for the future of the industry and the utility of the consumer is a negative effect.
With the emergence of a new entrant into a market, prices will naturally fall and due to economies of size and scope and efficiencies and this may unwittingly, on the part of the incumbent, cause exit of the entrant. This was the case in the early 1980's in the UK, where there was a proliferation of new petrol stations and prices dropped and this forced many to close, the courts ruled there was no case for predation. Easly, Masson and Reynolds give a full analysis of a case where E is unable to determine whether a cut price response by I, to new entry is a matter of superior cost efficiency or predation. So how can we tell when it does happen? The United States since 1975 have used the Areeda-Turner test. This simply states that a pricing policy is predatory if prices fall below the marginal cost of the incumbent. This seems to make sense as no firm should put prices below marginal cost for any other reason that to eliminate competition. However, it is a very limited test as there are many cases where firms will in fact lower prices below marginal cost for many different reasons. Firstly, a firm could put p < MC for the purpose of moving down their learning curve and not for predation. In addition, if E makes a mistake and builds new capacity on too large a scale then the ensuing price war could push p < MC, in a struggle to stay in the industry. Predatory pricing will be even more difficult to detect in such cases of investment in capacity as only detailed market knowledge can reveal if capacity was excessively expanded for purposes of predation.
It was not until 1986 that a stricter test was found. This was in the Matsushita case (Matsushita Electrical Industrial Co Vs Zenith Radio Corp. 475 U.S 574 (1986)). Matsushita was alleged to have been using predatory prices to sell televisions in the USA for twenty years. The court was willing to acquit Matsushita on theory alone. They presumed that no rational firm would willingly price below MC for a period of twenty years and therefore there was no threat of predation. Similar logic was used in the case of Brown and Williams (Brooke Group Vs Brown and Williams 125 L.ed 2d 767 (1993)). This case involved generic brands of cigarettes. Brown and Williams set their prices below MC but because of the market's complex oligopoly conditions, the chance of recouping the losses in subsequent periods was so low that predation was ruled out. In addition, the impacts on the plaintiff had not been fatal and the court decided that competition had not been harmed. The precedent from these two cases is that any alleged victim of predatory pricing now has the burden of proving the defendant priced below marginal cost AND later recouped those previous losses AND prove a harmful impact on itself or competition as a whole in the market. At present, although the US courts still rely on simple cost-based tests, they require evidence of the possibility for recoupment after predation, and in some cases of intent, also. The extra conditions have created a great level of ambiguity between the courts and the implementation costs have risen. Consequently, fewer and fewer cases of predation have been proven legally, ironically just as the concept has been modelled as an economic possibility.
In practice, the US supreme court only clarifies the first two as necessary conditions for "illegal" predatory pricing. However, the US legal system has for a long time shown disregard for cases of predation. They are usually dismissed due to the consensus view expressed in the opening quotation. Analysis here has shown that predation does exist and is on many occasions the rational choice. It still remains very difficult to establish credible difference between competitive and anti competitive behaviour coupled with the ambiguous welfare effects make the legal perspective difficult to gauge. The courts are wrong when they imply that predation is an unlikely event, but overall they have the right convictions to not giving predation cases much weight, but it is more than likely for the wrong reasons.
Europe in general, has less of a tradition of predation than North America. The legal standing of predation in Europe is based around the abuse of a dominant position under article 86 in the Treaty of Rome. As mentioned in the title of this essay one of the major precedents, set was in 1985 when a ruling was made against Akzo Chemie BV. AKZO Chemie BV v Commission [1991] ECR I-3359. Akzo was fined for predatory pricing against ECS. The basis of this ruling was intent as opposed to actual evidence of price-cutting below marginal cost. ECS had evidence that on separate occasions Akzo threatened them with aggressive predation if they did not exit other markets that the two were both engaged in. The European commission put a lot of weight on this and, in fact, no investigations took place into whether Akzo priced below cost or imposed any significant damages on ECS. This case highlights the difficulties of decisions of predatory pricing. This evidence of intent, which policy makers take very seriously, means that the decision was come to very easily. This means that in current terms predatory pricing can be prosecuted if there is considerable evidence of intent to back it up. If not then cases very rarely are successful. A good example of this is ADM. In 1995, Archer Daniel Midland was sued by over seventy firms for alleged predation. As in most private cases, most companies use evidence of conspiracy and sue on strong evidence of guilt. Nevertheless, all the incumbent has to do is deny all charges and therefore force the entrant to gather its own information on the actual guilt and this can be very difficult to gather precise cost information on another company. Therefore, it can be seen that unless a mistake is made like Akzo and actual evidence of guilt are available then the onus is transferred from the incumbent to the entrant to show predatory pricing and as this is very difficult, cases of predation are very rarely successful.
. Article 86
"Any abuse by one or more undertakings of a dominant position within the common market or in a substantial part of it shall be prohibited as incompatible with the common market in so far as it may affect trade between Member States."
(a) Directly or indirectly imposing unfair purchase or selling prices or other unfair trading conditions;
(b) Limiting production, markets or technical development to the prejudice of consumers;
(c) Applying dissimilar conditions to equivalent transactions with other trading parties, thereby placing them at a competitive disadvantage;
(d) Making the conclusion of contracts subject to acceptance by the other parties of supplementary obligations that, by their nature or according to commercial usage, have no connection with the subject of such contracts.