------------------------------------THE DYNAMICS OF A STABLE CARTEL: THE EXPRESS 1851-1913

Peter Z. Grossman
Washington University

The most successful cartel in U.S. business history was almost certainly the railroad express cartel. This cartel, which controlled the parcel post service in the United States, persisted for more than half a century, by one reckoning from 1851 to 1913. During that time, the five firms of the cartel agreed on prices, territory and governance procedures, and , by and large, maintained a harmonious environment within the industry. Stable cartels are, of course, rare historical phenomena. Cartel members typically find it too attractive to violate agreements in order to capture market share. This paper explores the reasons why the express industry proved to be an exception to the general rule.

The express cartel began with a territorial agreement signed in 1851 between Adams and the American express companies to divide the country east of the Mississippi River between them. Over the next decade, three other firms--Wells, Fargo & Co., the United States Express, and the Southern Express--were permitted entry. Thereafter, the five companies apportioned market share. Initially, firms were designated geographic territory, but later they were assigned monopoly control over railroad routes. Where routes overlapped (and members could have competed), the cartel firms fixed prices. Governance procedure were established to adjudicate disputes between members, although in times of crisis, ad hoc bodies might be formed to settle disagreements. Conflicts occurred, but the cartel never broke down. Its eventual demise came at the hands of government that effectively regulated it out of existence in the early part of the twentieth century.

The paper focuses on two key elements in the success of the express cartel--elements that would seem essential for the long-run success of any cartel: 1) its proficiency at deterring (or at least limiting) entry, and 2) its ability to prevent defections among its members. Indeed, in the case of the express cartel, these appear to have been linked. To deter defection, cartel members threatened one another with severe retaliation, typically money-losing price wars. This paper argues that such threats were credible because entry barriers allowed the cartel to function as a monopoly and earn rents. Thus, there was a substantial payoff to long run stability even at the cost of short run losses. To an extent, the express seems to bear out the model of Abreu (1986) who argued that the extent of collusion was determined by the severity of credible punishment strategies. In the case of the express, credibility emerged from the character of the industry.

Entry Deterrence

The express faced potential entry from two kinds of organizations: new express companies and railroads. But the industry had created barriers that for the most part prevented entry by any outside organization. The first, and most significant, barrier resulted from a series of contractual arrangements that allowed the cartel to form a national transshipment system. The system facilitated the shipment of packages throughout the country by any cartel firm. Therefore, any potential entrant faced the prospect of duplicating not only fixed capital, but also a procedural arrangement that lowered transaction costs to cartel members. Even when an entrant had a production cost advantage, if it were denied access to the transshipment system, the new firm still might have had higher costs overall than the incumbents.

The cartel members also maintained capital sufficient to produce a level of output above the amount demanded of a monopoly producer. This represented a precommitment to remain in the industry, and entrants were left without available profit margins that could easily be captured. An entrant had two alternatives: attempt to grab market share from an existing firm or stay out. The former strategy guaranteed a large expense and a price war.

In models with precommitment, price wars against entrants may be credible by definition. But the history of the express industry provides clear evidence of the employment of such strategies. This was made plain throughout the early history of the express, most notably in 1867 when an outside firm, the Merchants Union Express, attempted what was effectively the last large-scale entry into the express industry. The entrant tried to duplicate a significant portion of the cartel system--largely in the territory controlled by American Express. After spending money to create an express network, the Merchants Union faced an all-out price war. Within a matter of months, the Merchants Union lost the better part of $20 million--its paid in capital. Although American Express could probably have won if the conflict persisted, the incumbent's own losses were severe enough so that it chose to buy out the Merchants Union with an issuance of new stock. Nevertheless, the entry attempt proved that when threatened incumbents would fight a money losing price war, and could impose the higher costs on the entrant. Even though the original shareholders of the Merchants Union received a buyout, they clearly had made a poor investment; American Express paid only $0.45 to the dollar. As the express grew richer in the following years, the cartel's wherewithal to fight made a successful general entry even less likely, and none was attempted.

Entry attempts by the railroads posed a different problem. A railroad entry was always local; it could only take over the express service on its own line. But there was a great temptation for rail companies to try it. Because railroads held a cost advantage by control of transportation, operation of the express would appear to have allowed them full monopoly rents instead of the share they received through express contracts. But the contractual network--the transshipment system--made the cost advantage illusory. Denial of access to the transshipment system raised transactions costs sufficiently so that entry was almost always a poor decision by a railroad. Railroads were further discouraged from entry by threats of price wars.

