CONFLICT AND PERFORMANCE IN MARKETING CHANNELS OF DISTRIBUTION: A MODIFIED MODEL Dissertation for the Degree of Ph. D. MICHIGAN STATE UNIVERSITY MORRIS PERRY 1989 CONFLICT AND PERFORMANCE IN MARKETING CHANNELS OF DISTRIBUTION: A MODIFIED MODEL By Morris Perry A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Marketing and Transportation 1989 ABSTRACT CONFLICT AND PERFORMANCE IN MARKETING CHANNELS OF DISTRIBUTION: A MODIFIED MODEL By Morris Perry There is general consensus among channel analysis that conflict affects performance. However, there is dissensus about the functional form of the relationship between the two constructs and under what situations a specific kind of relationship exists (linear or nonlinear). The objective of this dissertation was to investigate the effect of conflict on performance using a broadened conceptual and empirical framework that is perceptually based. To test the hypotheses established for this study, an experiment was conducted. The 2x2x2 quasi-experimental design with repeated measures on the dependent variable included four conflict situations which were chosen to have high and low levels of conflict, involvement, and compliance. Data were collected from two groups of respondents. One group provided data on the perceived relative effectiveness of resolution behaviors to resolve conflict, and a second group provided data on the perceived impact of the resolution behaviors on four performance measures. Both groups provided data on multiple measures of conflict and involvement. Morris Perry First, it was found that the effect of the conflict x involvement x compliance interaction was statistically significant for each of the four measures of performance. Second, it was found that the joint effect of conflict x involvement was significant for high and low compliance; conflict x compliance was significant for low involvement and insignificant for high involvement; and involvement x compliance was significant for high conflict and insignificant for low conflict. These findings provide tentative evidence that the relationship between conflict and performance is not linear. To ascertain whether the interaction effects were or were not contaminated by uncontrolled covariate variables (dealership size, experience, competition), analysis was done to remove their effects to isolate the pure effects of the experimental treatments. It was found that the three-way interaction effects remained unchanged across the four performance measures after removing the effects of the covariates. Also, the joint effects remained unchanged after removing the effects of the covariates. At least one covariate was significant in the full and joint effect models. These findings and theoretical contributions and managerial implications are discussed. ACKNOWLEDGMENTS The contributions by some special people facilitated the completion of this dissertation. First, I would like to extend my grateful appreciation to Dr. Forrest "Sam" Carter, chairperson of the dissertation committee, who has stuck by me for a long time with lots of encouragement and support during the entire process of writing this dissertation. He has contributed tremendously to my conceptual and methodological development. I also would like to express special thanks to Dr. S. Tamer Cavusgil for his valuable time and insights and for agreeing to work with me as a committee member when I was most in need. I also am very grateful to Dr. R. Dale Wilson, a dissertation committee member, for his valuable comments and suggestions. I was very fortunate that the members of my committee allocated their time to oversee and direct research efforts in an area that is outside their own research interests. My appreciation also goes to Dr. Robert W. Nason, chairperson of the department, for his support and facilitation of the data- collection process. Last but not least, I am indebted to Drs. Dogan and Sevo Eroglu, colleagues and best friends, for their support, encouragement, and friendly harassments, especially when distractions sometimes got in the way. iv TABLE OF CONTENTS Page LIST OF TABLES..................................................................................... viii LIST OF FIGURES...................................................................................... xii Chapter I. INTRODUCTION ..................................................................... 1 Research Approaches ........................................................... 1 The Structural Approach .............................................. 1 The Functional Approach .............................................. 2 The Commodity Approach ............................................... 3 The Behavioral Approach .............................................. 3 CriticalComponents of the Behavioral Approach . 5 The Importance of the Conf1ict-Performance Relation From a Theoretical Perspective ............ 8 The Research Problem ................................................... 8 Previous Findings ........................................................... 8 The Importance of the Conf1ict-Performance Relation From a Managerial Perspective ................ 10 Managerial Problem ....................................................... 10 Collaboration as a Mode of Managing Long-Term Channel Exchange Relationships ........................ 10 Inherent Problems in Collaborative Exchange Relationships ........................................................... 12 Contributions of the Research ....................................... 15 Plan of the Dissertation ............................................... 18 II. LITERATURE REVIEW ................................................................... 19 The Channel of Distribution ........................................... 19 Conflict in Channels of Distribution ........................ 24 Causes of Conflict ....................................................... 26 Effects of Conflict................................... 27 Conflict as a Process................................................... 29 Performance in the Distribution Channel .................... 29 The Classical Models of the Conflict-Performance Assumption................................................................... 32 A Behavioral Model of Channel Member Conflict and Performance........................................................... 39 v Page Involvement....................................................................... 43 Compliance....................................................................... 45 A Proposed Theoretical Framework of Channel Member Conflict and Performance Relationships . . 51 Establishment of Hypotheses ....................................... 56 III. METHODOLOGY............................................................................... 60 Experimental Design ........................................................... 60 Overview........................................................................... 60 Operationalization of Conflict andInvolvement . 61 Operationalization of Compliance ............................ 70 Measurement of the Dependent Variable .................... 79 Validation....................................................................... 87 Development of Data-Collection Instruments .... 87 Pretest of the Questionnaire .................................... 89 Data Collection............................................................... 90 Manipulation Checks ....................................................... 92 Analysis Methods ............................................................... 95 IV. RESULTS...................................................................................... 96 Manipulation Checks ........................................................... 96 Conflict........................................................................... 97 Involvement....................................................................... 104 Compliance....................................................................... 