Capítulo 5: Cartéis e Oligopólios

c h a p t e r 5 146 Cartels People of the same trade seldom meet together even for merriment and diversion but the conversation ends in a conspiracy against the public or in some contrivance to raise prices It is impossible indeed to prevent such meetings by any law which either could be exe cuted or would be consistent with liberty and justice But though the law cannot hinder people of the same trade from sometimes assem bling together it ought to do nothing to facilitate such assemblies much less to render them necessary Adam Smith In any market firms have an incentive to coordinate their production and pric ing activities to increase their collective and individual profits by restricting mar ket output and raising the market price An association of firms that explicitly coordinates its pricing or output activities is called a cartel A cartel that includes all firms in a market is in effect a monopoly and the member firms share the monopoly profits Cartels are more likely to occur in oligopolistic markets where there are only a few firms than in competitive markets It is easier to reach and maintain an agreement on price or output when the number of firms is small But even with out an explicit agreement firms may act so as to raise their collective profits The factors that increase the likelihood that a cartel succeeds also affect whether oli gopoly firms can elevate their prices above the competitive level Thus a study of cartels is also a study of oligopolies In the next chapter we use game theory to analyze oligopoly behavior 3641AWLCARLCh05pp146180qxp 12715 1249 PM Page 146 Why Cartels Form 147 Fortunately for consumers although firms have an incentive to coordinate activities to restrict market output and raise prices each member of the cartel has an incentive to cheat on the cartel agreement Each cartel member wants to produce more output than is best for the cartel collectively As a result cartels tend to break apart even with out government intervention When a cartel partially breaks apart so that some firms act outside of the cartel or when not all firms in the market join the cartel in the first place the cartel may act like a dominant firm facing a competitive fringe of nonmember firms As discussed in Chapter 4 entry of new fringe firms into a market can destroy the market power of a dominant firm or a cartel Thus only cartels that do not fall apart through lack of co operation and that exist in markets in which entry is difficult can maintain market power for substantial lengths of time Four key questions are examined in this chapter 1 Why do cartels form 2 What factors cause some cartels to last and others to break up even without government intervention 3 How harmful are cartels 4 What have governments done about cartels Why Cartels Form United we stand divided we fall Union gives strength Aesop Why is Adam Smith correct that firms want to form cartels The answer is that each individual firm wants to increase its own profit But why should a firms profit go up when the firms in a market form a cartel After all each competitive firm is maximiz ing its profit How can the firms do better by forming a cartel if each is already maxi mizing its profit The answer involves a subtle argument In a competitive market each firm consid ers how much a reduction in its own output benefits it and ignores the gains to other firms which benefit from a reduction in total market output to the extent that reduc tion raises the price In contrast a cartel takes into account the benefits to all its mem bers of the reduction in each firms output Thus a competitive market in which each firm ignores the collective gain from its output reduction produces more output than a cartel To illustrate the nature of this collective gain consider two polar cases First sup pose that a market is made up of many identical competitive firms each of which is a price taker In contrast suppose that all the firms join together to form a cartel and act as a monopoly Figure 51a shows a typical firms marginal cost curve The sum of the individual firms marginal cost curves is the market supply curve which is shown in 3641AWLCARLCh05pp146180qxp 12715 1249 PM Page 147 148 Chapter 5 Cartels a Typical firm b Industry MC MC Poor fee teen By Market vo demand Im co Firms quantity g On4m Qe Industry quantity O Figure 51b labeled C along with the market demand curve The competitive out put Q is determined by the intersection of this supply curve with the market de mand curve Figure 51b with each firm producing gq units of output Figure 51a and the market price is p Why does it pay for the cartel to reduce output from the competitive level At the competitive output the cartels marginal cost is greater than its marginal revenue Fig ure 51b so it pays the cartel to reduce its output Because the demand curve slopes downward the marginal revenue curve lies below the demand curve and marginal revenue is less than marginal cost at the competitive output Q Thus it pays for the cartel to reduce output from the competitive levelbut by how much It should lower output until its marginal revenue equals marginal cost which guarantees that profits are maximized The cartel increases its profits by lowering the aggregate cartel output to Q where MR equals MC Figure 51b The price rises to p Because the cartel is made up of 7 identical firms it requires each firm to reduce its output to g Qm In this example the identical firms share in the extra profits equally Why doesnt each competitive firm reduce its own output below the competitive level At the competitive equilibrium each competitive firm sets its marginal revenue TAs with a monopoly the cartel can restrict output and let the demand curve determine price or raise price and let the demand curve determine output The two approaches are equivalent Creating and Enforcing the Cartel 149 2Because a pricetaking competitive firm faces a horizontal demand curve its marginal revenue curve is also horizontal and identical to the demand curve Thus the competitive firms MR curve is horizontal at the competitive price pc where the market supply or MC curve hits the market demand curve in Figure 51b 3See the discussion in Chapter 3 of the elasticity of demand facing a single competitive firm and Equation 31 4An externality is a good or bad that is not priced by the market equal to its marginal cost and has no incentive to further lower its output2 If it were to reduce its output by one unit it would lose profits because the marginal revenue on the last unit produced the price would exceed its marginal cost Thus each competi tive firm is maximizing its profits at the competitive output The gain in collective activity comes from the very slight slope to the competitive firms demand curve Although economists often say that each competitive firm acts as though it faces a horizontal demand curvethat it cannot raise its price by lowering its outputthat is not absolutely correct The demand curve does have a slight slope A competitive firm that stops producing might raise the market price by a small amount3 That small slope can be ignored when talking about a single firm but it can not properly be ignored when talking about all firms collectively If all firms cut back by say 10 percent the market price definitely rises however if only one firm cuts back by 10 percent the effect on price is so small that it is hardly measurable Each competitive firm decides that it doesnt pay to reduce its output sig nificantly because its gain is less than its cost If it reduces its output by one unit its gain is the trivial amount by which price rises times the units it produces whereas its loss is the price it would have received for this last unit A competitive firm ignores the good it does other firms by reducing its output and increasing the market price it places no value on the gains of other firms This gain by others is an externality4 Working cooperatively the cartel members gain from the out put reductions of each firm When all firms belong to the cartel all the gains from re ducing output and raising price go to the cartel which divides the gains among its members Here the externality created by each firm in reducing its output has been internalized by the cartel As a result it pays the cartel to reduce total output below the competitive level even though it would not pay any competitive firm to reduce its output individually Creating and Enforcing the Cartel Socrates Tell me whether you think that a city or an army or a band of robbers or thieves or any other company which pursue some unjust end in common would be able to effect anything if they were unjust to one another Thrasymachus Of course not Socrates When we say that any vigorous joint action is the work of unjust men our language is not altogether accurate If they had been thoroughly unjust they 3641AWLCARLCh05pp146180qxp 12715 1249 PM Page 149 150 Chapter 5 Cartels could not have kept their hands off one another Clearly they must have possessed justice of a sort enough to keep them from exercising their injustice on each other at the same time as on their victims For the thorough villains who are perfectly un just are also perfectly incapable of action Would you join a cartel if all the other firms in your market were forming one Such behavior is usually illegal in the United States and many other capitalist countries so no doubt you would refuse on moral and legal grounds Suppose it was not a moral person like you who was being asked this questionsuppose it was your slightly shady cousin Would your cousin join an illegal cartel conspiracy Well it depends Your cousins first thought is likely to be Whats in it for me It should be obvious to him that it is in his best interest to let all the other firms in the market form a cartel that does not include his firm Then the cartel would restrict out put driving up the price while his firm could produce as much as it wanted Of course every other firm in the market makes the same calculation Now suppose the other firms tell him that unless his firm agrees none of the others will join the cartel and restrict output Your cousin now realizes that he cant have his cartel and produce as much output as he wants too He can only obtain the higher price if his firm agrees to a reduction in output Your cousin then thinks What do I have to lose If the cartel is caught by the gov ernment and convicted my firm will have to pay a fine But if the chance of being caught is small or the fine is low it may be worth it to me That is if the expected loss from such a fine is low enough your cousin joins the cartel But your cousin is always looking for an edge Once hes joined the cartel he says to himself Why shouldnt I cheat and produce more output than the cartels agreement permits After all the cartel probably wont know whos producing the extra output Of course if all firms in the cartel think this way the cartel will fall apart The success of the cartel then turns on its ability to enforce its agreement Figure 51 illustrates why a firm has an incentive to cheat on the cartels agreement As explained above the cartel members agree to restrict output to Q which drives the price to p the monopoly price Figure 51a shows the cost curves of your cousins firm which is one of 7 identical firms in the market and in the cartel The cartel wants your cousin to produce q Qm output the output corresponding to his firms share of the cartel output But at the cartels price p your cousins firm can maximize its profits by producing g units of output where its marginal cost curve equals p Thus although it is in the cartels best interest for every firm to restrict out put it is in your cousins best interest for every firm except his own to restrict output Cartels have little effect on prices if members do not cooperate For example in Kuala Lumpur representatives of four pepperproducing countries decided that they would set a minimum price for black pepper Even though the pepper cartel Brazil India Indonesia and Malaysia produces more than 95 percent of the worlds pepper Based on Plato 1957 378 The names of the speakers