At the same time, the cartel depended on railroad transportation and offered inducements to railroad companies to maintain harmony and discourage entry. Essentially, the standard contract gave the railroads a share of gross express revenues, and the express sought to make the percentage high enough to discourage conflict. Institutional change after 1887 gave additional incentives for railroad-express cartel cooperation. The creation of the ICC forced the railroad companies to reveal. and at times alter, their business practices. The express were initially exempt from ICC oversight. Express revenues, which the railroads received, could be counted on to reflect profit-maximizing behavior instead of regulatory caprice.

Cartel Interaction

In 1880, a railroad pamphleteer portrayed the express cartel as a harmonious organization. This depiction was largely accurate. By a series of contracts, the members fixed rates over common routes and apportioned control over railroad lines to members. The contracts also stipulated governance procedures including bodies to adjudicate and penalize cheating. Because conditions changed, prices and territorial division were frequently revised, but by and large agreements were followed. Though penalties and adjudication procedures existed, there is little evidence that they were ever employed.

This was especially true with respect to prices. Although common routes, and with them the opportunity for price competition, expanded throughout the period, firms seldom deviated from price agreements. In fact, as the opportunity for price competition grew so did the willingness to employ severe punishments to prevent it. The threat of severe retaliation--a money losing price war on a large scale--was the key deterrent. As noted earlier, this was made credible by the success of the cartel's ability to maintain a monopoly. Firms gained monopoly rents. Since price competition suggested that the rents would be dissipated, there was the willingness to accept temporary losses to prevent them.

In fact, the monopoly character of the industry raises a theoretical point. Unlike a standard cartel where the alternative to collusion is competition, the alternative to the express cartel may well have been a single firm monopoly. More generally, if a cartel can maintain successful entry barriers there is no inherent reason why, if it breaks up, a single firm may not be able to sustain those barriers and gain control over the entire market. Price defections then might be seen as an attempt by one firm to drive out the others and gain monopoly control of the market. In that case, a more extreme punishment strategy becomes credible: defection is met by an all-out price war that continues until the defector is forced from the market--a punishment of collective predation. Thus defection is tantamount to suicide, and consequently no one defects; collusion is sustained. While in general the express cartel did not issue such drastic threats, members did so on occasion, and there is evidence that the cartel saw collective predation as a real option. This was especially true of the early years of the cartel. Predation had been used by cartel members against rival express firms, and was threatened in early cartel disputes.

Less drastic price war threats were issued far more frequently and appear to have been taken seriously. As the model would suggest, their credibility was seldom tested since potential defectors knew that cartel members would have felt compelled to retaliate. Agreements became largely self-enforcing; cartel price agreements were generally upheld for fifty years.

Breakdowns, on the rare occasions they occurred, can be traced to exogenous events. (This supports conclusions by Porter in 1983 analysis of a leading rail freight cartel, the Joint Executive Committee.) For example, one period of intracartel strife can be linked to the entry attempt into the express of the New York, Lake Erie and Western Railroad in 1886. But after a period of price warfare, the railroad entry was defeated and cartel price agreements were reinstated.

Territorial disputes between express companies also occurred and, in fact, were more common than price disputes. Territory followed rail lines and thus were continually changing. As lines expanded, monopoly territories would suddenly have competition; overlapping routes would increase. Nevertheless, the cartel was generally able to adjust to changes in the railroad system. Members traded routes to maintain approximate market shares, or on occasion, they shared revenues and expenses on express traffic over certain lines.

Territorial conflict in the early 1890s, however, produced the greatest period of stress of the cartel's history. The United States Express Company sought to gain territorial advantages by offering railroads better contract terms. This effort was largely thwarted by the rest of the cartel though at a high cost to all firms. Indeed, it produced a unique and exceptionally severe punishment. Cartel members felt compelled to buy up a controlling block of US Express stock. This punishment temporarily ended the firm's autonomy and allowed the cartel to determine the defector's policies from within.

The express cartel succeeded largely because it was credibly able to punish both entrants and internal price defectors. This case poses a challenge for cartel theorists. The standard literature separates entry barriers and cartel stability. In the case of the express, the two were linked and as a result the cartel remained relatively stable for more than half a century.