109 Reliability........................................................................... 110 Perceived Relative Effectiveness of Conflict Resolution Behaviors ............................................... 113 Impact on Performance ................................................... 119 Sample Representativeness ............................................... 128 V. TEST OF HYPOTHESES............................................................... 132 Theoretical Framework Revisited .................................... 132 Test of the Full Effect Model....................................... 133 Test of Simple, Simple Models (Two-Way Interac­ tion Effects)............................................................... 138 Test of Main Effects....................................................... 142 Relative Magnitude of Experimental Effects and Power.............................................................................. 150 Test for the Effect of Moderator Variables on the Experimental Treatment Effects .................... 157 Test for Parallel Regression Slopes ........................ 159 Fit of Separate Regression Slopes ............................ 161 vi Page VI. CONCLUSIONS............................................................................... 162 Discussion of Results ....................................................... 162 Limitations of the Study ............................................... 168 Theoretical Contributions and Managerial Implications ............................................................... 171 Future Research Directions ........................................... 175 Summary of the Research ................................................... 177 APPENDICES A. SELECTED INTERVIEWS ... ................................................... 180 B. QUESTIONNAIRE TYPE A............................................................ 221 C. QUESTIONNAIRE TYPE B............................................................ 228 BIBLIOGRAPHY .......................................................................................... 235 vii LIST OF TABLES Table Page 3.1 High Conflict-High Involvement CRB........................................ 72 3.2 High Conflict-Low Involvement CRB........................................ 73 3.3 Low Conflict-High Involvement CRB ........................................ 74 3.4 Low Conflict-Low Involvement CRB............................................ 75 4.1 Empirical Categorization of Conflict Issues by Group 1 Dealers....................................................................... 99 4.2 Results of Paired T-tests for Differences in Means (Group 1 Subjects)........................................................ 101 4.3 Empirical Categorization of Conflict Issues by Group 2 Dealers............................................................................ 103 4.4 Results of Paired T-tests for Differences in Means (Group 2 Subjects)........................................................ 105 4.5 Standardized Reliability Coefficients for Involvement Items............................................................................... 107 4.6 Univariate Statistics for Resolution Responses (CS = New Vehicle Allocation and Distribution Conflict) . . 114 4.7 Univariate Statistics for Resolution Responses (CS = New Product Technology Purchase) ........................ 115 4.8 Univariate Statistics for Resolution Responses (CS = Parts Availability).................................................... 116 4.9 Univariate Statistics for Resolution Responses (CS = Customer Satisfaction Index (CSI) Rating Survey) . . . 117 4.10 Univariate Statistics for the Performance Impact of CRR: CS « New Vehicle Association/Distribution .... 120 4.11 Univariate Statistics for the Performance Impact of CRR: CS = New Project Technology Purchase......... 123 viii Page 4.12 Univariate Statistics for the Performance Impact of CRR: CS = Part Availability.................................... 124 4.13 Univariate Statistics for the Performance Impact of CRR: CS = CSI Survey Use ........................................ 126 4.14 Percentage Comparison of the Sample and Population on Selected Demographic Variables .................... 131 5.1 MANOVA Results of the Full Model (Dependent Variable = New Car Sales Volume) ........................................... 134 5.2 MANOVA Results of the Full Model (Dependent Variable = Market Share) ........................................................... 135 5.3 MANOVA Results of the Full Model (Dependent Variable = Profits)........................................................................... 136 5.4 MANOVA Results of the Full Model (Dependent Variable = Overall Performance) ............................................... 137 5.5 MANOVA Results of Two-Way Interaction Model, Adjusting for High Compliance (Dependent Variable = New Car Sales Volume)................................................................ 139 5.6 MANOVA Results of Two-Way Interaction Model, Adjusting for Low Compliance (Dependent Variable = New Car Sales Volume)................................................................ 139 5.7 MANOVA Results of Two-Way Interaction Model, Adjusting for High Conflict (Dependent Variable = New Car Sales Volume)................................................................ 140 5.8 MANOVA Results of Two-Way Interaction Model, Adjusting for Low Conflict (Dependent Variable = New Car Sales Volume)................................................................ 140 5.9 MANOVA Results of Two-Way Interaction Model, Adjusting for Low Involvement (Dependent Variable = New Car Sales Volume)................................................................ 141 5.10 MANOVA Results of Two-Way Interaction Model, Adjusting for High Involvement (Dependent Variable = New Car Sales Volume)................................................................ 142 5.11 MANOVA Results of Main Effect Model Given High Compliance-Low Involvement (Dependent Variable = New Car Sales Volume) ........................................... 143 ix Page 5.12 MANOVA Results of Main Effect Model Given High Involvement-High Compliance (Dependent Variable = New Car Sales Volume) ............................................ 144 5.13 MANOVA Results of Main Effect Model Given High Involvement-Low Compliance (Dependent Variable - New Car Sales Volume) ............................................ 144 5.14 MANOVA Results of Main Effect Model Given Low Involvement-Low Compliance (Dependent Variable = New Car Sales Volume) ............................................ 145 5.15 MANOVA Results of Main Effect Model Given High Conflict-High Compliance (Dependent Variable - New Car Sales Volume) ............................................ 146 5.16 MANOVA Results of Main Effect Model Given High Conflict-Low Compliance (Dependent Variable = New Car Sales Volume) ............................................ 146 5.17 MANOVA Results of Main Effect Model Given Low Conflict-High Compliance (Dependent Variable = New Car Sales Volume) ............................................ 147 5.18 MANOVA Results of Main Effect Model Given Low Conflict-Low Compliance (Dependent Variable « New Car Sales Volume) ............................................ 147 5.19 MANOVA Results of Main Effect Model Given High Conflict-High Involvement (Dependent Variable = New Car Sales Volume) ............................................ 148 5.20 MANOVA Results of Main Effect Model Given High Conflict-Low Involvement (Dependent Variable = New Car Sales Volume) ............................................ 148 5.21 MANOVA Results of Main Effect Model Given Low Conflict-High Involvement (Dependent Variable = New Car Sales Volume) ............................................ 149 5.22 MANOVA Results of Main Effect Model Given Low Conflict-Low Involvement (Dependent Variable = New Car Sales Volume) ............................................ 149 5.23 Magnitude of Effect on Performance (Full Models) .... 