A Cartel of Widget Producers

In normal business, independent traders compete with one another, whether they are buyers or sellers. But it is possible to think of situations in which all buyers, or all sellers, gain by reducing competition in a controlled fashion. In such situations, traders may form cartels.

A cartel can be defined as a group of independent traders in a commodity that agrees to limited cooperation rather than complete competition, in order to influence (or even control) the price of that commodity. A cartel can only be put together when all the members perceive that they will benefit more by joining the cartel than they would as individual traders.

But there is nothing altruistic about a cartel. A cartel can only persist as long as the members actually receive the expected relative benefit, and can continue to expect further relative benefit. A member will leave a cartel as soon as it perceives a relative benefit in doing so. Even though they all adhere to the trade limitations agreed on by the entire cartel, the individual members remain competitors, and are free to compete in other ways.

A successful cartel gathers benefits at the expense of outsiders, and those benefits are shared by the members. A cartel will always have internal stresses, as each member tries to negotiate the internal agreement to give it a larger benefit than those of its fellow members. These stresses will be present even in a cartel whose members are honorable enough to keep to the agreement. But since the benefits to the cartel are derived by artificial manipulation of the open market, even greater benefits will accrue to a cartel member who cheats on the agreement. Therefore, a successful cartel must be able to punish a cheating member with a penalty that is much heavier than any potential gain from cheating. Each member must also be assured that its fellow members are not cheating. Obviously, a cartel where cheaters are not punished, or cannot be punished, is not likely to last long. Therefore, a successful cartel must have its own effective internal monitoring and policing system.

A successful cartel can expect stresses from outside, because, by definition, cartel members gather benefits that outsiders would like to appropriate for themselves if they can, or share if they must. Depending on the relative pressures that can be applied by the cartel and the outside systems, a successful cartel is likely to require considerable resources and power before it begins to operate.

Cartels are often rather short-lived. Changes in market conditions can remove the reasons for cartel formation. Sometimes even a short-lived cartel can make a dramatic difference to the traders that formed it. Therefore, the "breakdown" of a cartel does not necessarily mean that it has been "broken" from outside: it may simply have become a redundant organization that is abandoned freely by its constituent members.

Cartels can operate in different ways. I have chosen to use a "widget" cartel to make the argument, but the principles apply whether "widgets" are railroad cars of copper ore, gold bars, plastic toys, or any other item that has any value.

Suppose that a group of widget producers perceives that the market price they receive for widgets is "too low." (It doesn't matter whether the price really is "too low" or not, since that's a subjective decision made by the buyer or the seller. What is important is that the producers feel they could be making a lot more money on widgets.) Suppose that the producers blame "oversupply" for the "low" prices‹they feel that there are too many widgets on the market. Each widget producer can try to drive out competitors, so that there will be fewer widget producers. The only way to do this on the open market is to drop the selling price until the weaker widget producers are driven out of business: by competition, in other words, in a "price war". During this time, however, all widget producers can look forward to receiving very low prices, and making little or no profit.

Suppose now that many widget producers feel that they cannot afford to gamble on trying to drive competitors out in a price war. They may instead make a mutual decision to change the market so that widgets sell for more than they would expect to receive in an open market. There are several ways they might try to do this. They might limit widget production, so that market forces then keep the price high without further action or organization. This is the simplest way to keep prices high, but it does require that actual production be less than full potential production. The necessary limitation on production has to be agreed among the members. Once the limitation is decided, the cartel must be in able to limit the global supply to that level. Usually this means that all major producers must be members. If this is the case, then the cartel members must make the most difficult decision: how many widgets each of them will be allowed to produce. Each member will be anxious to produce as many widgets as possible, since the cartel's success will raise widget prices, and the level of income and profit will be high. Each member must perceive its own allotted widget quota as fair, or at least it must be convinced that it will make more money by joining the cartel than it would make on the outside. Since widget prices will be raised by a successful agreement, each member of the cartel could make more money by cheating on the agreement, producing more widgets than its quota, and selling them secretly. If everyone were to do this, the widget trade would quickly return to pre-cartel levels, with a large supply of widgets selling at much lower prices.

