Multiunit Organization and Multimarket Strategy: Volume 18

This paper explores whether membership in ethnically based groups influences conduct and performance in a spatially dispersed industry. We test two propositions. First, when group members own several units in a given market, prices and revenues might be similar to when just one person owned these units. Second, if group membership provides such performance benefits, members may locate their units to obtain these benefits. Using over a thousand hotels located in rural Texas, we test the role of ethnic group membership by examining hotels owned by individuals with the surname “Patel”. This surname is common in the hotel industry, with some estimates of Patels owning one in four U.S. motels. Though most are not directly related, much anecdotal evidence suggests that Patels try to lessen the competition among their units. Interestingly, we find that for Patel-owned hotels, proximity of other Patels confers no performance gains. While collocating at a regional level, Patels do not collocate at the finer zip code level, which prevents them from coordinating to enhance performance.

In this paper we differentiate between the creation and subsequent exploitation of multimarket contact. We examine specific factors that influence the likelihood that a firm will seek to develop a purposive set of overlapping markets with specific competitors, as opposed to developing naë contacts based on an internally derived logic. We suggest that competitor identification, organization structure, ease of competitive response and industry structure (including the presence of network externalities, barriers to entry, and industry growth rate) all play important roles in determining the ability of managers to seek rivalry-reduction that has been found to follow the development of market overlap. We offer several propositions that serve to define the boundaries of research into mutual forbearance and multimarket competition, and that may help to explain empirical results obtained to date.

Firms face a choice in the organization of production. By concentrating production at one site, they can enjoy economies of scale. Or, by dispersing production across multiple facilities, firms can benefit from product-specific efficiencies and enhanced organizational learning. When choosing to organize in multiple units, firms must also decide where to locate these units. Concentrating production geographically can enhance economies of scale and facilitate organizational learning. On the other hand, dispersing facilities might allow the firm to lower transportation costs, reduce risks, and forbear competition. To examine these tradeoffs, we compare exit rates of single-unit organizations to multiunit organizations and their constituent plants in the U.S. footwear industry between 1940 and 1989. Our results suggest that, multiunit organizations benefit primarily from enhanced organizational learning, competitive forbearance and the diversification of risk. Nevertheless, these benefits appear to come at the expense of organizational adaptability.

Interorganizatonal relationships in general and chain relationships in particular are a critical channel for interorganizational learning. Learning may not only be a result of interorganizational relationships, however; it may also be a primary cause of them. We examine this idea in the empirical context of Manhattan hotels and their relationships with hotel chains. Our analysis shows that hotels are likely to form relationships when they have very low, or very high levels of their own operating experience. The relationship between hotel and chain is less likely to dissolve when the chain has more operating experience in the hotel's local market, and more likely to dissolve when the chain has more operating experience in non-local markets. The duration of a chain relationship has a ∩-shaped effect on its dissolution, indicating that relationships go through a honeymoon period, and that the parties to a relationship learn to better collaborate over time.

This paper addresses the topic of knowledge transfer between firms that meet in multiple markets. Theoretical arguments are developed regarding the issues thought to influence the process of knowledge transfer between multimarket firms, in general, and also regarding the emergence and influence of macrocultures on this process. We use a structural equivalence lens to develop this theory, suggesting that multimarket competitors are likely to share macrocultures to the degree that their markets overlap. The structural equivalence of multimarket competitors influences the direction that they are likely to look for knowledge resources.

This paper discusses the design of hypothesis tests about multimarket contact and uses a sample of chemicals firms to illustrate researchers' choices and to provide new evidence about the effects of multimarket contact in product and innovation markets. Multimarket contact has a large, stable and statistically significant association with a firm's citations of the patents of another firm. The effect is present even when controls are added for each paired firm's pertinent characteristics (including the effects of its product and innovation markets) that would be expected to affect citations or reflect the effects of the research underlying the citations. New probability measures of multimarket contact are introduced and discussed, and their performance is compared with simple count measures.

Two duopoly market structures are experimentally constructed. Payoff tables describe basic conditions in an X market and a Y market. In one market structure two rivals choose X quantities and a different set of rivals choose Y quantities. These markets develop equilibria independently. A second market structure creates a multiproduct producer that chooses both an X and Y output. Separate X and Y duopolists are rivals to the multiproduct producer. The multiproduct/multimarket rival creates a horizontal connection for the X and Y markets. Subjects choose outputs from payoff tables for at least thirty five periods. Compared to the independent duopolies, the horizontal connection raises outputs in the X market and lowers outputs in the Y market. These impacts are statistically significant. We find Granger causality for output choices moving from X to Y and from Y to X in the connected markets; however learning that moves through the horizontal connection is less influential than learning within the market. Further testing shows that competitive learning travels better than cooperative learning through the multimarket firm.