152 5.24 Magnitude of Effect on NCSV (Simple, Simple Models) . . 154 x Page 5.25 Magnitude of Effects on NCSV (Main Effect Models) ... 156 5.26 MANOCOVA Results in the Full Effect Models.............................158 xi LIST OF FIGURES Figure Page 2.1 A Modified (Behavioral) Model of Channel Member Conflict and Performance ................................ 34 2.2 A Classical Model of the Conflict-Performance Assumption........................................................... 40 3.1 A 2 x 2 x 2 Design of Conflict, Involvement, and Compliance........................................................... 62 3.2 Conflict Situations for Levels of Conflict and Involvement........................................................... 65 4.1 Conceptual Categorization of Conflict Issues Across Levels of Conflict and Involvement . 97 xii CHAPTER I INTRODUCTION Research Approaches Historically, the literature on the development of marketing theory in terms of channels of distribution is characterized by three approaches. They include the institutional (structural) approach, the commodity approach, and the functional approach. More recently, the channel literature has characterized channels of distribution as broader social systems which encompass the institutional and functional approaches. This is referred to as a behavioral approach. The Structural Approach Channel institutionalists advocate the design of distribution networks across industry types (Bucklin, 1966). The institutional approach deals with "the set of institutions, agencies, and establishments through which the product must move" (Stern & El- Ansary, 1982, p. 3). For example, the marketing channel structure for Proctor and Gamble (P&G), a major marketer of personal care items, is characterized by its own internal marketing organization (e.g., sales managers and representatives) and independent external organizations (e.g., distributors and retailers). The nature of the 1 2 relationship between P&G, its distributors, and retailers is manifested by a administered vertical marketing system arrangement. Thus, P&G’s marketing efforts depend to a large extent upon coordination of the efforts of the internal and external marketing organizations that comprise its channel structures. The institutional approach to marketing channels emphasizes the types of distributors (wholesalers), facilitating agencies, and retailers who participate in different channel arrangements. Also, the evolution or change in institutions over time has added an enriched dimension to the study of channel institutions (Alderson, 1957; Stern & El-Ansary 1982). For example, consider in retailing, the birth of chain supermarkets (e.g., Kroger, A&P) to compete with independents or the rise of mass merchandisers (e.g., K-Marts, Walmarts, Zayre) to compete with traditional department stores. The Functional Approach Many channel analysts have adopted the functionalist approach to the study of marketing channels of distribution (Alderson, 1957, 1965). The functional approach delineates the distinct, yet essential marketing operations in the creation of time, place, and possession utilities (Stern & Gorman, 1969). Channel "functions" or "flows" are used interchangeably and are synonymous. There are eight universal flows or functions: physical possession, ownership or title, promotion, negotiation, financing, risking, ordering, and payment (Vaile, Grether, & Cox, 1952, p. 113). 3 Alderson (1957) advocated that the functionalist approach to marketing theory begin with the study of organized behavior systems. Thus, marketing functions are performed by individuals acting within systems--behavior systems. For example, General Motors discharges physical possession, ownership, and promotion of its automobiles to retail dealers and end users. It also engages in negotiation, financing, and risking with dealers to facilitate other flows. Auto dealers, in addition to performing or sharing in performing channel functions, are expected to attract customers, persuade them to buy cars, and provide after-sale service. The Commodity Approach The commodity approach underscores the activities involved in marketing specific classes of goods (Stern & Brown, 1969). Those who have adopted this approach feel that there is a clear distinction between the marketing of consumer goods (e.g., toothpaste, detergent) on the one hand and the marketing of industrial goods (e.g., steel) on the other. Furthermore, adopters of this approach contend that there are differences in the marketing and distribution activities for different types of consumer goods (e.g., shopping versus specialty goods) and different types of industrial goods--e.g., raw materials versus operating supplies (Alderson, 1957). The Behavioral Approach The behavioral approach to the study of marketing channels of distribution was spearheaded by Alderson’s (1957, 1965) work on 4 organized behavior systems and the functionalist approach. From an empirical perspective, channel analysts have relied upon organizational behavior and sociopsychological theories to study the dynamics of channel interaction among exchange partners. The social systems or behavioral approach to channel theory and research details important social variables (e.g., power, control, conflict, cooperation, coordination, satisfaction, position, and role rela­ tionships) . This approach is a much broader one for the study of channel behavior than the myopic economic systems’ perspective. It reveals an interaction between structure and function which requires the simultaneous study of both. Consequently, the behavioral dimensions add depth to understanding channel performance, and they permit participating channel members to exercise control over the types of interactions that may strengthen and/or weaken channel competitiveness. McCammon and Little (1965) concluded that channels of distribution are elaborate economic, political, and social systems that usually involve many decision makers and often extend over a wide geographical area. A sophisticated understanding of this phenomena requires an eclectic approach--an integration of concepts from a variety of disciplines, including sociology, economics, political science, cultural anthropology, regional science, marketing, and social psychology, (p. 322) To illustrate the usefulness of this broader approach, Stern and Brown (1969) describe the success of wholesaler-sponsored (voluntary) cooperatives relative to retailer-organized cooperatives. A sociological perspective allows an examination of 5 social power among participants in the two systems, where the locus of power resides, and the use of power to resolve conflict. The success of both types of cooperatives requires more than an economic perspective. The economic perspective is limited because primary focus is placed on efficient performance of marketing functions. Additional understanding can be gained from the study of behavioral dimensions. Critical Components of the Behavioral Approach The behavioral dimensions in channels of distribution are selected for conceptual and empirical investigation when an "eclectic approach" is used to understand the observed behavior of channel members, beyond that supplied by economic theory (Stern & Brown, 1969). The relevant dimensions of channel behavior, e.g., roles, conflict, power, leadership, cooperation, control, coordination, satisfaction, and communication, form the basic concepts for developing and testing channel behavior theories. The theoretical foundations of the behavioral approach consist of hypothesized relationships between (among) these relevant dimensions of channel member behavior. Channel analysts are empirically investigating the hypothesized relationships between these behavioral dimensions (1) to enrich development of channel theory, (2) to provide a base for determining the probability that a channel will achieve certain outcomes (e.g., performance), and (3) to help managers better understand the behavior of firms on other channel levels and to design strategies to achieve