If widget prices go up, the buyers may react in three different ways. They may decide that they can make do with fewer widgets, or they may decide to wait for a while before they actually buy more widgets, or they may use some substitute to do the same job. Therefore, a successful cartel usually deals with a commodity for which demand is "inelastic," that is, there is a strong and continuous demand for it, and nothing else can easily be substituted for it.

If widget prices go up, entrepreneurs may be tempted to open new production facilities. The cartel would be faced with two unpalatable outcomes: they might co-opt each new producer, which would mean giving up a share of the widget production quota to accommodate the newcomer into the cartel; or they might have to drop widget prices to keep the newcomer out, essentially removing the effect of the cartel. Therefore, a successful cartel usually deals with a commodity for which supply is also "inelastic," that is, new supplies cannot be brought on to the market quickly, either because the time scale for bring new production on to the market is very long, or because the capital needed to open new production facilities is very high.

Overall, if a producers' cartel is to be successful by limiting production, it means that

Alternatively, producers might limit supply rather than production. Perhaps production schedules are rather inelastic, that is, it might be costly to shut down or re-open a widget factory, so that it has to be kept operating even if it is very difficult to sell the product. (For example, mines have to be kept drained even if no ore or coal is being produced; so a company would tend to keep on mining too.)

Perhaps demand for a product is very elastic, that is, there are years of very high demand, and years of very low demand. In such a case, producers may decide to keep producing, but stockpile the products, keeping them off the market at times of low demand (and low prices), to be sold later when demand (and prices) are higher.

A cartel of producers can operate this way, provided they have a product that can store without deteriorating, and are willing to go to the expense of storage. The expense may not be entirely for physical storage: the cartel may need a lot of capital to keep production going while sales are low. Once again, the members of the cartel must act together to regulate the market.

All producers' cartels are relatively simple in principle to operate, because all the likely opposition is "downstream," from buyers. But increasingly in the 20th century, problems have arisen because there are different interpretations of "producer." For example, is Australia the world's largest producer of diamonds, or is it the Argyle Diamond Company, which is largely owned by the Malaysia Mining Corporation of Kuala Lumpur and Rio Tinto of London? What would happen if the interests of these parties diverged? The tensions between the multinational corporations that are so often the technical "producers," and the national governments that are political and geographical "producers," have not yet been fully studied by economists.

But a number of political thinkers have questioned the relationship between the OPEC states, the giant oil companies that distribute the world's oil, and the oil-consuming nations whose citizens are the dominant owners of the oil companies, and have wondered about the problems of conflicts of interest that are inherent in the dynamics of international oil. Much the same problems exist, on a less important geopolitical level, in many other primary industries. These questions bring us to the next levels of complexity, those of

Traders' Cartels

John D. Rockefeller saw very quickly that if he could influence oil trading to make a lot of money, it was no use trying to do anything about production. Then, more than now, production levels changed very quickly. Entrepreneurs would bring in a new oil discovery, and oil prices would crash. They built again with increasing demand, and would crash again. There was no way of storing oil production; there was too much chance of a cartel being broken by a new find from outside; and in those days the capital cost of entering the oil exploration business was minimal. Rockefeller chose instead to attack oil refining and trading, where capital requirements were higher, and where all oil production had to pass through a rather narrow bottleneck. The result was the astounding short-term success of the Standard Oil Trust in building a virtual monopoly over the oil business before 1911. Standard Oil was not a producer at this time, but rather a trader in oil.

This allows us to see that the giant oil companies of today are producers only in some countries, such as the USA. They are not producers in most of the OPEC countries (though they may very well employ geologists to find oil and engineers to pump it out of the ground). They are traders, acting in the middle ground between the producers and the buyers, just as Standard Oil did in the late 19th century.

Obviously, then, traders can build successful cartels: the most successful one is the CSO, the "Central Selling Organization" of the De Beers diamond company. Once based on the dominant production from De Beers mines, this is now a trading organization that still controls 80% of the world's gem diamond supply. Many of the motivations for trading cartels are the same, and most of the conditions still apply. There are additional complications, however. Traders try to buy low and sell high. Thus the prices they pay, as well as the prices they ask, are subject to fluctuation and to attack. A traders' cartel is vulnerable to whatever pressure the producers can bring to bear "upstream," as well as the "downstream" pressures already discussed. Producers can form a cartel (or a counter-cartel) upstream of traders just as well as upstream of the "buyers" we discussed already.