have been added and some material has been dropped Creating and Enforcing the Cartel 151 6Chaos in the Cartel Pepper Producers Pick a Purchasers Price San Francisco Chronicle August 8 198349 7Generally price fixing is discussed rather than output reductions because it is believed to be more common Other firms prices are sometimes easier to observe than their output levels In a study of antitrust cases from 1890 through 1969 Posner 1970 especially p 424 found that only 16 percent of the cases had only explicit production or sales quotas 8The use of the terms tacit collusion and conscious parallelism has led to confusion and added fuel to legal disputes One ambiguity in the terms arises because the oligopoly price lies between the competitive and monopoly prices Although many economists use these terms to mean that the oli gopoly price is at the monopoly or cartel level others use them to refer to the case where the non cartel price is elevated above the competitive price A similar ambiguity arises with the term collusion Economists often use the word to refer either to nonconspiratorial behavior lawful actions where firms do not engage in explicit price fixing or to conspiratorial behavior illegal explicit price fixing but lawyers commonly use the term to refer only to conspiratorial behavior A separate ambi guity in the law exists around the meaning of the word agreement see Carlton Gertner and Rosen field 1997 and could raise the price it has never been able to do so because its members keep un dercutting the cartels minimum price6 Factors That Facilitate the Formation of Cartels Once a cartel forms the firms must agree to fix price or equivalently reduce output if it is to be successful7 Why are there successful cartels in some markets but not in others Unfortunately we know a great deal about cartels that get caught but very lit tle about those that escape detection As a result it is not known whether the cartels that find themselves in court are unsuccessful or merely unlucky Some evidence sug gests that cartels that end up in court are actually unprofitable and hence perhaps atypical Asch and Seneca 1976 Other evidence Suslow 1998 suggests that cartels tend to form in less profitable industries Many characteristics of markets and firms that contribute to successful pricefixing conspiracies have been identified using studies of cartels that have ended up in court Stigler 1964a Hay and Kelley 1974 These characteristics may be roughly divided into those that allow a cartel to raise the market price in the first place and those that prevent the cartel agreement from breaking apart due to cheating by members The following sections describe some of the major factors that facilitate the formation of cartels Factors that lead to their survival are discussed in the next section See Example 51 for a description of one of the most important cartels in American history Large firms may decide independently to behave as though they had a cartel arrangement without a formal meeting that is each one can cut its output and hope that the others will do likewise Inevitably in oligopolies firms take their rivals actions into account as discussed in more detail in Chapter 6 When firms in an oligopoly coordinate their actions despite the lack of an explicit cartel agreement the resulting coordination is sometimes referred to as tacit collusion or conscious parallelism8 Stigler 1964a explains that a theory of oligopoly could be based on cartel theory even in the absence of explicit agreements 3641AWLCARLCh05pp146180qxp 12715 1249 PM Page 151 152 Chapter 5 Cartels An Electrifying Conspiracy On a steamy Saturday May 9 1959 at 230 PM reporter Julian Granger of the Knoxville NewsSentinel sat at his desk reading routine handouts from local publicity sources The usual weekly newsletter from the Tennessee Valley Authority TVA an nounced that several contracts had been awarded Pretty dull stuff But then Granger read that one contract had been awarded to Westinghouse for transformers for 96760 The newsletter went on to note that AllisChalmers Gen eral Electric and Pennsylvania Transformer quoted identical prices of 112712 How could three companies bid prices that were identical to the penny under a sys tem of secret sealed bids Later in the release he read that two other companies had quoted identical prices on a 273200 contract On yet another contract for conductor cable there were seven identical bids of 19843824 Identical down to the 24 Grangers story which appeared on page 25 of the second section of the Knoxville NewsSentinel on May 13 1959 drew little reaction The lead of his second story on May 17 1959 stated TVA purchasing records revealed today that at least 47 large and small American manufacturers have taken part in identical bidding on a wide va riety of items in the past three years He noted that the TVA had no choice in many cases but to award the contract on a chance drawing from a hat Could these identical bids have happened by accident Oh come now Granger showed that some identical bids quoted the same delivered prices even though dis tances from delivery points varied by hundreds or thousands of miles for equipment you couldnt exactly weigh on a bathroom scale General Electric and Westinghouse the two biggest electrical equipment manu facturers had equal bids more frequently than other firms Between 1946 and 1957 they raised prices 10 times on switching gear in a parallel pattern The announce ments of these increases came within a few days of each other After the second story Granger went to local suppliers and got nowherethey seemed afraid to talk Granger did learn however that the Knoxville Utility Board had received a long series of identical bids from the electrical equipment manufactur ers up to 11 at one time Their purchasing agent Karl Strange told Granger that he had noticed an increase in the practice since the end of World War II The ScrippsHoward papers reprinted the first two articles nationwide On May 19 Senator Estes Kefauver inserted Grangers second story verbatim in the Congressional Record of the hearings before the Subcommittee on Antitrust and Monopoly Eight days before Grangers first story broke Ralph J Cordiner Chairman of the Board of General Electric GE testified before the subcommittee on a bill to pro mote more vigorous competition in a variety of industries He asked Is it assumed that companies in industries affected by this bill have the ability to administer prices in a manner not responsive to market supply and demand He responded to his own question If so the assumption is false because these companies are just as much subject to competitive market conditions as any others He continued In all instances the prices are completely subject to the force of competition in the market Example 51 3641AWLCARLCh05pp146180qxp 12715 1249 PM Page 152 Creating and Enforcing the Cartel 153 place and the value the customer believes he is receiving He concluded that the thencurrent antitrust laws were well enforced Exactly six months before Cordiner testified so sanctimoniously about competi tion seven of his top executives met with their competitors in the Hotel Traymore in Atlantic City New Jersey to jack up the price of power switchgear assemblies 125 million in annual sales Until that time the firms had an agreement that their sealed government bids would be divided by each of the conspiring companies in the fol lowing ratios General Electric 42 percent of the market Westinghouse 38 percent AllisChalmers 11 percent and ITE 9 percent Apparently the initial conspirators made room for a major new entrant at this meeting Federal Pacific the newcomer to the cartel was given permission to quote prices slightly lower than the others for a brief period in order to establish its assigned share of the market GE and Westinghouse agreed to lower their shares to 39 percent and 35 percent respectively giving Federal Pacific 7 percent of the market Over the next 12 months at least 35 such meetings were held with GE playing a prominent and active leadership role Senator Kefauver the chairman of the Subcommittee on Antitrust and Monopoly moved the hearings to Knoxville in September 1959 The hearings demonstrated that the prices of heavy industrial electrical equipment had increased 50 percent since 1951 At the hearings many examples of identical bids were presented Even when the bids were not identical the companies followed a rotation pattern where one bid was low one high and the other two identical The next time around the order of the firms might change but there would be one low one high and two equal bids Presumably the firms took turns winning the bidding in the proportions agreed to earlier Appar ently they agreed to alternate winning the bidding according to a phase of the moon for mula The firm to give the low bid was determined by the fullness of the moon The Philadelphia Antitrust Office spent 18 months tracking down evidence used to indict 29 manufacturers practically the entire heavy electric industry and 44 of their top executives Attorney General William P Rogers issued the first official an nouncement of the indictments on February 16 1960 According to the first indictments returned by a federal grand jury sitting in Philadelphia Westinghouse AllisChalmers Federal Pacific Electric Company GE ITE and many of these firms top executives had engaged in a conspiracy at least since 1956 The defendants were accused of fixing and maintaining high prices allo cating the business among themselves submitting noncompetitive rigged bids re fusing to sell equipment to other manufacturers of electrical equipment or raising prices to them so that they could not effectively compete Conspiracy in this industry was facilitated by the relatively small number of firms and the large market shares of the largest firms Electrical manufacturing had a four firm concentration ratio sum of sales by the four largest firms divided by total indus try sales of over 50 percent compared to about 25 percent in all manufacturing The concentration ratios were above 75 percent in all specific product areas involved and over 95 percent for turbogenerators power transformers power switchgear assem Continued 3641AWLCARLCh05pp146180qxp 12715 1249 PM Page 153 154 Chapter 5 Cartels blies distribution transformers lowvoltage power circuit breakers isolated phase buses bushings and lightning arresters Eventually 45 executives and 29 corporations were indicted Most of them made no defense in the face of this overwhelming evidence Generally vice presidents and division managers took the fall The seven men with the highest positions received jail sentences but those were typically only on the order of 30 days In addition 24 people received suspended sentences Total fines to firms reached nearly 2 million while individual fines were 137500 In addition to the government suits nearly 2000 private suits were filed General Electric settled its lawsuits for over 200 million including 674 million to the TVA and 1 million to other federal agencies and Westinghouse for more than 100 million Total damages paid by all companies topped 400 million Many articles and books presented this story as a triumph of the system over evil conspirators but this conclusion is hard to understand Many if not most of the top directors of these firms were not personally indicted or punished and the total fines were a small fraction of the monopoly profits the cartel earned over its life span Sul tan 1974 1975 argues that the conspiracy did not significantly raise prices however Bane 1973 and Lean Ogur and Rogers 1982 find that prices did rise According to a US Congress report this longlasting conspiracy may have raised prices by nearly 10 percent Other estimates have been over twice that for specific products Electrical manufacturing accounts for about onetwelfth of total manufacturing and about 3 percent of all economic activity About 30 percent of this manufacturing was electrical apparatus which had 5 billion in shipments in 1958 The indictments re ferred to only about 175 billion of these annual sales about 10 percent of all electri cal manufacturing