individual and 6 interorganizational goals (Alderson, 1965; Beier & Stern, 1969; Bonoma, 1976; Brown, 1979; Brown & Frazier, 1978; Brown, Lusch, & Koenig 1984; Brown, Lusch, & Muehling, 1983; Butaney & Wortzell, 1988; Churchill, 1979; Cron & Levy, 1984; Dickson, 1983; El-Ansary, 1979; El-Ansary & Stern, 1972; Eliashberg & Michie, 1984; Etgar, 1976a, 1976b, 1978; Frazier, 1983a, 1983b; Frazier, Gill, & Kale, 1989; Frazier & Summers, 1986; Gaski, 1984, 1986; Heskett, Stern, & Beier, 1974; Hunt & Nevin, 1974; John, 1984; Lederhaus, 1984; Lusch, 1976b, 1977, 1978a, 1978b; Lusch & Kasulis, 1976; Lusch & Ross, 1985; Mallen, 1963, 1964; Michie & Sibley, 1979; Raven & Kruglanski, 1970; Ridgeway, 1957; Robicheaux & El-Ansary, 1976-77, 1977; Rosenberg & Stern, 1970, 1971; Ruekert & Churchill 1984; Schul, Pride, & Little, 1983; Shugan, 1985; Simon, 1953; Speh & Bonfield, 1978; Stern & Reve, 1980; Stern, Schultz, & Grabaer, 1973-74; Walker, 1970; Walker & Shooshtari, 1982; Warshaw, Crawford, & Tank, 1985; Wilkinson, 1973, 1974, 1979). The one area that has received the least evaluation has been the relationship between conflict and performance. Channel member performance and conflict have been virtually ignored except for a few studies (Brown, 1978, 1979; Lusch, 1976a; Pearson, 1973; Pearson & Monky, 1976; Rosenbloom, 1973). Channel performance is the outcome or output of the collective efforts of the institutions in the channel system. Channel performance is often measured "objectively," e.g., asset turnover, return on assets, customer service level, inventory turnover, cost, sales, profitability, return on investment, sales per square/cubic 7 foot of selling area, number of stockants, etc. (Lusch, 1976a; Pearson, 1973; Sibley & Michie, 1981). Yet objective measures of performance are seldom tied to other behavioral constructs useful for a genuine understanding of channel behavior. It is important that performance measures used be associated with the channel relation being investigated, given the multitude of factors which affect performance. Most studies have not done this because of the difficulties associated with developing a meaningful behavioral index of performance (Brown, 1980; Lambert & Cook, 1979). One exception is the study by Sibley and Michie (1981) on power and performance. They developed a behavioral index of performance and related it to sources of power. Channel conflict is a state of frustration brought about by a restriction of role performance (Stern & Gorman, 1969, p. 156). It develops as a result of mutual interdependencies among channel members. Performance in distribution channels can be evaluated in terms of system effectiveness, system equity in serving various markets, system productivity (inputs-to-outputs ratios), and system profitability-financial efficiency (Stern & El-Ansary, 1982, pp. 462-63). It is frequently noted that the literature lacks a quantitative measure of performance for aggregate channel performance. Yet, measures do exist to evaluate individual channel member performance. 8 The Importance of the Conflict-Performance Relation From a Theoretical Perspective The Research Problem This research study attempted to examine conflict and performance from a behavioral perspective. The primary focus was on conflict resolution behaviors by channel members to resolve conflict and the impact of those behaviors on performance. Given that channel members interact because of mutual interdependencies, conflict--constructive or destructive--is inevi­ table. It is necessary that conflict and performance be examined from a behavioral perspective in order to gain a better understand­ ing of their association. This may provide valuable insights for theoretical development and empirical investigation in the area of channels of distribution. Previous Findings There is general consensus among channel analysts that the conflict process affects performance. However, there is widespread dissent concerning the functional form of the relationship between the two constructs. Theoretically, it has been hypothesized that the relationship can be positive, negative, or curvilinear (Alderson, 1965; Little, 1968; Rosenbloom, 1973). The dissensus over this issue seems to stem from not having a paradigm that systematically identifies when the relationship is of a particular form. Also, there is no systematic way to determine under what conditions a particular functional relationship will be manifested. Empirical investigations (Brown, 1978; Lusch, 1976a, 1976b; Pearson, 9 1973; Reve, 1977; Walker, 1970) of the classical models of conflict and performance found unsettling results. These studies found statistically insignificant or weak significance relationships. Furthermore, the results of empirical evidence to date reveal that the relationship is not generalizable across industries or dyadic arrangements within different channels in the same industry. Consequently, future research efforts must take into account the notion that the relationship between conflict and performance seems to be situation specific. As suggested by Brown (1978), specific issues of conflict and, more importantly, specific dimensions of performance may prove to be the key to finding significant--statistical and operational--results. In addition, it is crucial that the operational definition or measurement of both constructs incorporate the "behavioral" dimension to gain insights into their true relationship(s). Also, this approach may be useful in determining whether conflict and performance have a direct or indirect relationship. Channels of distribution are essentially collaborative exchange relationships which occur through business activities. The theoretical significance of this research is closely connected to managerial practices in general. As a result, the conf1ict-performance assumption has valuable ramifications from a managerial perspective. 10 The Importance of the Conflict-Performance Relation From a Managerial Perspective Managerial Problem Across most industries, service and nonservice, to effectively compete means addressing the issues of increased global competition, rapidly changing environmental conditions, and advancing technology. Domestically, deregulation, disinflation, price wars, and escalating costs have all had an impact on management’s competitive strategies. Specifically, two trends are emerging as a result of the dynamics of market competitiveness. First, there are increased efforts by management to engage in long-term planning as a catalyst to survival and growth. Second, there is increased focus on "productivity" and efficiency in light of the intensity and growth of worldwide competition. For example, auto and steel producers, threatened by foreign competition, are discarding weak product lines, trimming work forces, modernizing plants, and clamping tight controls on material costs and inventories (Fortune. 