Because a traders' cartel is vulnerable to attack from both upstream and downstream, it is likely to be shorter-lived than a producers' or a purchasers' cartel. One answer to that problem is for a cartel to try to become vertically integrated, so that it can control much more of the flow of goods through from production to final sale. The giant oil companies, and the giant aluminum companies, are the best examples of such vertical integration earlier in the 20th century. But such examples are few, because of all the conditions that must be satisfied before such cartels can work successfully. Note too that the giant oil companies no longer control production in the way they did forty years ago, and a trend of the 1990s is for the larger companies to slowly move out of refining and retailing in the United States too, because of shrinking profit margins. Thus I can go to a gas station and fill up my tank with "BP" brand gasoline, knowing that BP has not pumped, shipped, refined, or sold one drop of it. BP produces oil in Alaska and ships much of it to Japan, and the gasoline I have bought was refined by Tosco from any crude oil it could buy cheaply!

There are thus many intrinsic restraints on the succes of any would-be cartel: the conditions for success are quite limiting. It is not surprising that there have been many short-lived cartels. It is surprising how many long-lived and successful cartels there have been.


Aluminum was a chemical curiosity when it was first discovered in 1807 by Sir Humphry Davy, but its extraction was so difficult that the first large pieces were not available until 1855. It was quickly discovered that the metal gained a thin but effective protective coating of aluminum oxide in air, so that it was very durable. Aluminum had a high melting point, but was easy to cast, mill, and polish; it was light and fairly strong, and clearly had a lot of potential as a useful metal. Unfortunately, the process for purifying it was expensive and required a lot of electricity, which was simply not available at the time. The small amounts that were first produced were costly. Aluminum sold for $500 per pound, twice the cost of gold. Yet the metal was fashionable, much in demand for luxury items because of its novelty and scarcity. A dining set of aluminum was presented with great pomp to the Emperor Napoléon III of France, and was used at royal banquets! The Emperor financed projects to make the metal cheaply, because he was interested in it for military purposes: light body armor and helmets, for example, for French Army cavalry. By 1859 the price was down to $17 per pound (the price of silver at the time), and by 1887 to $8 per pound, still far too expensive to compete with existing industrial metals except for luxury or prestige items. In 1884 the Washington Monument was capped by a thin sheet of aluminum weighing a little over 6 pounds, at a total cost of $225.

An intensive search for an industrial-scale method of producing aluminum was solved simultaneously in France and the United States in 1886, when Paul Héroult and Charles Hall independently discovered and patented the method that is still used today. (Hall invented the smelting process in his woodshed, but died as the wealthiest inventor in the United States, and the major benefactor of Oberlin College. Héroult invented the process in the family tannery, which he had dropped out of college to run.) The race was now on to create and satisfy a global demand for aluminum and its products, and that has succeeded to the point that aluminum production now exceeds that of all other metals except iron.

Aluminum, Energy, and Recycling
Almost all aluminum is produced from bauxite, the name for a group of ores that are basically aluminum oxide Al(OH)3. The ore is refined by heating with a mixture of alkaline salts to produce about half the volume of alumina, which is purified aluminum oxide, Al2O3. Using enormous amounts of electricity, alumina is then heated to drive off the oxygen and produce ingots of pure aluminum, Al, for use in everything from planes and submarines to cans, kitchenware and foil.

The cost in dollars, and the cost in environmental impact of such an energy-intensive process, both depend very much on the cost and style of power generation. Making a tonne of aluminum means, on average, releasing 13.1 tonnes of CO2 to the atmosphere. But if the power for the process comes from a coal-fired power plant, the CO2 load rises to 20.6 tonnes, while if the process is driven by hydroelectric power, then it falls to 5.6 tonnes. Furthermore, since the major power drain is in the separation of aluminum from its oxide, the simple melting and processing of recycled aluminum represents a tremendous power (and cost) savings: a tonne of aluminum made from recycled metal costs only 0.5 tonnes of CO2 released.

The major success of recycling aluminum in the United States is one of the few unmitigated environmental successes of the last two decades, probably because of the (enlightened) self-interest of the companies that realized the potential cost savings in recycling. 63 billion aluminum cans were recycled in the United States in 1992, enough to make two-thirds of the "new" aluminum cans produced.