sales Even assuming that prices were only 10 percent too high on only 175 billion worth of annual sales purchasers paid roughly 175 million too much during each year of the conspiracy which apparently lasted for decades From the viewpoint of the firms involved this experiment in cartel behavior probably looked like a great success even after they were caught and punished The threat of penalties for illegal price fixing apparently was not viewed as a suffi cient deterrent by GE and Westinghouse These same firms have been charged and punished repeatedly since the Sherman Antitrust Act first went into effect in 1890 There were 13 US Department of Justice antitrust cases and 3 Federal Trade Com mission cases against GE and Westinghouse between 1911 and 1952 The govern ment won all of these cases obtaining convictions nolo contendere pleas or consent decrees in each Walton and Cleveland 1964 1620 Presumably the conspirators should have learned to be more careful as a byprod uct of the publicity if not the fines in the 1960s cases However GE and Westing house were accused of conspiring to fix prices on turbogenerators starting with a new pricing policy GE announced in May 1963 just two and onehalf years after they were found guilty in this bidrigging case Sources Fuller 1962 Walton and Cleveland 1964 and US Congress Joint Committee on Internal Revenue Taxation Staff Study of Income Tax Treatment of Treble Damage Payments under the Antitrust Laws Washington DC Government Printing Office 1965 39 cited by Posner 1975 3641AWLCARLCh05pp146180qxp 12715 1249 PM Page 154 Creating and Enforcing the Cartel 155 9If the noncartel price is close to the cartel price then firms may not believe that joining the cartel is profitable given the legal liability they potentially face from belonging to a cartel 10See the extensive discussion of the Organization of Petroleum Exporting Countries OPEC cartel at our web site at wwwawbccomcarltonperloff OPEC 11The Restrictive Trade Practices Act passed by Parliament in 1956 required that all contracts or agreements among suppliers in restraint of trade be reported to the Registrar of Restrictive Practices This law has been modified substantially since then In 1973 the Office of Fair Trading took over this responsibility and agreements among service industries had to be reported as well This agency was empowered to challenge agreements that were contrary to the public interest A special Restrictive Practices Court decides whether such agreements are prohibited In contrast to US law this court can accept the argument that benefits outweigh damages and allow price fixing A new Competition Act was passed in 1980 to facilitate investigations by the Office of Fair Trading Three major factors are necessary to establish a cartel First a cartel must be able to raise price above the noncartel level without inducing substantial increased competi tion from nonmember firms Second the expected punishment for forming a cartel must be low relative to the expected gains Third the cost of establishing and enforc ing an agreement must be low relative to the expected gains The Ability to Raise the Market Price Only if a cartel is expected to raise the price above the noncartel price and keep it high do firms join9 The more inelastic the demand curve facing a cartel the higher the price the cartel can set and the greater its profits If the cartels demand curve is inelastic relatively vertical at the current price raising price can significantly raise revenues that is quantity demanded falls by a smaller percentage than price rises and profits In contrast if a potential cartel faces an elastic demand curve relatively horizontal raising price causes revenues to fall because quantity would fall by more than price increases and profits may rise only slightly See Example 52 Entry by nonmember firms or close substitutes produced in other industries prevents a cartel from raising price If the cartel controls only a small share of the relevant market which includes all close substitutes firms not in the cartel undercut the cartel and pre vent it from raising the market price that is the demand curve facing the cartel is rela tively elastic Even if all firms initially in a market form a cartel and raise the price the higher price may induce enough new firms to enter that the cartel is unable to keep the price high in the long run That is the longrun elasticity of demand facing the cartel is very high especially relative to the shortrun elasticity Obviously the longer the car tel can expect to keep the price high the greater the current value of creating a cartel Low Expectation of Severe Punishment Cartels only form if members do not ex pect the government to catch and severely punish them Large expected penalties re duce the expected value of forming a cartel in the first place Before they were made illegal in the United States in 1890 explicit cartels were much more common During periods when the Department of Justice has been relatively lax in enforcing the laws pricefixing conspiracies have been more prevalent Posner 1970 Internationally where cartels are legal they have been more common than in the United States10 Some governments have created cartels as discussed below As long as coordinating firms did not use unlawful acts of violence intimidation or fraud British courts did not stop price fixing in modern times until 195611 A survey 3641AWLCARLCh05pp146180qxp 12715 1249 PM Page 155 156 Chapter 5 Cartels The Viability of Commodity Cartels Attempts have been made to cartelize the market for many of the major internation ally traded commodities Most of these initiatives have failed however as the cartels fell apart quickly or were unable to raise prices substantially Eckbo 1976 studied 51 formal international cartel organizations in 18 indus tries with the earliest agreement in 1918 and the latest in 1964 He defined a cartel as successful if it raised the price at least three times the marginal production cost of the member with the highest cost Only 19 cartels 37 percent were successful by this criterion One of them the iodine cartel lasted 61 years The remaining success ful cartels had formal agreements that lasted from 2 to 18 years with a median life time of 5 years and a mean of 66 years Only 5 of the 19 lasted 10 years or longer Of these successful cartels 3 out of 9 for which there is information broke down for nonmarket reasons such as government intervention or war Of those that col lapsed for marketrelated reasons 7 out of 16 44 percent had internal conflicts among cartel members whereas 9 56 percent ended because of external forces such as competition from nonmembers the usual case or reactions by buyers Two factors that allow a cartel to persist and raise prices are that 1 it can detect and prevent cheating by members and 2 it faces a relatively inelastic residual de mand curve at noncartel prices The cartels residual demand curve is likely to be in elastic if it has a relatively large share of the market the market demand is not very elastic and noncartel members have inelastic supply curves The longestlived cartel in Eckbos survey iodine 18781939 made all sales through a central cartel office in London which prevented members from cheating Maintaining a cartel is not sufficient for success however if it cannot raise prices The Organization of Petroleum Exporting Countries OPEC the International Bauxite Association IBA and the Conseil Intergouvernemental des Pays Exportateurs de Cuivre International Council of Copper Exporting Countries or CIPEC differ in their market power because of their different market shares and the residual demand elasticities they face OPEC quadrupled the world oil price initially IBA tripled the price of bauxite but CIPEC has been unable to raise copper prices significantly When OPEC was formed it had approximately twothirds of the worlds oil re serves and a similar fraction of the noncommunist worlds oil production By 1975 IBA accounted for 85 percent of total noncommunist world bauxite production In contrast CIPEC accounts for only about onethird of the noncommunist worlds copper production Of Eckbos successful cartels 15 out of 19 79 percent had four firm concentration ratios over 50 percent In 14 of them 74 percent the cartels share of total production exceeded 75 percent Of the 9 successful cartels about which we have enough information 7 faced in elastic demand curves elasticities less than 1 in absolute value In 8 of the 9 cases no shortterm substitutes for the commodity were available outside the cartel al Example 52 3641AWLCARLCh05pp146180qxp 12715 1250 PM Page 156 Creating and Enforcing the Cartel 157 though for 7 cartels there were longterm substituteswhich may be why they even tually ended Pindyck 1977 1979 shows that dynamic longrun adjustments in commodity markets are also important OPEC faces a relatively inelastic fringe supply Despite major price increases nonOPEC petroleum producers have not substantially in creased their supply in the short to medium run Similarly the world demand for bauxite is extremely inelastic up to a limit price even in the long run In contrast in the short run and even more so in the long run secondary copper which is produced from scrap is very responsive to price As a result CIPEC faces a much larger longrun elasticity than shortrun elasticity If CIPEC were to raise its price very much others would increase their production from scrap Given these differences it is little wonder that OPEC and IBA could raise prices while CIPEC could not These factors may also explain why still other natural re sources have not been successfully cartelized Pindyck holds that other minerals such as iron ore manganese ore lead tin zinc and nickel would also face high longrun residual demand elasticities due to secondary supplies from scrap Recently IBA has suffered setbacks because Brazil and other producers have not restricted output It continues as a research group Indonesia and Grenada produce 98 percent of the worlds nutmeg and agreed to form a nutmeg cartel in 1987 For the 15 months before the formal agreement how ever Grenada operated informally under an Indonesian guideline The two Countries claimed they did not intend to force prices higher but merely wanted to ensure that there is no price cuttingpresumably from the informal cartel level They were not worried about the impact of the cartel on demand Nutmeg has no close substitutes it has a distinctive taste so bakers are unlikely to change their recipes appreciably One long successful cartel is the De Beers diamond cartel which throughout the twentieth century has been the largestselling agent of most of the worlds diamonds Even the Soviet Union which was the secondlargest diamond exporter after South Africa sold all its diamonds through the De Beers cartel for the last quartercentury The breakup of the Soviet Union threatened the stability of De Beers though Russia apparently has returned to the cartel Traditionally as new mines were developed De Beers gave them sufficient market share so that they agreed to sell through De Beers and accept its production control system When Tanzania decided to act independently De Beers depressed the price for the quality of stones sold by Tanzania forcing it to rejoin the syndicate Sources Fisher Cootner and Bailey 1972 Eckbo 1976 Pindyck 1977 1979 Fisher 1981 Alan J Wax Spicy New Cartel Sets Nutmeg Prices San Francisco Chronicle May 25 198720 Clyde H Farnsworth OPEC Isnt the Only Cartel That Couldnt New York Times April 24 1988 3 Diamonds Friends Again The Economist March 2 1996 3385960 3641AWLCARLCh05pp146180qxp 12715 1250 PM Page 157 158 Chapter 5 Cartels 12If the number of firms involved in each case is arranged in ascending order then the middle num ber is the median number of firms If there were 5 cases with 2 4 5 8 and 9 firms involved the me dian number of firms would be 5 Imagine a graph that plots the cases so that the horizontal axis shows the number of firms involved and the vertical axis shows the number of cases involving a given number of conspiring firms The mode is the highest point on the plot The mode is