Studies of multimarket contact typically hypothesize (and find) that the relationship of contact levels between firms and the likelihood of aggressive competitive behaviors is linear and negative. This includes the likelihood of entry moves by multimarket rivals. However, a linear negative relationship between market entry and multimarket contact, which implies that deterrence exists at lower levels of contact, makes it unlikely that firms' will enter into each others' markets in order to create that deterrent.

We address this theoretical issue by distinguishing between multimarket contact and multimarket competition. Lower levels of multimarket contact provide models for a focal firm that both guide and legitimize its strategic moves. As multimarket contact levels increase, multimarket competition eventually ensues as multimarket rivals recognize their mutual interdependence and further entry into shared markets is deterred. We develop the logic for this non-linear relationship by incorporating theory from the strategic, institutional, and decision-making literatures that has direct implications for the way in which contact between firms across multiple markets is likely to affect managerial cognitions and firm actions.

Since firms must enter into the markets of their (multimarket) competitors to build multimarket structures which yield competitive benefits, we conclude by developing several propositions about how such entry might best be achieved to avoid retaliatory responses by multimarket competitors.

The so-called information revolution has loosened many tongues in the academic, business and policy worlds. The communis opinio is that the diffusion of modern information technologies in the global village is about to radically change the rules of the competitive game in many, if not all, industries. What the new rules may be, nobody really knows for sure. The New Economy, being characterized by the penetration of information technologies and the dominance of network arrangements, is born. Firms face the daunting task of developing a response to this new environment. How they respond depends in part on what they learn from their contact portfolio. Which contacts they have, and what information they get out of these contacts, shapes their response to the New Economy. A case in point is the Internet. Apart from stimulating the emergence of new value-adding products and efficiency-enhancing processes, an argument is that the Internet will shake up the century-old market institution by introducing information transparancies and network economies. This chapter's objective is fivefold. First, we summarize the current state of the art in the multimarket competition literature. Second, we develop a multichannel contact framework that builds upon the established multimarket contact theory. Third, we briefly evaluate the New Economy debate. Fourth, we analyze what the emergence of this New Economy may imply for issues of competitive rivalry and strategic management in the context of multichannel contact management. Fifth, we illustrate part of the argument by briefly discussing the case of Java software.

The expansion of markets and industrialization greatly increased the benefits of specialization in the U.S. economy. However, since the benefits of specialization can only be realized through trade, specialization significantly increases the volume of market transactions in the economy. The analysis presented in this paper suggests that a better understanding of the historical changes in the nature of market transactions costs, especially those related to information, is likely to provide considerable insights on the rise of the modern business enterprise and a richer understanding of the industrial organization of the U.S. economy.

Research linking firms' hypercompetitive strategic actions and performance focuses on firm level performance, and assumes that the impact of each firm's strategic actions affects only a limited number of other firms. We advance a complementary perspective for multiunit-multimarket firms in which the action performance linkage operates at the market-unit level (i.e. within a specific geographic product or service market), and strategic actions within a market-unit are assumed to affect all other firms participating in that market-unit. Supporting the perspective, our analysis of Canadian insurance firms shows that these firms' performance in a marketunit depends on the number and simplicity of their strategic actions in that market unit - both absolutely and vis-d-vis those of others competing in the market unit. Our study offers new insight into how multiunit-multimarket firms compete across a series of market-units - and how they can compete more effectively.

The growing literature on multimarket contact and mutual forbearance in management and economics has produced an inflation of multimarket contact measures. The lack of validation of these multiple measures has hindered the accumulation of consistent knowledge and comparison of empirical findings. This paper investigates the measurement of the multimarket contact concept. Specifically, we review the existing measures of multimarket contact, identify the main differences among them, and evaluate their reliability and discriminant and predictive validity at multiple levels of analysis, both cross-sectionally and longitudinally. The results indicate substantial differences in the reliability and discriminant validity of these measures. Predictive validity depends critically on the level of measurement and on whether longitudinal or cross-sectional correlations are considered.