Early History of the Aluminum Cartel

The processes for extracting aluminum on a large scale were not invented until the late 19th century, and because they were patented, a few companies were able to dominate the industry right from the start. This pattern has continued, but it has been maintained not simply by its historical beginnings, but by cartel action by all the large existing companies.

Alcoa's Vision: Pre-emptive Strike Against Competition
Let's look at Alcoa's position from its point of view, say around 1894. The company had closed down its only American rival, and its production was increasing dramatically. Its costs per pound were dropping with economy of scale, its sales looked strong, and it could look forward to 15 years of patent protection before another rival company could be started in North America. But its founders had a longer vision than that. What could and should Alcoa do now, to cripple any prospective competion even in the far distant future after the patent protection had expired?

Alcoa's early electrical supplies were generated from coal-fired steam engines, but in 1893 the company was already looking outside the Pittsburgh area. In June 1893, Alcoa became the first customer of the new Niagara Falls Power Company, signing up for hydroelectric power in advance of construction. The company built smelters at Niagara Falls, and its costs dropped dramatically as soon it moved all its smelting to the new plants in 1896. Alcoa could drop the price for aluminum, to attract new business for new uses for the metal, yet make a healthy profit at the same time.

But the experience of planning the Niagara plants told Alcoa that start-up costs for new aluminum plants (at the scale the new industry would be operating at) would be heavy. Yet there were such economies of scale that they had little to fear from any small-company rivals. The two major considerations for any potential large-company rival would be access to large deposits of bauxite, and access to huge quantities of cheap electricity. Alcoa moved to block both those avenues, as a pre-emptive strike against future competition. The strategy was brilliant, and the way they executed it was ruthless and effective. In the end, Alcoa staked out all the best sources of North American bauxite for itself, and sequestered all available cheap electricity.

Bauxite had little commercial value before the 1890s, so geologists had not explored for it. The known deposits were few, and because of the bulk of the material, any new deposits had to be close to waterways or the ocean, because the only economic way to ship bauxite, or refined alumina, was by boat.

Alcoa bought up three major mining companies operating in the best-known bauxite deposits in North America, which were in Georgia, Alabama, and Arkansas: Georgia Bauxite and Mining in 1894, General Bauxite Co. (in Arkansas) (1906), and Republic Mining in 1909. Not only did this ensure Alcoa's future supply of bauxite, it prevented any competitors from access to these resources. Any competitor would have to prospect for and find a new bauxite deposit (which would be further from the American industrial heartland at the time), bring it to production, and set up shipping facilities for it. Given that Alcoa's bauxite was coming from proven high-quality deposits, by cheap and established shipping routes, this immediately placed a substantial cost barrier in the face of any company even thinking of entering the aluminum industry.

An Alcoa board meeting in 1896 discussed the other prong of the strategy, cheap electricity, which for all practical purposes meant hydroelectric power. Despite the Niagara Falls contract signed the previous year, the board of directors discussed reaching into Canada, where there were very promising sites for hydroelectric power. It would be necessary, said the board, "to preclude the establishment in the future of works to manufacture aluminum at some of the Canadian waterpowers" (a very clear statement of the strategy). Alcoa thereupon signed two exclusive long-term contracts with new projects, one at Shawinigan Falls, Québec (1899), and the other with the St. Lawrence River Power Company (1903), and built its own power plant at Niagara Falls (1906). Any potential rivals would now have to face developing new power sources at more remote locations.

Alcoa continued its strategy long after patent protection had ceased. It organized bauxite mining companies in Surinam and Guyana in 1916 and signed 75-year exclusive leases with them; it built dams along the Tennessee River after 1914 and in North Carolina in 1915. It built the company towns of Bauxite, Arkansas (for mining) and Alcoa, Tennessee (for smelting). The only major challenge to its power policy came in 1925 when an aggressive power company built a major project on the Saguénay River, in Québec: at the time, the power potential of the new site was three times the power to produce all the world's annual aluminum production, and it would be produced at a quarter of the typical cost. Alcoa's response was to buy out the company: it cost 9% of the shares in Alcoa, but it brought yet another safety buffer against the entry of any potential rival.