the most common number of conspirators 13The median number of firms involved varied by industry In natural resources markets the median number of firms was 13 The corresponding numbers were 7 in manufacturing 11 in distribution 15 in construction 4 in financial institutions 4 in transportation and 8 in services 14Hay and Kelley 1974 studied horizontal pricefixing conspiracies that were prosecuted by the US Department of Justice Antitrust Division Their study excluded price fixing by various profes sional groups because they were not covert but included virtually all other cases that were filed and won in trial or settled by nolo contendere no contest pleas Pleading nolo contendere is equivalent to pleading guilty for the purposes of sentencing but is not an admission of guilt by the defendant When such a plea is accepted by the court a trial is not necessary Occasionally courts accept such pleas over the objection of the Department of Justice of industrial trade associations carried out by the Political and Economic Planning agency in 195356 found that 243 of the 1300 associations 19 percent attempted to fix prices Phillips 1972 Low Organizational Costs Even if a potential cartel could raise prices in the long run and not be discovered it will not form if the cost of initial organization is too high The more complex the negotiations the greater the cost of creating a cartel Four factors keep the cost low facilitating the creation of a cartel Few firms are involved the market is highly concentrated all firms produce a nearly identical product and a trade association exists Setting up a secret meeting without the governments knowledge is relatively easy when there are few firms involved Even if there are many firms in a market the largest firms may meet and establish a cartel dominant firm that does not explicitly include the smaller fringe firms Of the 606 Department of Justice pricefixing cases 191072 examined by Fraas and Greer 1977 the average number of firms involved in each case was 167 whereas the median was 8 and the mode was 412 That is a few cases involving a large number of firms raised the average but the most common type of case involved 4 firms and half the cases involved 8 or fewer firms13 Of the Department of Justice pricefixing cases January 1963December 1972 studied by Hay and Kelley 1974 only 65 involved 50 or more conspirators14 The average number of firms in the remaining cases was 725 Although only 26 of the cases involved 4 or fewer firms nearly half 48 involved 6 or fewer firms and 79 involved 10 or fewer firms Of the global cartels from 1990 to 2003 studied by Connor 2003 the median number of corporate participants was five More than half 77 of these cartels had six or fewer firms Only 13 had 10 or more participants but most of those were or ganized by quasiofficial European trade associations Even where cartels are legal as are many international cartels not involving US firms the number of firms is crucial For example a long period 192872 of success ful cartelization by two countries of the world mercury market was followed by years 3641AWLCARLCh05pp146180qxp 12715 1250 PM Page 158 Creating and Enforcing the Cartel 159 15Scott 1991a shows that multimarket contact can be important Large conglomerates may be po tential rivals in a number of markets simultaneously They can communicate about all these markets at once It is also potentially more costly to deviate from a cartel agreement in one because it risks destroying all the cartel agreements To the degree that multimarket contacts are important the de gree of concentration in a single market may be misleading 16To minimize the systematic bias from excluding cases for which the concentration measure could not be determined directly if the number of firms was known Hay and Kelley 1974 calculated min imum concentration ratios by assuming each firm had an equal share of unsuccessful attempts at price fixing by a larger group of countries MacKieMason and Pindyck 1986 If a few large firms make most of the sales in a market and if they coordinate their activities they can raise price without involving all the other smaller firms in the market For example Spain and Italy which controlled 80 of the worlds production of mercury formed a successful cartel that did not formally involve five other produc ers MacKieMason and Pindyck 1986 Empirical evidence supports the view that cartels are more likely in concentrated in dustries15 In 42 of the Department of Justice pricefixing cases studied by Hay and Kelley 1974 the fourfirm concentration ratio the sum of the market shares of the four biggest firms was over 75 in another 34 of the cases the ratios were between 51 and 75 Thus in 76 of the cases the concentration ratio was greater than 50 Only 6 of the cases had concentration ratios less than 25 The overall average was 67716 Of the global cartels studied by Connor 2003 cartel members usually con trolled over 90 of the markets sales Moreover when entry caused the cartels share to drop below 65 cartel activity typically ceased Similarly the existing evidence shows that cartels are often found in smaller geo graphic areas In the US Justice Department pricefixing and other antitrust cases from the passage of the Sherman Act 1890 through 1969 studied by Posner 1970 nearly half 474 the conspiracies were in local or regional markets 376 were na tionwide and 87 involved foreign trade The smaller the geographical area of a market the more likely it is that a few firms have a large share of the business Firms have more difficulty agreeing on relative prices when each firms product has different qualities or properties Each time a product is modified a new relative price must be established It is easier for a cartel to spot cheating when all it has to examine is a single price It is relatively difficult to detect price cutting that is achieved by an increase in quality a firm could increase its quality and hold its price constant if it wanted to increase sales without explicitly violating the pricing agreement In virtually all the pricefixing cases studied by Hay and Kelley 1974 the product was relatively homogeneous across firms In the few exceptions complicated products or services were allocated on a jobbyjob basis that facilitated coordination or a single issue was isolated for the agreement For example a group of swimsuit manufacturers agreed to delay endofseason discounts Similarly virtually all the recent global con spiracies Connor 2003 involved homogeneous products 3641AWLCARLCh05pp146180qxp 12715 1250 PM Page 159 160 Chapter 5 Cartels Trade associations by lowering the costs of meeting and coordinating activities among firms in a market facilitate the establishment and enforcement of cartels Most industries have trade associations that meet regularly Not all industries with trade as sociations necessarily form cartels However as Adam Smith observed such meetings are conducive to pricefixing agreements and trade associations are often the mecha nism by which large groups coordinate activities In the Hay and Kelley 1974 study of Department of Justice pricefixing cases trade associations were involved in 7 out of 8 cases in which more than 15 firms conspired and in all cases involving more than 25 firms Overall 29 percent of the cases involved trade associations Fraas and Greer 1977 found that 36 percent of all pricefixing cases involved trade associations Moreover the median number of firms involved was 16 when there was a trade associ ation compared to 8 for all cases Posner 1970 found that 436 percent of all an titrust cases involved trade associations Enforcing a Cartel Agreement Even if a market consists of a small number of firms producing a homogeneous good with no close substitutes has an inelastic demand curve and faces no threat of entry a cartel cannot succeed if members can and want to cheat on the agreement Some of the factors that lead to the formation of a cartel also help it to detect cheating and en force its agreement Detecting Cheating Cartel agreements are easier to enforce if detecting violations is easy Four factors aid in the detection of cheating There are few firms in the market Prices do not fluctuate independently Prices are widely known All cartel members sell identical products at the same point in the distribution chain With relatively few firms the cartel may more easily monitor each one and in creases in one firms share of the market an indication of price cutting are easier to detect Further moral or immoral suasion may be easier when there are only a few conspirators See wwwawbccomcarltonperloff Broker Hay and Kelley 1974 found that most of the pricefixing conspiracies lasting 10 or more years were in markets in which there were few firms and the largest firms made most of the sales When a large number of firms was involved conspir acies were generally discovered very quickly especially because details about some of the largegroup organizational meetings often were printed in local newspapers In contrast Posner 1970 found that of the detected cartels large ones lasted as long as smaller ones He found that 52 percent of conspiracies involving 10 or fewer firms had lasted for 6 or more years whereas 64 percent of larger conspiracies per sisted that long Presumably the more firms involved in a conspiracy the more 3641AWLCARLCh05pp146180qxp 12715 1250 PM Page 160 Creating and Enforcing the Cartel 161 17Of the cases studied by Hay and Kelley 1974 detection was due to grand jury investigation of an other case in 24 to complaint by a competitor in 20 to complaint by a customer in 14 to complaint by local state or federal agencies in 12 and to complaint by current or former employ ees in 6 Each of the following methods was responsible for detection in 4 of the cases com plaint by a trade association official investigation of conduct or of performance by the Antitrust Division report of a newspaper and referral to the Antitrust Division by the Federal Trade Commis sion Each of the following methods was responsible for 2 of the cases complaint by an anony mous informant merger investigation and private suit 18Hay and Kelley 1974 found some cases in which sales to government agencies were explicitly ex cluded from the agreement Apparently cartel members believed that price fixing was more likely to be detected and prosecuted if directed against the federal government In other cases some market segments were excluded from agreements in order to reduce potential friction among cartel mem bers Fraas and Greer 1977 found that 19 percent of all cases over a longer period involved bid rig ging Posner 1970 determined that 74 percent of all cases involved sales to the government and 67 percent involved other bidding cases 19Witness Says Mob Is into Highrises San Francisco Chronicle April 30 1988A7 likely is discovery by the government In general conspiracies are uncovered through information provided by private parties rather than by Department of Jus tice investigations17 If a market has frequent shifts in demand input costs or other factors prices in that market have to adjust often In that case cheating on a cartel arrangement may be difficult to detect because it cannot be distinguished easily from other factors that cause price fluctuations Cheating is easier to detect if prices are known Some cartels have arranged for firms to inspect each others books In Posners 1970 study of an titrust cases at least 62 percent of the cases involved exchange of information whereas 43 percent involved policing fines and audits Of course books can be faked so such inspections cannot prevent all violations of the cartel agreement In some cases governments help For example they often report the outcome of bidding on government contracts so that cheating is instantly observable by the cartel see Example 53 A quarter of the cases Hay and Kelley 1974 examined involved some form of bid rigging18 Example 51 describes a phases of the moon scheme used by manufacturers of electri cal products to rotate the winning of sealed bids No firm could hope to win out of turn because its treachery against the cartel would be instantly