1986, p. 21). As a result of more thorough reviews of cost efficiencies, the "make or buy" issue is being examined or reexamined. Collaboration as a Mode of Managing Long-Term Channel Exchange Relationships With a renewed interest in market competitiveness through increased productivity and efficiency, channel members are adopting strategies intended to develop "longer and stronger" relationships with their exchange partners. Thus, firms cannot overlook the 11 importance of "collaboration" or a lack thereof to sustain survival and growth (Rosenberg & West, 1984). In terms of strategic planning, distribution decisions are considered strategic ones which have long-range implications. Conceptually, Salmond and Spekman (1986) argue that when fully developed, collaboration is: "a bilateral mode of managing exchange in which exchange partners adopt a high level of purposeful cooperation to maintain the trading relationship over time" (p. 162). As trading or exchange relationships develop over time, they may be managed in a more or less collaborative manner. In order to have trading relations which are managed collaboratively, conceptual analysis and managerial practice suggest that open communication, conflict resolution, coordination of work, and long-term planning for the future are the key ingredients (Salmond & Spekman, 1986). Collaborative exchange relationships have a future orientation. MacNeil (1980) argues that mechanisms for projecting future exchange or trading have evolved in all societies. He emphasizes that repeated exchange between trade partners cannot adequately be described as a bundle of discrete transactions. The very complexity of modern technology calls for processes and structures tying even the most specific and measured exchanges into patterns of ongoing relations. Furthermore, MacNeil (1980) argues that exchanges between partners can be seen as a generalized set of transactions that must be considered in light of past relationships and anticipated future transactions. Collaboration is important because of, first, its impact on productivity and efficiency, and, second, its impact on changes in distribution channels in general (Adkins & Diller, 1983; Porter, 1986). Given the global competition in the auto industry, channel members must establish collaborative exchange relationships such that inter-firm communication is open, conflicts are resolved in a problem-solving mode, work affecting both parties is coordinated, and the exchange parties take each other into account as they plan future exchange (Salmond & Spekman, 1986). The survival and growth of trading parties in channels of distribution in general must develop and maintain longer-term, more cooperative relationships. Inherent Problems in Collaborative Exchange Relationships Inherent in collaborative exchange is interdependence. Since interdependence is the root of exchange, it provides opportunities for cooperation, competition, and constructive and destructive conflict (Ross & Lusch, 1982; Stern & Gorman, 1969). In channel relationships, collaboration offers the potential for subsequent conflict which can create problems with respect to (1) self-interest being served and (2) unclear definition of roles--role incongruity. Stern and Gorman (1969) advocate that interdependence is the root of conflict since each party has individual goals and preferred methods of achieving them. Thus, when incompatible activities occur such that an action by one party prevents, obstructs, interferes 13 with, injures, or in some way makes less effective the actions of the other party, the parties are in a state of conflict. Another problem that arises as a result of collaborative exchange is the unclear role definition between channel members. In the distribution channel, suppliers have to relate to channel intermediaries as customers, employees, competitors, and partners. Thus, there is the opportunity for incongruent role specification, role expectation, and role performance. Consequently, suppliers have to determine how they will relate to intermediaries in order to promote collaborative channel activity which strengthens its (channel) long-term competitiveness. As customers, channel intermediaries are the targets of supplier marketing efforts covering a multitude of products. The objective of the supplier’s marketing efforts is to establish a working relation with intermediaries to effectively build patronage at the retail level such that customer satisfaction is established with the supplier. The joint objective of the suppliers and intermediaries is to create and maintain product/brand loyalty and store patronage. This can be achieved more effectively and efficiently by communicating, coordinating work, cooperating, and planning with the other partner in mind. However, there are times when destructive conflict will develop, which adversely affects the performance of all the parties in the exchange relationship. Consequently, channel members must use problem solving to resolve the destructive effects of conflict on the competitiveness of the channel. 14 As employees, channel intermediaries are employed by suppliers to create time, place, and possession utilities for their products/brands among ultimate consumers. These utilities are created through an effective and competitive marketing program directed toward mass consumer groups or distinct market segments. Therefore, the supplier must provide the kinds of benefits (trade promotions, trade advertising, consumer promotion and advertising, financing, payment terms, etc.) which will retain intermediaries as trading partners in an attempt to foster channel competitiveness in the long run. However, as suppliers relate to intermediaries as employees, the benefit package offered may not always be acceptable. This can create the opportunity for various levels of conflict which can be harmful to channel efficiency, growth, and competitiveness. Given that suppliers create form utility and intermediaries create time, place, and possession utilities, it is imperative that they collaborate over time to recognize conflict and resolve it and thereby eliminate potential destructive consequences. The most effective, efficient, and competitive channels are those which provide the best bundle of market utilities to the consumer. Suppliers relate to channel intermediaries as competitors in the sense that the goals of the parties can be in opposition; that is, exchange has some zero-sum aspects and one party maximizes its gains at the expense of the other (Salmond & Spekman, 1986, p. 163). For example, the supplier’s goal of obtaining a high price is competitively linked to the intermediary’s cost-minimization goal. 15 However, if channel members are treated as bilateral exchange parties, competition can be converted into cooperation and constructive conflict rather that destructive conflict. In an unbalanced power relationship in favor of either the supplier or intermediary, there is the tendency to relate to trade partners as subordinates. This type of unilateral mode exhibits a fairly structured, almost hierarchical relationship, in which one party will repeatedly consent to the power attempts of the other. The weaker party is co-opted, and cooperation depends upon the good will of the dominant party for the benefits of the exchange. Consequently, cooperation is one-sided. The bilateral nature of collaboration necessitates that each party to an exchange cooperate with the other, yielding to the needs of the other while at the same time asserting its own needs and goals (Salmond & Spekman, 1986, p. 163). Contributions of the Research Given that this research has both theoretical and managerial significance, the objectives or contributions are worthy of explanation. Therefore, analysis of the research objectives is presented with an emphasis on the scope of the research. Interaction between channel participants leads to conflict (functional or dysfunctional), which in turn affects their performance. Admittedly, the bulk of interaction in the channel leads to cooperation (Bowersox, Lambert, Taylor, & Cooper, 1982). However, cooperative interaction, on occasion, is overtaken by 16 heightened states of conflict. Consequently, it is these occasional episodes of conflict, constructive and/or destructive, that are most worthy of investigation. In light of existing empirical findings on the relationship between conflict and performance, the present study offers two major contributions. First, a theoretical framework is provided for the conflict-performance relationship, and second, a quasi-experimental design is developed which may explain the disparity in the empirical findings on the nature and form of the relationship between conflict and performance. The contribution of the theoretical framework centers on the managerial implications of the variety and severity of conflict resolution behaviors that are unique to a particular industry. Each conflict situation faced by exchange partners offers a different range of resolution behaviors which vary in severity in terms of their impact on performance. It seems plausible that the more options channel members have