It's not only the evidence from the 1896 board meeting, but the data, that show Alcoa was following a deliberate strategy. Around 1917-1920, Alcoa owned or had an exclusive lease on about 500 years' worth of bauxite reserves, and owned or had exclusive rights to 4­5 times the hydroelectric capacity it actually needed. After the Saguénay purchase of 1925, Alcoa had rights to ten times the power it actually needed. You can be sure that the excess electricity would not have been sold to a rival aluminum company! Now although Alcoa had accumulated the properties and leases fairly cheaply, the cost of maintaining the buffer was not zero: it was an investment in the future, assuring Alcoa of continued mastery of the North American aluminum industry and market. As such, it was a brilliant and successful strategy, in purely economic terms.

Alcoa, as the sole significant North American producer, was in a position to charge what the market would bear. Again, the data show that Alcoa, which had been running at a profit margin around 4 to 6% in the 1890s while it dropped prices and increased the market for aluminum, began stabilizing prices around 1900 and increasing its profit margin to more like 15%. Between 1905 and 1919, profit margins were around 20%.

How the cartel evolved.

At the turn of the century, the only danger to Alcoa came from outside: foreign rivals might begin to export cheaper aluminum to North America. Alcoa lobbied for, and received, protection by import duties on foreign aluminum ingots. The duty was 8 cents a pound after 1897, but this is in relation to a selling price around 33 cents a pound. Import duties were much higher on finished goods. This gave Alcoa an advantage in the American market that simply could not be successfully attacked by foreign companies.

There were five significant early aluminum companies: one American, two French, one Swiss, and one British. The European companies competed fiercely with one another at first, then signed a series of cartel agreements after 1896.

In 1901 the first world-wide aluminum cartel was formed by the five major companies, who were producing 90% of the world's aluminum at the time. Alcoa joined in the negotiations, but formed a 100%-owned Canadian subsidiary (the Royal Aluminum Company, eventually Alcan) two days before the agreement was finalized, to sign on its behalf, avoiding legal problems at home in the United States under American anti-trust legislation. Each company would be free to sell within its own country in a closed market, at a price around 1 cent a pound higher than elsewhere. Then each company was allotted a guaranteed share of the rest of the world market (Alcoa, as Alcan, received 21%). As a result, all the firms were able to charge around 36¢/pound in Europe (instead of the 22 cents they had been receiving), while Alcoa charged 33¢/pound in North America for a few years to stimulate demand, then charged to the cartel price of 36¢/pound from 1906 to 1908, by which time it was making an annual net profit of 35% on stockholder equity.

By 1911 Alcoa was running into the same antitrust accusations that had led to the break-up of the Rockefeller Standard Oil Trust. A complaint from the Department of Justice ended without trial, with Alcoa cheerfully signing a pledge to avoid antitrust behavior within the United States (something it was bound to do by law in any case), and particularly not to join in any more European cartels. Five days later its newly formed, totally-owned Canadian subsidiary (Alted, now Alcan) signed all the usual undertakings with the European cartel members. Alcoa operations in the United States carried out all the provisions of the cartel agreements without actually signing anything.

Alcoa may have been saved by the outbreak of World War I. The war caused a tremendous increase in aluminum demand, at the same time that it disrupted the European economies. New companies were built in neutral countries, especially in Norway, which was in the process of developing its enormous hydroelectric potential.

But as soon as the war was over there was surplus producing capacity. The Europeans began to dump surplus aluminum on the American market, and Alcoa tried to protect its market. This time it did not propose a cartel agreement, partly because Andrew Mellon, a major shareholder in Alcoa, became Secretary of the Treasury in 1921.

Instead, Mellon organized a protective tariff of 5 cents a pound on imported aluminum with effect in 1922. Alcoa raised its domestic prices by 6 cents a pound, and began to attack its rivals in Europe, buying large holdings in companies in Norway, Italy, Yugoslavia, and Spain that had overextended their resources. In 1923 it bought half of Norsk Aluminum. By 1923 the Europeans were ready to talk about market controls, and a new formal cartel agreement was signed in 1926. Alcoa's Canadian subsidiary Alcan signed on its behalf.

By 1930 the Alcoa/Alcan combination controlled 50% of world production. Alcoa was insulated from accusations of complicity in the cartel largely because Andrew Mellon was Secretary of the Treasurer from 1921 to 1932. When the Attorney-General began to prepare an antitrust case in the 1920s, President Coolidge took care of the problem by appointing him to the Supreme Court!