exposed when the gov ernment announced the winner Vincent The Fish Cafaro a former member of the Genovese organized crime family told senators the mob rigged bids in New York City controlling the concrete industry and construction unions19 He said the contractors and unions that won construction jobs through bid rigging were required to kick back 2 percent to the 2 Percent Club an organization run by the Genovese Gambino Lucchese and Colombo families of New York City He estimated that at least 50 percent of the highrise construction in New York had a mob connection and added Legitimate guys aint got a chance to win contracts for those buildings According to Mr Cafaro the 2 Percent Club split up all of the jobs worth over 2 million Contracts worth over 5 million went to mobrun companies 3641AWLCARLCh05pp146180qxp 12715 1250 PM Page 161 162 Chapter 5 Cartels Concrete Example of GovernmentAided Collusion Members of a cartel have an incentive to undercut an agreed collusive price If cartel members cannot detect such cheating the cartel is likely to fail If a government agency publishes the prices each firm charges it may facilitate the maintenance of a cartel as appears to have been the case recently in Denmark Stigler 1964a explained how the same factors that lead to a successful cartel also can result in successful tacit collusion Many of the conditions necessary for success ful tacit collusion in the readymixed concrete market were already in place prior to the government actions The readymixed concrete market in Denmark consists of a relatively small number of firms The two largest firms have plants throughout the country and compete with a number of smaller firms Because readymixed concrete can be kept in a mixer truck for only about two hours after the mixing it is typically shipped no more than 20 miles from the production site As a consequence relatively few firms compete in a given area Usually fewer than 5 of the 115 plants in Den mark can serve a particular customer Although the national fourfirm concentration ratio is 57 many local markets are monopolistic and the average market share of the largest firm in an area is 70 In 1993 the Danish antitrust authority started gathering and publishing firmspe cific transaction prices for two grades of readymixed concrete The government hoped that by providing buyers with price information seller competition would be stimulated and price would fall According to Albæk Møllgaard and Overgaard 1997 the governments actions led to successful tacit collusive behavior Prices rose and the variance in prices across firms was nearly eliminated The average prices of the reported grades increased by 1520 within less than a year compared to an annual inflation rate of no more than 2 No dramatic changes in costs or other factors explain this large increase indeed wages a major component of costs fell during this period Moreover the variance in firms prices dropped from about 30 around the average to only about 24 Thus it appears that the firms were able to set a higher price as a consequence of the governments information program When smaller contractors complained about the arrangement they were given the right to split all jobs worth over 3 million He said the Genovese family was a very disci plined organization with strict rules and capital punishment for serious violations Public availability of information can greatly simplify cartel enforcement Publicly announcing price increases and decreases well in advance is one method of making price information available to all interested parties An extreme special case of sharing information occurs when a single sales agent or pool is used by all firms for all their sales as was the case in 3 percent of the cases Fraas and Greer 1977 examined and in 6 percent of the cases studied by Posner 1970 Sales agents are commonly used in European cartels Example 53 3641AWLCARLCh05pp146180qxp 12715 1250 PM Page 162 Creating and Enforcing the Cartel 163 20So long as a cartel raises its marginal cost curve by more than its average cost curve such actions increase profits Salop Scheffman and Schwartz 1984 21Hay and Kelley 1974 argued that bid rigging and allocation of jobs among cartel members occur in industries in which orders are relatively large lumpy compared to total sales If some firms are vertically integrated the same firm produces inputs manufac tures the product and sells at the retail level it may be difficult for the cartel to de termine at what point in the distribution chain cheating occurs In contrast if all firms sell to the same type of customer for example at the retail level cheating is easier to detect Cartels with Little Incentive to Cheat A cartel may find enforcement easy under certain circumstances Members have no incentive to cheat on the cartel agreement if their marginal cost curves are relatively inelastic their fixed costs are low relative to total costs their customers place small frequent orders or they have a single sales agent If a firms marginal cost curve is nearly vertical it has little to gain by cheating on the cartel agreement because it costs too much to substantially increase its output In Figure 51a if the marginal cost curve were nearly vertical q would be close to qc Marginal cost curves are likely to be nearly vertical if firms are operating near their full capacities Indeed cartels may force their marginal cost curves to be more vertical by signing union contracts that require double wages for overtime work or using similar techniques Maloney McCormick and Tollison 197920 Suppose a firm incurs a large fixed cost to build a plant in which it can produce at a constant marginal cost at any output level up to the plants capacity Such a firm has substantial unutilized capacity when demand falls such as during a recession It has an incentive to lower its price below the cartel level to stimulate its sales If there are many customers in a market who make small purchases no firm has an incentive to lower prices below the cartel level If it does so without announcing the price cut other customers are unlikely to learn of the price cut hence its sales will not rise If the firm advertises its price reduction the other cartel members will learn of the cut and retaliate In contrast when only a few customers place large infrequent orders a cartel has trouble detecting and preventing cheating21 Firms have an incentive to grant price reductions to large buyers to keep them as customers Legal cartels can try to prevent cheating by requiring that a single agent or organi zation sell output from all firms For example the iodine cartel one of the longest lived international cartels 61 years 18781939 made all its sales through a central office in London Eckbo 1976 See Example 54 Methods of Preventing Cheating Unless a cartel can detect violations of its price fixing agreement and prevent reoccurrences member firms engage in secret price cutting or output expansions that destroys the cartel Although economists and lawyers understand a number of mechanisms that aid cartels in enforcing their agreements the most successful cartel agreements and their enforcement mechanisms may be unknown Here we concen trate on six methods fix more than just price divide the market fix market shares use mostfavorednation clauses use meetingcompetition clauses and establish trigger prices 3641AWLCARLCh05pp146180qxp 12715 1250 PM Page 163 164 Chapter 5 Cartels Relieving the Headache of Running a Cartel The major use for bromine in the late 1800s was as a headache remedy and sedative The bromine industry was not concentrated and consisted of many small producers Ordinarily such an industry would not be a good candidate for a cartel because of the large number of producers In the early 1880s the price of bromine fell by 40 percent In 1885 the National Bromine Company bromine pool was formed It purchased the bromine pro duced by all manufacturers The bromine pool then sold bromine to two indepen dent distributors of chemicals Powers and Weightman of Philadelphia and Malinckrodt Chemical Works of St Louis which were obligated collectively to buy the entire output of the pool In turn these two distributors sold to customers in their territories The bromine pool required that manufacturers sell exclusively to it and that if any manufacturer violated this provision contracts with all other manu facturers could be terminated Moreover if a manufacturer entered the industry and did not contract with the bromine pool contracts with other manufacturers could be terminated The two distributors who were obligated by contract to purchase all the pools output accumulated large inventories of bromine which they threatened to dump on the market if any manufacturer failed to cooperate with the pool Indeed inventories were sold during price wars in 1886 and 1888 in order to punish com petitors and to restore pricing discipline The distributors were much better able than individual manufacturers to monitor sales to customers and thus to detect whether any manufacturers were making secret sales For their role in the cartel the independent distributors took a significant share of the cartel profits In 1892 this share was renegotiated to give a larger portion to manufacturers The original pool was replaced by W R Shields which performed similar functions The very successful bromine cartel lasted from 1885 to 1902 During its reign the average price of bromine was about 25 percent higher than the average in the years before the cartels formation There were only three periods of extended price wars over the cartels roughly 20 year life span The cartel ended because Dow Chemical Company developed a low cost method of processing bromine in the 1890s Dow initially signed contracts with the two bromine distributors that the bromine cartel had used as its exclusive distrib utors But in 1902 Dow had grown so large that it decided not to rely exclusively on distributors to sell its products and began selling directly to customers The pool fell apart and the price of potassium bromide the major bromine product plunged 45 percent in two months Undoubtedly the demand by owners of small bromine processing firms for potassium bromide as a headache remedy increased Source Levenstein 1993 Example 54 3641AWLCARLCh05pp146180qxp 12715 1250 PM Page 164 Creating and Enforcing the Cartel 165 22Posner 2003 51 notes The machinery of cartelization may include sales quotas exclusive sales agencies industrywide pricefixing committees the levying of penalties for infractions provi sions for the arbitration of disputes the establishment of an investigative apparatus product standard ization allocation of customers and the division of geographical markets To prevent cheating successful cartels must do more than just set a price Posner 1970 400 finds that at least 14 of all Department of Justice antitrust cases in volved explicit collusion on terms besides basic price and this figure apparently does not include explicit rules on dividing the market exchanging information or sales quotas22 Some cartels succeed in preventing cheating by assigning each firm certain buyers or geographic areas which allows cheating to be detected easily Fraas and Greer 1977 found that 26 of pricefixing cases involved market allocation schemes Posner 1970 found that 78 of the antitrust cases involved an allocation of cus tomers 146 involved a division of territories and 18 involved a division of product markets or 24 overall The twocountry mercury cartel used a geo graphic division of markets Spain supplied the United States and Italy supplied Europe Another effective technique is for members of a cartel to agree to fix market shares say at their precartel levels See Example 55 As long as market shares are easily observable no firm has an incentive to cut its price If it lowered its price its share would increase and other firms would retaliate For example cartel members who detect changes in the output levels of other firms could adjust their own output to maintain their proportionate shares of market output Osborne 1976 Spence 1978a 1978b All firms expect this reaction so no firm has an