to resolve their conflicts, the less impact (constructive or destructive) they will have on performance. On the other hand, the fewer the number of resolution options available, the greater the impact each conceivably may have on performance. Thus, the nature and breadth of a channel member’s conflict resolution behaviors vary across conflict situations which arise that are unique to specific channels and industries. Channel analysts, theoretically and empirically, have advocated the importance of channel control exercised by the channel captain (Bucklin, 1968, 1973; El-Ansary & Robicheaux, 1974; Speh & Bonfield, 17 1978; Stern, 1970). The dominant channel member usually tries to exercise control over the resolution responses that are acted upon by other channel members. For example, in the franchise channel of distribution for automobiles, the manufacturer generally has the locus of power and control over dealers. According to Bucklin (1973), manufacturers’ incentives to exercise control stem from three sources: intrasystemic competition, inadequately trained middlemen, and the coordination of heterogeneous decisions (p. 40). Thus, unless manufacturers can exercise control over middlemen, they may make distinct decisions in order to maximize their individual profits. Yet, the dominant channel member must show others how to improve their operations in order to establish a solid base for control. Given the severity of alternative conflict resolution behaviors, it seems paramount that the manufacturer exercise control over the functions and actions of middlemen to resolve conflict. Where middlemen are used in the distribution channel, it also includes the design of control procedures to insure compliance with the desired marketing mix (Bucklin, 1973, p. 39). In order to capture an appreciation for the objectives of this study, a review of the literature on conflict and performance should be examined. Furthermore, the conceptual and empirical perspectives from channel analysts on the relationship between the two constructs should substantiate the appropriateness of the objectives/ contributions. 18 Plan of the Dissertation This research investigated different types of conflict situations and the effects of new car dealers’ conflict resolution behaviors on dealer performance. Chapter II provides a review of the background literature for this study. It begins with a review of the theoretical overview of a marketing channel of distribution. Then, the conceptualization of channel conflict is reviewed as a social process and in terms of its causes and effects from both an interorganizational and a sociopsychological perspective. Next, the performance construct is reviewed. Next, the literature review deals with the conceptual development and the empirical investigations of the relationships between channel conflict and channel performance. This review of the literature is the basis for establishing a revised model of conflict and performance and concomitant hypotheses at the end of the chapter. Chapter III contains a description of the methodology used to test the hypotheses. Chapter IV contains the results from the first phase of data analyses. Chapter V contains the results from the second phase of data analyses and Chapter VI follows with discussions, conclusions, limitations, theoretical contributions and implications, and future research directions. CHAPTER II LITERATURE REVIEW In this chapter, six major issues are discussed. First, the interorganizational framework of distribution channels is discussed. Second, channel conflict is analyzed in terms of (1) the nature of conflict, (2) issues which lead to conflict between channel members, (3) the causes of conflict, (4) the effects of conflict, and (5) the conflict process. Third, channel performance is examined in terms of the empirical and theoretical treatment to date. Fourth, a theoretical and an empirical exposition is examined on the relationship between channel conflict and channel performance. Also, a discussion follows on the "classical" models, which depicts the functional forms of the relationship. Fifth, a proposed "behavioral" model of the channel conflict-performance assumption is discussed. Finally, a proposed theoretical framework of conflict and performance is discussed. The Channel of Distribution There are a plethora of definitions for marketing channels of distribution. Conventionally, definitions of a distribution channel focused on the functions (functionalism) and institutions (institutionalism) used to move a product and/or its title from the 20 point of productions to the point of consumption (Alderson, 1957). Given that the channel is unobservable, a series of institutions at each level in a vertical arrangement perform marketing functions. In the early 1950s, channels of distribution took on a broader prospective. It was felt that to view a channel strictly in terms of functions, institutions, and commodities was too narrow and devoid of inherent complexities and dynamics of vertically aligned firms. McCammon and Little (1965) promulgated that a broader approach offered by Vaile, Grether, and Cox (1952) should be adopted. They depicted a channel as a combination and sequence of agencies through which one or more of the marketing flows (physical possession, ownership, promotion, negotiation, financing, risking, ordering, payment) moved. In the 1960s, it was recognized that this conceptual framework for analytical treatment of vertical channel arrangements was insufficient. McCammon and Little (1965) suggested that the analysis should delve into the "vertical relationships" that exist between/among the firms therein as well as the marketing flows. Specifically, they viewed a channel as being an operating system with the following attributes: --Interrelated components --Striving to achieve mutually acceptable goals --Sequencing activities and flows --Participating voluntarily --Usually administered by one firm --Regulated by a code of behavior or group norms This was the period in which the analysis of distribution channels took the social system’s approach. For example, Alderson 21 (1957) labeled channels as "organized behavior systems." Marketing channels were viewed as interorganizational systems (Fisk, 1961; Stern & Brown, 1969; Vaile, Grether, & Cox, 1952). Given this viewpoint, Stern and El-Ansary (1977) offered a more contemporary view of channels of distribution: "Marketing channels can be viewed as sets of interdependent organizations involved in the process of making a produce or service available for use or consumption" (p. 5). The mutual interdependence among firms in a channel gives rise to the importance of the behavioral dimensions (control, power, conflict, cooperation, satisfaction, coordination, roles), their interrelationships, and their effect on channel member performance as well as total channel performance. Marketing channels comprise a colossal division of labor among the institutions and agencies comprising them (Bucklin, 1965; Stigler, 1951). This division of labor represents a partitioning of the marketing functions performed with increased specialization as these functions are performed more efficiently and effectively by different institutions and agencies. Thus, specialization and functional differentiation lead to interdependency among channel participants relative to the tasks performed. Consequently, these interdependencies are considered as the foundation of the behavioral dimensions in the channel literature (Alderson, 1965; McCammon & Little, 1965; Stern, 1969; Stern & El-Ansary, 1977). 