In 1931 the cartel was formalized as a gigantic company, Alliance Aluminium Compagnie of Switzerland, including every major aluminum producer except Alcoa. The cartel was very successful during the 1930s. Aluminum prices were kept stable through the worst of the depression until re-armament programs raised the demand again. Alcoa's market share of aluminum sold in the United States averaged over 90% from 1899 to 1940, and even some of the 10% it did not produce came from its sister company Alcan.

Finally, Mellon got into trouble with Congress, not for complicity in the aluminum cartel, but because it was felt his mismanagement of the Treasury had contributed to the severity of the Depression. President Hoover finally rescued him from resignation and/or impeachment by appointing him ambassador to Britain. In 1937 the long-awaited complaint was filed against Alcoa, but the court proceeding dragged on for 2 years. The simple fact was that Alcoa produced 100% of the bauxite, 100% of the alumina, and 100% of the primary aluminum produced in the United States. The more complex question was whether Alcoa had accomplished that market share fairly.

World War II had begun by the time Judge Caffey had fought through 58,000 pages of testimony and found Alcoa not guilty on every one of 130 charges (it took him 9 days to read his verdicts). The Government's appeal was not resolved until 1945, with a decision against Alcoa. This time, Judge Learned Hand ruled that the very existence of the monopoly was a Bad Thing, no matter whether Alcoa maintained it unfairly or not. "I think," said a senior Alcoa official, "we are being penalized for our successful free enterprise", and that judgment seems to be largely true in terms of Alcoa's internal policies. It seems that Alcoa's influence on the international cartel, never proved in the case, may have had more to do with the verdict than it should have done. The suit had lasted longer than the American Civil War, and meanwhile, by 1945, Alcoa and Alcan were producing 64% of world aluminum.

Even before the United States entered World War II, Alcoa assured the Government that it would be able to meet wartime needs for aluminum. However, Alcoa was wrong. The demand was unprecedented, and Alcoa was falling short even early in 1941. Eight new aluminum smelters were built by the US Government during World War II, several of them associated with the giant new hydroelectric projects of the Northwest, doubling the capacity of the United States. After the war these plants were considered surplus, and they gave the government the opportunity to break Alcoa's American monopoly‹they were offered for sale, but Alcoa was not allowed to bid. Even then, Alcoa's long-term advantages in bauxite and electricity supplies made it difficult to contemplate open competition. Reynolds and Kaiser eventually bought the new plants, but only after they were made available at about one-third of their cost (and much less than Alcoa was prepared to pay).

The Bauxite Producers

The tropical nations that have most of the richest bauxite deposits in the world do not have many smelters and refineries to process the ore. Most of the final processing is done in economically advanced countries. Some of this discrepancy is pure economics: the great hydroelectric surpluses of the northern nations, especially Canada, the US, and Norway offer very cheap electric power in enormous quantities. However, there more be more than simple economics to this situation.

Suppose a poor bauxite-producing nation wants to nationalize its aluminum industry. The consequences may be much more severe for the nation than for the aluminum company. Usually companies do not depend on a single nation for bauxite supplies, so they can turn to other sources who are only too anxious to sell more bauxite. On the other hand, the producer nation may be left with bauxite deposits that it can neither sell (because the cartel blacklists it), nor process (because it lacks the refineries and smelters, and the electricity, and cannot afford the capital investment to build them).

So if a producer nation depends heavily on bauxite income, it cannot afford to nationalize the company, and does not have much leverage to insist that refining and smelting be located there. Therefore, the separation of processing from mining is a very powerful weapon that a company can use against a poor bauxite-producing nation. The aluminum companies, as long as they act as a traders' cartel, are in a strong position to dictate terms to most bauxite-mining nations. Most of the value is added in the later stages of aluminum production. The mining end of the industry is typically high-volume, low-profit.

Early on, the few giant aluminum companies made a great deal of money, not just by operating a cartel that kept consumer prices high, but also by gaining access to bauxite ores very cheaply, because the deposits were in areas that had very weak political and economic leverage. The companies ran the aluminum industry to maintain that advantage.

Page last updated May 1999.

Return to Geology 115 main page.

Return to Geology Department Web page