incentive to increase its own output only to earn lower profits after retaliation As wwwawbccom carltonperloff Conjectural Variations discusses fixing market shares can result in the cartel price A mostfavorednation clause in a sales contract guarantees the buyer that the seller is not selling at a lower price to another buyer Salop 1986 A variant of such clauses was used in sales of large steamturbine generators The two major sellers General Electric and Westinghouse see Example 51 each used clauses in their con tracts stating that the seller would not offer a lower price to any other buyer current or future without offering the same price decrease to the initial buyer This rebate mechanism created a penalty for cheating on the cartel If either company deviated from the agreement by cutting its price it would have to cut prices to all previous buyers as well A meetingcompetition clause in a longterm supply contract or in an advertise ment guarantees the buyer that if another firm offers a lower price the seller will match it or release the buyer from the contract Salop 1986 Such a clause makes it difficult for a firm to cheat because buyers will bring news of lower prices to the cartel Thus surprisingly these clauses could be associated with high cartel prices rather than the low ones they seem to guarantee 3641AWLCARLCh05pp146180qxp 12715 1250 PM Page 165 166 Chapter 5 Cartels Vitamins Cartel In the 1990s there was a massive worldwide cartel involving many different vita mins including biotin folic acid and vitamins A B1 B2 B5 B6 C and E among others Vitamins have a wide variety of uses as additives to human and animal diets and in skin and healthcare products The various vitamins are not substitutes for each other Vitamin production is highly concentrated among a few firms At the time of the cartel the three largest producers were HoffmanLaRoche which since has sold its vi tamins business which produced 40 to 50 percent of all vitamins BASF with a 20 to 30 percent share and Aventis formerly RhonePoulenc with a 5 to 15 percent share These major manufacturers produced many of the same vitamins Over half of all vita min sales were for vitamins A and E which were sold by all three major producers Allegedly beginning in 1989 HoffmanLaRoche BASF and RhonePoulenc held meetings to discuss allocations of market sales around the world so as to reduce com petition and soon thereafter other firms became involved in a worldwide cartel The cartel fixed market shares for each vitamin by country agreed on price in creases specified target prices and minimum prices and shared information to en sure that each firm was abiding by its allocation Sometimes the firms explicitly discussed large individual customers and agreed on those customers prices and how much of the customers needs each manufacturer would supply The firms met regularly There were four levels of meetings the highest level in volving senior executives who determined overall strategy and adherence to the agree ments the next level involving marketing executives about two or three times a year another meeting usually quarterly involving marketing managers of individual products to monitor the implementation of the allocations and finally quarterly meetings of regional marketing managers to discuss pricing implement price in creases and adjust allocations The budget meetings in August were used to outline allocations for the coming year together with price increases Example 55 All cartel members could agree that if the market price drops below a certain level called a trigger price each firm will expand its output to the precartel level Fried man 1971 that is all firms will abandon the cartel agreement In this case a firm that cuts its price might gain in the extremely short run but would lose in the end due to the destruction of the cartel by this predetermined punishment mechanism One reason to use trigger prices is that in some markets firms have difficulty distin guishing between cheating by other firms and random fluctuations in price due to fluc tuations in demand or supply costs It is possible however for cartels to modify their punishment methods to prevent cheating even when random shocks occur Green and Porter 1984 If firms were to permanently revert to competitive behavior whenever they detected a fall in price the cartel could be destroyed by a random fluctuation in 3641AWLCARLCh05pp146180qxp 12715 1250 PM Page 166 Creating and Enforcing the Cartel 167 The price was usually raised in increments of 5 percent and the effective date of a price increase was often April 1 of the next year HoffmanLaRoche typically initi ated the price increases with the other firms following suit The detailed exchanges of sales information allowed the firms to monitor adherence to their sales allocation If a firm had sold too much in a given year it could be required to buy supplies from the other firms to restore the original sales allocation Although the exact amounts by which the cartel raised prices are subject to dis pute the increases during the cartel period were sizable For example the average price for vitamin A rose by 40 percent and that of vitamin E increased over 60 per cent from 1990 to 1998 The price of vitamin C rose by about 30 percent during the identified cartel period and fell by 50 percent thereafter Starting in the late 1990s RhonePoulenc participated in the US Justice Depart ments corporate leniency program The first cartel member to confess to the DOJif it is not a ringleader or enforcer in the conspiracy and if the DOJ is not aware of the il legal activityis granted automatic amnesty RhonePoulenc revealed the existence of the cartel and details about its operations so as to avoid antitrust fines Subsequently other cartel members agreed to pay multimilliondollar fines These same firms had to pay fines to Canadian and European Union competition authorities as well The fines were substantially larger than had been collected in previous antitrust cases The largest US fines were 500 million for HoffmanLaRoche 225 million for BASF and 72 million for the Japanese company Takeda Two Hoffman LaRoche executives went to jail The biggest Canadian fines were the 48 million Canadian collected from HoffmanLaRoche the 18 million from BASF and the 14 million from RhonePoulenc The major European fines were 462 million for HoffmanLaRoche about 300 million for BASF and 37 million for Takeda Pri vate antitrust settlements collected additional amounts from the cartel members Source Official Journal of the European Communities Commission Decision of 21 November 2001 Case CompE137512Vitamins 23If the firms revert to their precartel output level the price falls to the precartel level as well A more severe punishment a price below the precartel level may be used instead With a lower price it may be possible to shorten the punishment period For an illustration of how cartel prices might be set to minimize cheating see Davidson and Martin 1985 Where members of a cartel disagree on how to behave some kind of voting mechanism may be used Cave and Salant 1987 price rather than price cutting by one firm Instead if the firms agreed to behave competitively only for a predetermined length of time and then to revert to the cartel behavior a random fluctuation in price would not destroy the cartel permanently23 One attraction of this scheme is that even if the agreement temporarily breaks down it can be reestablished without further meetings In a market in which random 3641AWLCARLCh05pp146180qxp 12715 1250 PM Page 167 168 Chapter 5 Cartels 24Bernheim and Ray 1989 Evans and Maskin 1989 Farrell and Maskin 1989 and others point out that instead of going into a punishment phase cartel members may renegotiate their cartel agree ment These papers indicate however that it may be possible to form agreements that avoid this renegotiation problem price fluctuations can mask cheating on the cartel agreement by firms such an agree ment could lead to recurrent sharp declines in price and cartel profit levels When a random drop in price occurs cartel members punish themselves unnecessarily Nonetheless this mechanism may be attractive to the cartel because if the punish ment period when all firms produce large levels of output is long enough it is never in a firms best longrun interest to cheat on the cartel Thus cartel members realize that the price only falls below the trigger price because of random fluctuations be cause no firm ever engages in price cutting The cartel must keep punishing itself however if it stopped price cutting would occur24 Example 56 provides an example of the American railfreight industry in the 1880s that may illustrate such behavior Cartels and Price Wars Many observers seeing large price fluctuations in a market argue that the firms in that market are trying to form a cartel that keeps breaking apart They conclude that government intervention is not required because competitive forces keep destroying the cartel Yet these fluctuations could be part of a rational longrun cartel policy in volving trigger prices as discussed in the preceding section This triggerprice argu ment holds that price wars occur more often during unexpected business cycle downturns recessions and depressions when price is likely to decline in response to lowered demand Green and Porter 1984 Staiger and Wolak 1992 We expect then that cartels are more likely to terminate during a price war Other economists argue that price wars should occur in periods of high demand Rotemberg and Saloner 1986 They reason that the benefit from undercutting the cartel price is greatest dur ing booms To see whether either or both theories are realistic Valerie Y Suslow 1998 investi gates the stability of cartels over the business cycle by examining 72 international car tel agreements covering 47 industries during the period 192039 Because major European countries had no systematic antitrust legislation prior to World War II these cartels were legal and had formal written contracts As of 1927 cartels were legal in Switzerland whereas Belgium France Spain Italy and the Netherlands did not explicitly prohibit them Under German law cartels were legal however Germany passed antitrust legislation in 1923 that was designed to guard against abuses of economic power In 1930 Great Britain adopted a resolution rec ognizing cartels as a fact of economic life but calling for the principle of publicity which required compulsory notification registration and publication of the cartel agreements Other European countries followed Great Britains policy in the mid 1930s It was not until after World War II that France passed legislation to control cartel activity 3641AWLCARLCh05pp146180qxp 12715 1250 PM Page 168 Creating and Enforcing the Cartel 169 How Consumers Were Railroaded During the 1880s a cartel of US railroads openly operated as the Joint Executive Committee JEC Prior to the Sherman Act of 1890 no law prohibited such a car tel As Porter explains the JEC appears to have used a triggerprice strategy Green and Porter 1984 The JEC agreement allocated market shares rather than the absolute quantities shipped Each railroad set its rates individually and the JEC office reported weekly accounts so that each railroad could see the total amount transported Because total demand was quite variable each firms market share depended on both the prices charged by all firms and unpredictable market fluctuations Entry occurred twice between 1880 and 1886 the period Porter 1983a studies In each case the cartel passively accepted the entrants allocated them market shares and thereby allowed the cartel agreement to persist On a number of occasions however when the cartel thought that cheating had occurred it cut prices for a time and then returned to the cartel price Porter finds that noncooperative periods averaged about 10 weeks in duration and occurred in 1881 1884 and 1885 The 1881 and 1884 incidents each occurred about 40 weeks after a new firm entered He notes however that these price wars were not triggered by an unexpected tapering off of demand Porter also finds that price was 66 percent higher and quantity was 33 