22 Reve and Stern (1979) emphasized that the increasing focus on the interdependencies among channel members has developed for the following reasons: 1. interdependent relationships are sufficient conditions to define a system, 2. varying degrees of interdependencies among channel members give rise to power relations between the groups comprising the channel, and 3. a dependent relationship gives seeds of conflict (p. 407). Some channel analysts have evaluated distribution channels as political economies which are a combination of social and economic systems (Achrol, Reve, & Stern, 1983; Anderson & Narus, 1984; Arndt, 1983; Frazier, 1983a, 1983b; Stern & Reve, 1980). This framework for comparative analysis permits integration of both the economic and sociopolitical approaches as determinants of channel member behavior. In addition, this approach reveals that the eco­ nomics of performing marketing functions is the basis of channel member interaction. Mutual interdependency is an outgrowth of func­ tional specialization. As a result, the behavioral or social dimen­ sions (power, conflict, cooperation, etc.) of channel member interactions are inherent in the economic performance of marketing functions. Furthermore, these interactions must be considered in the context of the internal and external environments of channel systems. The onset of recognized interdependency in distribution channels has led to an increased level of vertical integrations or group membership in more structured-type organizations. Thus, channel analysts recognized that the design and management of the 23 channel for both title and product flows contribute to the long-run success, survival, and growth of channel members. Consequently, increased attention was given to vertical marketing systems (VMS). McCammon (1970) contrasted the nature of the VMS to the conventional channels over a number of salient attributes. The differences that favor VMSs include: **greater coordination of marketing efforts, **a greater capacity to take advantage of economies of scale, **reliance on fewer channel strategists but no more support and operating staff, **emphasis on the total channel system’s performance more than on the member’s performance, **reliance more on specialists and scientific decision making, and **more open attitudes toward change in the system. Given these advantages, the conventional channels stand little chance to compete with VMSs. Furthermore, research results reveal that in vertical marketing systems, more coordination and efficiencies occur when compared to the conventional channels (Etgar, 1976). The three forms of vertical marketing systems include corporate, administrated, and contractual systems (McCammon, 1965). Channel coordination is achieved through three means--ownership, leadership, and legal contract, respectively, in each of the abovementioned systems. Because channel systems compete, the vertical marketing system has the essential tools to achieve and maintain a competitive advantage over conventional channels. 24 In the channel literature, competition is viewed as rivalry among systems of distribution--intertype competition (Palamountain, 1955). In addition, competition is viewed as rivalry among firms or organizations on the same level in the channel. Vertical marketing systems minimize intrachannel competition among participants so as to maximize overall channel performance. Consequently, the task for participants within any distribution network is to discover ways to cooperate in developing an interorganizational system that will minimize suboptimization. Therefore, the attention has been on the coordination of efforts among successive stages of distribution in a vertical arrangement (Mattsson, 1969; McCammon, 1970). Although organizations cooperate more frequently, it is inevitable that conflict will surface periodically. Conflict in Channels of Distribution Marketing channels represent a colossal division of labor among the institutions and agencies comprising them (Bucklin, 1965, 1966). This division of labor gives rise to specialization and functional differentiation. They, in turn, lead to mutual interdependence among channel members relative to their task performance. Consequently, these interdependencies among channel participants are considered central behavioral dimensions in the channel literature (Alderson, 1965; McCammon & Little, 1965; Stern, 1969; Stern & El- Ansary, 1977). Among the behavioral constructs in the channel literature is channel conflict. Definitions of conflict in a marketing channel 25 setting have been adapted from the social psychological and inter­ organizational behavior literature. Theoretically, Lusch (1976a, p. 383), Brown and Day (1981), who used the "latent-affective-manifest" framework, Rosenberg (1974, p. 9), and Firat, Tybout, and Stern (1975, p. 385) defined channel conflict similar to that of Stern and El-Ansary (1977): channel conflict is a situation in which one channel member perceives another channel member to be engaged in behavior that is preventing or impeding him from achieving his goals, (p. 283) According to Stern and Gorman (1969, p. 156) and Etgar (1979), channel conflict is present: . . . when a component (channel member) perceives the behavior of another component to be impeding the attainment of its goals or effective performance of its instrumental behavior patterns, (pp. 61-62) Thus, it is important to recognize that the definitions of conflict have been broadened to be "perceptually" and "overtly" based. Mutual interdependencies among organizations in distribution channels create conflict of interest. Conflict can and does arise in channels comprised of independently owned institutions and agencies--conventional channels--as well as in vertically integrated systems. One could argue that if interdependency among channel participants is "inherent" in the channel network, it is also arguable that conflict is an "inherent" attribute because dyadic dependency breeds the seeds of conflict. The connection between interdependency and conflict has been documented in the social literature and the channel literature (Assael, 1969; Cadotte & Stern, 1979; Dutton & Walton, 1966; Fink, 1968; Firat, Tybout, & 26 Stern, 1975; Kochran, Huber, & Cummings, 1975; March & Simon, 1958; Reve, 1978; Rosenberg & Stern, 1971; Schmidt & Kochran, 1972). Thus, the greater the level of interdependence, the greater the opportunity for interference with goal attainment (Schmidt & Kochran, 1972, pp. 361-63), and, hence, the greater the potential for conflict among vertically aligned organizations. Conflict can take place within the channel over a number of issues. Examples include: (1) how much inventory should be carried by various participants, (2) suppliers bypassing distributors and dealers via direct selling, (3) whether prices and services are being maintained at "adequate" levels, and (4) the right of a given intermediary to represent a particular product within a given territory (Reve & Stern 1979). Causes of Conflict A number of causes of conflict have been advanced by channel analysts. Little (1968) advocated such causes as misunderstood communications, divergent functional specialization, goals of the channel member, and failure of joint decision-making processes. Assael (1968) suggests ideological differences among channel members and differing economic objectives. The most comprehensive list of causes of conflict in marketing channels is offered by Stern and Gorman (1969). They identify the following as the fundamental underlying causes of channel conflict: (1) role incongruities, (2) resource scarcities, (3) perceptual differences, (4) decision domain disagreements, (5) goal incompatibilities, and (6) communication difficulties (pp. 57-61). 