percent lower in cooperative periods As a result the cartel as a whole earned about 11 per cent more revenues in cooperative periods Sources MacAvoy 1965 Ulen 1980 and Porter 1983a See also Ellison 1994 Example 56 It should have been easier for these cartels to survive than for illegal ones in the United States German French or British firms were participants in roughly half the cartels and US firms were involved in onethird of them In the 1940s US firms were indicted for their participation in 10 of these international cartels According to Suslow the median cartel lasted slightly more than 5 years 75 percent lasted more than 2 years and 20 percent lasted more than 10 years There was an in dustry pattern Of the singleepisode cartels 40 percent involved chemicals with only 6 percent in metals In contrast 46 percent of the multipleepisode cartels involved metals with only 17 percent in chemicals In the 42 cartel episodes in which the number of firms is known 83 percent had 10 or fewer firms 64 percent had 5 or fewer and 39 percent had 3 or fewer Of the 74 percent of the 39 cartels for which there is marketshare information each had a world market share of over 50 percent Thus as with US cartels these international cartels involved relatively few firms with large collective market shares 3641AWLCARLCh05pp146180qxp 12715 1250 PM Page 169 170 Chapter 5 Cartels TABLE 51 Market Conditions Facilitating Global Price Fixing in Lysine Citric Acid and Vitamins A and E in the Early 1990s Market Conditions Lysine Citric Acid Synthetic Vitamins A E High seller concentration CR4 Global market 95 80 95 US market 97 90 100 Few cartel participants 4or5 4or5 3 High cartel supply control 9599 6570 95100 Low buyer concentration CR4 30 40 20 Homogeneous product Perfect High High High barriers to market entry Large plant scales 150 million 150 million Probably Sunk investment costs Yes Yes Yes Technology secret Yes Yes Yes Slow building of new plants 3 years 3 years 3 years Buyers observation of market prices None Some Little Annual market growth 10 steady 8 steady 23 steady CR4 is the share of sales of the four largest firms in the industry Source Connor 2003 Suslow estimates the probability that a cartel will fall apart at a specific time given that it survives until that time Controlling for other factors she found that cartels are more likely to fail during businesscycle downturns recessions and depres sions2 Moreover cartels that were alive during periods of growth were less likely to end than others In general greater volatility in aggregate economic activity over the lifetime of the cartel frequent upswings and downturns increases the probability of cartel breakdowns We have discussed several factors that help a cartel to form and to prevent cheating Many large successful cartels possess these properties Table 51 from Connor 2003 shows the presence of these market conditions in the lysine citric acid and vitamins A and E industries in the early 1990s when each had an operating cartel Entry plays a particularly important role see de Roos 1999 on lysine 5Hajivassiliou 1989 reaches similar findings based on his study of the 188086 US railroad car tel Porter 1983a and Lee and Porter 1984 examine this cartels behavior under the maintained as sumption of the Green and Porter model Town 1991 rejects both the Green and Porter and the Rotemberg and Saloner hypotheses for this cartel and concludes that price wars were not related to demand fluctuations Consumers Gain as Cartels Fail 171 Consumers Gain as Cartels Fail Firms that follow a cartels rules look with disfavor on firms that produce more than the cartel says they should Violators of the cartels rules may be called cheaters or worse by the firms that obey them Consumers however benefit from such noncartel behavior The violators of the cartel agreement produce more than the cartel wants which lowers the market price A numerical example illustrates the effects of noncompliance by some firms see Appendix 5A for details The market in this example includes 50 identical firms We assume that no more firms can enter this market Of the 50 firms 7 firms do not follow the cartels agreement to restrict output they sell as much as they want These firms are price takers The cartel is a dominant firm facing a competitive fringe as we studied in Chapter 4 The residual demand facing the cartel is obtained by subtracting the fringe supply from the market demand Figure 52b shows the residual demand curve thick dark blue line that lies below the market demand curve thin blue line at prices above the competitive firms shutdown level p 1026 The residual demand curve has a kink FIGURE 52 Imperfect Cartel a Noncartel firms j 20 b Cartel firms 50 7 30 50 50 Market demand Noncartel supply Cartel re marginal 30 30 cost 24 se ee ee ee eee ee DA RQ Market demand Residual demand 10 10 Residual marginal revenue 280 1000 240 800 1000 Quantity QO Quantity QO 6Because each competitive firms supply curve is q 10 p it would lose money if it produced positive levels of output at prices below 10 172 Chapter 5 Cartels Market Variables Under Various Degrees of Cartelization 50 Firms Industry Consumer Number of Market Market Profits Surplus Welfare Noncartel Firms Price p Elasticity Output 7 CS CS 7 Monopoly 0 3333 200 333 6667 2778 9445 1 3241 184 352 6524 3094 9618 10 2697 117 461 5318 5304 10622 20 2400 092 520 4360 6760 11120 30 2244 081 551 3743 7591 11337 40 2167 076 567 3391 8027 11418 49 21431 075 571 3267 8162 11428 Competition 50 21429 075 571 3265 8163 11429 CS Consumer Surplus is the triangle with area 1000 20p40 Cartels Market Share as a percentage 100 times the cartels sales divided by total sales DWL Deadweight Loss competitive welfare actual welfare Price Markup as a percentage 100p MCp in it at p 10 Because cartel firms have the same cost functions as noncartel firms the cartel cannot afford to produce at prices below p 10 either so the lower portion of the residual demand curve is not of interest The profitmaximizing cartel chooses its output 240 by setting its marginal revenue the curve marginal to its residual de mand curve equal to its marginal cost as shown in Figure 52b This output deter mines the cartel price 24 At that price the noncartel firms output is 280 as Figure 52a shows Table 52 shows what happens as the number of firms belonging to the cartel changes The market is in competitive equilibrium when all 50 firms act indepen dently and do not belong to the cartel 7 50 The competitive market price is 2143 and consumer surplus and total welfare are maximized At the other extreme if all the firms join the cartel 7 0 the cartel is a monop oly The monopoly price is 3333 or 56 percent higher than the competitive price Only 333 units of output are produced by the market or 58 percent as much as the competitive quantity 571 However each firms profits of 13333 is more than dou ble the competitive level 6531 Consumer surplus is only about onethird as great and total welfare is only 83 percent as great as under competition that is consumer losses are greater than the cartel gains This loss to society the deadweight loss to mo nopoly see Chapter 4 is 18 percent of sales and 71 percent of consumer surplus at the monopoly price Consumers Gain as Cartels Fail 173 As the cartel gains members the incentive of a cartel firm to cheat also grows be cause the discrepancy between a nonmembers profit and a cartel firms profit increases At every price nonmembers earn more than cartel members because nonmembers produce more yet sell at the same price as cartel members Consumers benefit if firms refuse to join the cartel If only one firm refuses to abide by the cartel rules and is a price taker market price is about 3 percent lower than the monop oly price deadweight loss is 10 percent lower and consumer surplus is 10 percent higher Table 52 also shows that it hardly pays for one firm to try to act as a price setter by itself If one firm forms a cartel consisting only of itself so that there are 49 noncar tel firms it faces a residual demand curve with only a slight slope It can reduce its output from the competitive level of 11429 to 1127 units thereby maximizing its profits which rise by 1 from 6531 to 6532 The 49 noncartel firms seeing the higher price and acting as price takers increase their output to 11431 units causing profits to rise by a phenomenal 2 to 6533each noncartel firms profits rise by more than those of the singlefirm cartel All firms profits rise in this case because the expansion of output by the 49 noncartel firms does not completely offset the reduc tion in output by the single firm Total market output falls from 57143 to 57138 units causing the price to rise from 21429 to 21431 The welfare losses from such a limited cartel are small The larger the market share of the cartel the greater the effi ciency cost See Example 57 Cartel Firm Noncartel Firm DWL as Cartels Market Price Markup of Sales Share Output Profits Output Profits 179 100 50 666 13333 159 94 48 672 12802 2241 25110 65 63 36 727 9695 1697 14399 25 46 25 800 8000 1400 9800 07 32 16 889 7108 1244 7738 01 18 8 1000 6670 1167 6809 00 2 1 1127 6532 11431 6533 0 11429 6531 3641AWLCARLCh05pp146180qxp 12715 1250 PM Page 173 174 Chapter 5 Cartels PriceFixing Laws Nothing is illegal if a hundred businessmen decide to do it Andrew Young In the late nineteenth century cartels were legal and operated in several US indus tries including oil powder railroads sugar and tobacco The Sherman Antitrust Act of 1890 was passed in response to this activity to protect trade and commerce against unlawful restraints and monopolies see Chapter 19 In 1914 the Federal Trade Commission Act established the Federal Trade Commission FTC and its Section 5 holds that unfair methods of competition are hereby declared illegal This act is still used by the FTC in prosecuting antitrust violations as is the Sherman Act by the US Department of Justice The Social Costs of Cartelization Posner 2003 estimates the social cost of several mainly international wellorga nized overt cartels using his theory discussed in Chapter 4 that all cartel profits are dissipated in rentseeking activity His results are shown in the table Social Costs as Industry Cartel Price Increase Elasticity of Industrys Sales Nitrogen 75 233 64 Sugar 30 433 35 Aluminum 100 200 75 Aluminum 59 363 56 Rubber 100 200 75 Electric bulbs 37 370 44 Copper 31 422 36 Castiron pipe 39 356 42 Note These figures are based on the Appendix in Posner 2003 Apparently two figures are given for aluminum because he has two estimates of the cartel price increase The elasticities are based on the cartels priceincrease data on the assumptions that the industry is charging the profitmaximizing monopoly price and that the de mand curve is linear Because of these restrictive assumptions Posner warns that these results should be viewed with some caution Nonetheless if these figures are anywhere near accurate the social costs of these cartels are large Source R A Posner Antitrust Law 2003 by The University of Chicago All Rights Reserved Example 57 3641AWLCARLCh05pp146180qxp 12715 1250 PM Page 174 PriceFixing Laws 175 27Addyston Pipe and Steel Co v United States 175 US 211 1899 This citation is from the US Reporter volume 175 and starts on page 235 The case was decided by the US Supreme Court in 1899 28United States v Trenton Potteries Co 273 US 392 1927 established a per se rule A later case Appalachian Coals Inc v United States 288 US 344 1933 however appeared to deviate from this per se rule United States v SoconyVacuum Oil Co 310 US 150 1940 firmly established the per se rule Fines and Sentences in US Department of Justice PriceFixing Cases 19701999 Number of Criminal Cases in Which Cases in Which Prison Years Cases Filed Fines Imposed Sentences Imposed 19701979 176 156 25 19801989 623 513 196 19901999 416 324 61 Source Connor 2003 TABLE 53 The Sherman Act makes illegal conspiracies whose sole purpose is to raise price For example in the Addyston Pipe and Steel case in 1899 outright bid rigging and the di viding of the market into regional monopolies were found illegal27 The Trenton Potteries case of 1927 and SoconyVacuum Oil Co case of 1940 deter mined that price fixing was a per se violation the action by itself is illegal regardless of whether the price set was above the competitive price28 These cases established that the violation is the attempt to charge the monopoly price the government does not have to show that the defendants succeeded in their attempt Posner 2003 Cartels formed solely to raise price are strictly prohibited Table 53 based on Connor 2003 shows the number of pricefixing cases the DOJ filed over the period 197099 how many fines were collected and the number in which individuals received prison sentences Example 58 shows that US Canadian EC and other antitrust authorities have heavily fined corporations engaged in global conspiracies since the early 1990s This approach to preventing price fixing is based on evidence of conspiracy rather than the economic effects of the conspiracy The government seeks evidence of con spiracies such as secret meetings in smokefilled rooms rather than economic evi dence such as price increases Cases involving only tacit collusion that is without explicit communications between the parties are not actionable under antitrust laws The current laws have been successful in eliminating overt but not tacit collusion Posner 2003 52 observes that the elimination of the cartel is an impressive and remains the major achievement of American antitrust Increasingly many other countries especially those in Europe actively try to prevent cartels 3641AWLCARLCh05pp146180qxp 12715 1250 PM Page 175 176 Chapter 5 Cartels Prosecuting Global Cartels Between World War II and the 1990s few private nongovernmentrun global car tels were observed although there were a number of governmentorganized cartels as discussed in Example 52 The period since 1990 has seen an explosion of private in ternational cartels that have been identified and prosecuted by antitrust agencies in North America and Europe From 1993 through July 2003 antitrust authorities dis covered engaged in public investigation handed down indictments or imposed fines 167 private cartels with corporate or individual participants from at least two countries By July 2003 128 were at least partially prosecuted and public investiga tions were initiated for another 39 Moreover approximately 35 secret US grand jury investigations of international cartels were ongoing Since the early 1990s US Canadian EC and other antitrust authorities have heavily fined corporations engaged in global conspiracies The prosecutions of 20 global multicontinent cartels after 1995 resulted in heavy fines in all cases and prison sentences in half From 1996 through mid2003 US Canadian EC and other antitrust authorities have imposed fines of over 53 billion on corporations engaged in international conspiracies Of the cartels 31 operated in two or more continents with most of them in North America Europe and Asia covering 51 of affected sales 20 had sales within more than one EU country 25 of sales 26 operated in a single Euro pean country 13 of sales 19 were only in the United States or Canada 11 of sales and only six international cartels were discovered elsewhere Compared to previous international cartels their markets encompassed more of the world and hence the damages they inflicted were accordingly greater Firms en gaged in international cartels that were successfully prosecuted by the US Depart ment of Justice DOJ had sales that exceeded 55 billion The DOJ collected over 900 million from international price fixers in 1999 alonefar more than it col lected in the previous 108 years of US antitrust enforcement For the 60 cartels for which he had adequate information Connor 2003 re ported that the cartels raised prices by an average of 28 25 in organic chemicals and 35 in other industries International cartel sales were concentrated in a few in dustries 39 involved intermediate organic chemicals 52 were other manufactur ers mostly metal cement plastics graphite products and the remaining 9 were in Example 58 SUMMARY Firms have incentives both to form cartels and to cheat on the cartel arrangement Firms want to join cartels if the cartel is capable of raising prices for sustained periods of time Prices are more likely to be significantly elevated above the competitive level when the cartel controls a substantial share of the markets output when it faces a relatively inelas 3641AWLCARLCh05pp146180qxp 12715 1250 PM Page 176 Summary 177 construction transport finance and other services However over 80 of sales in cartels discovered before 2000 involved food and agriculture The worlds major antitrust agencies have responded to these threats to worldwide economic wellbeing by imposing unprecedented sanctions 19902003 23 bil lion by the United States 36 billion by the European Union Of those corpora tions accused by the DOJ of criminal price fixing fewer than 1 were foreignbased firms prior to 1995 whereas more than 50 were nonUS corporations after 1997 The DOJ has convicted cartel executives from 12 foreign countries sending many to prison Similarly between 2000 and 2002 the EC fined 42 companies that were guilty of global price fixing of which 55 were nonEU firms Executives were fined in 50 of cases and received prison sentences in 33 of US cases Sixtytwo executives were fined and 43 fugitives were indicted The indi vidual fines added up to 244 million and four fines were greater than 350000 but the median fine was only 50000 Of the 62 executives fined 30 received prison sentences averaging 111 months Contrary to DOJ claims for all criminal pricefix ing cases Connor failed to find an upward trend in the length of prison sentences for participants in international cartels Canada prosecuted 18 cases between 1991 and 2003 Most of these cases followed US actions The initial Canadian prosecution started an average of eight months af ter a US conviction Canada fined 68 corporations 133 million US almost all of which were nonCanadian These fines were about 6 of the corresponding US fines Canada also fined four individuals a total of 600000 From 1990 to 2003 the EU prosecuted 35 cases of which 16 were global cartels The EU fined or granted amnesty to 259 corporations of which 30 were located outside the EU On average EU decisions lagged US prosecutions by 34 months The EU fines were about 72 of comparable US fines on the same cartel cases Although the antitrust agencies have acted more aggressively than in the past and are imposing larger fines cartels continue to thrive For the period 2000 through 2003 23 international cartels were discovered on average per year six times faster than a decade earlier Apparently firms believe that the fines are merely a cost of doing business as more than 50 corporations were members of multiple cartels up to 13 Source Connor 2003 tic demand curve and when entry is limited The expected rewards of forming illegal cartels are greater when detection by the government is unlikely and the fines are low Cartels fail due to cheating by member firms or by competition from firms outside the cartel Individual firms have an incentive to cheat on a cartel agreement because they can make higher profits by increasing output or undercutting the cartels price A cartel can maintain its agreement only if cheating can be detected and adequately 3641AWLCARLCh05pp146180qxp 12715 1250 PM Page 177 178 Chapter 5 Cartels punished Cartels have developed a number of techniques including division of the market and complex contract clauses in order to enforce their agreements When cartels succeed in raising prices there is a loss of consumer surplus The gain to the cartel is less than the loss to consumers The difference is a deadweight effi ciency loss The fewer the firms that go along with the cartel agreement the less mar ket power the cartel has and hence the less it harms consumers and society The US government and many others have antitrust laws that penalize firms that form cartels At least in the United States pricefixing cartels have been vigorously prosecuted PROBLEMS 1 Historically at each Organization of Petroleum Exporting Countries OPEC meeting Saudi Arabia the largest oil producer argued that the cartel should cut production The Saudis com plained that most OPEC member countries in cluding Saudi Arabia produced more oil than their cartel agreement allotted them Illustrate using a graph and explain why cartel members would produce more than the allotted amount given that they know that overproduction will drive down the price of their product 2 What are the main factors that increase the likeli hood of a cartel being successful 3 Use a graph to show why an increase in the market demand elasticity reduces a cartels monopoly power Show how an increase in the market de mand elasticity affects the elasticity of the residual demand curve 4 Problem based on Appendix 5A Show that the sum of a cartels output plus the output of noncar tel firms is less than the competitive output and that the corresponding price is higher than the competitive price 5 Problem based on Appendix 5A Show that a car tels price falls as the number of noncartel firms j increases Answers to oddnumbered problems are given at the back of the book SUGGESTED READINGS A good survey of modern thought on cartels is Jacquemin and Slade 1989 If you want to see how an actual cartel operates go to wwwawbccom carltonperloff Chapter 5 A Cartel at Work There you will find a link wwwusdojgovatr publicspeeches4489htm to a speech given April 6 2000 by James M Griffin Deputy Assistant At torney General Antitrust Division US Depart ment of Justice entitled An Inside Look at a Cartel at Work Common Characteristics of International Cartels in which Griffin outlines how the interna tional lysine cartel worked The Department of Jus tice can supply you with an actual tape of the lysine cartel meetings Contact the United States Depart ment of Justice Antitrust Division Freedom of In formation Act Unit 325 Seventh Street NW Suite 200 Washington DC 20530 Connor 2001 ex tensively analyzes global cartels since 1990 3641AWLCARLCh05pp146180qxp 12715 1250 PM Page 178 Appendix 5A The Effects of Cartel Size 179 APPENDIX 5A The Effects of Cartel Size This appendix derives the equations used in the example reported in Table 52 which shows how price and output vary with the number of cartel members The total num ber of firms is assumed to be fixed at no further entry is possible The market demand curve is linear Q a bp 5A1 where a and 6 are positive constants Q is market output and p is the price The elas ticity of demand is d b RP 4 TP 5A2 dp Q Q 4 Each firm has a linear marginal cost ZC of MC d eq 5A3 where q is the output of one of the x firms and d and e are positive constants As a re sult the competitive supply the output produced at the point where marginal cost equals price is d Q ng mp a 5A4 Competitive equilibrium is determined by setting the righthand sides of the quan titydemanded equation 5A1 and the quantitysupplied equation 5A4 equal and solving for p the equilibrium price The equilibrium quantity Q can be found by substituting p into Equation 5A1 or 5A4 The equilibrium values are aet nd 5A5 Po be an a bd n A6 Qe of be 6A6 180 Chapter 5 Cartels Now suppose that 7 firms in the market form a cartel and the remaining firms j n do not As shown in Figure 52b the residual demand Q is the market de mand minus the noncartel supply Q jg p 4 Q Q jg a bp PP 5A7 The cartel acts as a monopoly with respect to its residual demand and sets its mar ginal revenue MR equal to its marginal cost The cartels revenues R may be found by solving Equation 5A7 for p as a function of Q and multiplying that by Q to obtain ae jd eQ R pQ Ja A8 m PQ be j Q 6 By differentiating R with respect to Q we obtain the cartels marginal revenue ae jd 2e MR 5A9 nm bet ij be j Q GA The cartels marginal cost is MC d Q 5A10 n The quantity the cartel chooses to produce Q Q is determined by equating the cartels marginal revenue Equation 5A9 and marginal cost Equation 5A10 n jla bd 5A11 Qn be 2nj By differentiating Q with respect to j it can be shown that the cartels output falls as the number of nonmember firms rises