27 The standard list of causes of channel conflict, which has been supported in the channel literature, includes goal incompatibility, domain dissensus, and differences in the perception of reality (Cadotte & Stern, 1979; Etgar, 1979; McCammon & Little, 1965; Reve & Stern, 1979; Rosenberg & Stern, 1971; Stern & Heskett, 1969; Wittreich, 1962). However, Rosenberg and Stern (1971) found no empirical evidence to support the abovementioned causes as the fundamental grouping. Effects of Conflict Conventionally, conflict in the channel has been thought to be dysfunctional (Mattsson, 1969). It is often defined as behavior designed to destroy, injure, thwart, or control another member in an interdependent relationship (Mack & Synder, 1957). Thompson (1960) advocates that dysfunctional conflict should be avoided and eliminated. However, it is evident that conflict is frequent, yet highly functional (Assael, 1969; Coser, 1965, 1967; Litterer, 1966; Rosenberg & Stern, 1971; Rosenbloom, 1973; Stern & El-Ansary, 1977; Walton, Dutton, & Cafferty, 1969). Without conflict, channel systems are likely to become passive, noninnovative, and eventually nonviable (Reve & Stern, 1979). Nevertheless, conflict is direct, personal, and opponent-centered behavior as opposed to object­ centered behavior (competition) or joint-striving--cooperation (Fink, 1969; Mack & Synder, 1957; Stern, 1971a). Conflict created by functional interdependencies and the scarcity of resources can have positive consequences. However, 28 pathological conflict--moves that malign parties involved and the entire system--can be harmful (Boulding, 1965; Deutsch, 1973; Thomas, 1976). There has been a proliferation of studies in the channel literature which focus on three areas of conflict. They are: 1. Conflict in distributions (Alderson, 1965; Assael, 1968; Carlsson & Kusoffsky, 1966; Mallen, 1963; Palamountain, 1955; Stern & Gorman, 1969); 2. The sources and consequences of interorganizational con­ flict in marketing channels (Brown, 1978; Brown & Day, 1981; Etgar, 1979; Firat, Tybout, & Stern, 1975; Lusch 1976b; Lusch, Brown, & Muehling, 1983; Pearson, 1973; Pearson & Monoky, 1976; Reve, 1977; Ridgeway, 1957; Rosenberg & Stern, 1971; Walker, 1972); and 3. Management of conflict in vertical marketing systems (Anglemar & Stern, 1978; Assael, 1968; Hunger & Stern, 1976; Rosenberg, 1974; Stern, 1971b; Stern & Heskett, 1969; Stern, Sternthal, & Craig 1973a, 1973b, 1975; Thomas, 1976 Weigand & Wasson, 1974). The functional significance of conflict means that it has a positive impact on channel efficiency, while dysfunctional conflict has a negative impact on channel efficiency (Rosenbloom, 1973). Thus, on the conceptual level, these relationships seem to fit some part of real world behaviors, namely "interactions" in the channel network. However, there is a compelling notion that the nature of a conflict situation leads channel intermediaries to engage in response (reaction) behaviors which have either a dysfunctional (negative) impact, a functional (positive) impact, or no impact on channel efficiency (performance). Thus, to gain better insights into the conceptual issues concerning the impact of conflict on channel performance, some key research issues need further investigation: 29 1. The existence of a relationship between channel conflict and channel efficiency; 2. The direction and strength of the relationship between the two constructs; 3. The impact of environmental/situational factors on the relationship between the two constructs; 4. What issues of conflict and dimensions of performance are unique (specific) to industries and channel types within an industry; and 5. What intervening variable(s) exist which may camouflage the relationship between conflict and efficiency? Conflict as a Process Interorganizational dependency is the microscope through which conflict may be viewed as a "process"; that is, an organization proceeds through sequential stages of conflict--cognitive or latent conflict, affective or felt conflict, overt or manifest conflict, and conflict aftermath (Pondy, 1967). This perspective has been adopted by channel analysts in the study of channel conflict (Rosen­ berg & Stern, 1971; Stern & Gorman, 1969; Stern & Haskett, 1969). Cognitive and affective states of conflict generally precede the taking of overt opponent-centered actions by either party in a conflict situation. This means that organizations or channel members involved must first become aware of the conflict situation as well as personalize it so that hostile feelings develop (Pondy, 1967). Performance in the Distribution Channel Another channel construct that has been related theoretically and empirically to channel conflict is channel performance (Brown, 30 1978, 1980; Etgar, 1976a; Lusch, 1976a, 1978; Pearson, 1973; Reve, 1977; Rosenberg & Stern, 1971; Rosenbloom, 1973; Walker, 1970). These studies related financial measures of performance to perceptual conflict only to encounter disappointing results. However, Brown (1980) emphasized that future research on the conflict-performance assumption should develop the performance construct as behaviorally or perceptually based. Marketing channel performance, the output dimension, is accepted among marketing scholars and practitioners as the ultimate objective and the outcome of channel interaction between two or more firms (Robicheaux & El-Ansary, 1975-76, 1977). There are various levels of analysis at which performance evaluation can take place; that is, (1) on the management of the channel system, (2) the marketing functions or flows, and (3) the participating institutions (Bucklin, 1972; Stern & El-Ansary, 1977). The channel literature reveals that most measurements of performance have been quantitative. Quantitative measures of performance applied to all three levels include costs, sales, profitability, return on equity, return on investment, market share, input-output analysis, value added, and the strategic profit model (Brown, 1979; Rosenbloom, 1979; Sibley & Michie, 1981, p. 60). In spite of proponents of each quantitative measure of channel performance, sales analysis has been used most frequently, given the increasing popularity of the strategic profit model--assessment of the efficient use of financial and physical resources by the channel 31 or the firm. At the institutional level, other measures of performance used regularly are sales per square foot or cubic foot of selling area, number or percentage of stockouts, markdowns, merchandise, return on assets, asset turnover, customer service levels, and inventory turnover. Admittedly, these measures represent objective quantitative evaluations of channel performance or channel member performance that are determined by behavioral relationships or interactions. Furthermore, they are consistent with the conventional economic model. The theoretical basis for treating channel member performance as a behavioral construct has been most fully developed in the organizational behavior and social-psychological literature. Performance has been defined as "effectiveness" (McMahon & Perritt, 1973; Tannenbaum, 1968), "efficiency" (Rosenbloom, 1973), and "success" (Child, 1974; Woodward, 1968). Efficiency usually refers either to the inputs (e.g., costs) or to the input-output ratio (Rosenbloom, 1973). Effectiveness generally refers to the quality or the magnitude of the output (e.g., profits). Success generally captures both effectiveness and efficiency of performance. Nevertheless, Gaski and Nevin (1985) defined performance as "the degree to which a supplier’s relationship with a dealer contributes to fulfillment of the supplier’s objective" (p. 131). For the present research, channel member performance will be defined and measured using some perceptual measures developed in the channel literature and others which are appropriate for the channel relation investigated. 32 The Classical Models of the Conflict- Performance Assumption The conceptual foundation underlying the study of interorganizational relations in marketing channels is pervasive. It emphasizes that there are a number of relationships on the effects of conflict in marketing channels. For example, some channel analysts have postulated that channel conflict and channel performance are positively related. Other channel analysts concede that increased conflict leads to decreased performance; that is, a negative relationship. Furthermore, an overlapping group of channel analysts hypothesize that, over time, the relationship between the two constructs is both positive and negative; that is, the relation is curvilinear--a "threshold" effect. Thus, it is quite clear that the positions taken on the conflict-performance relation are not dichotomous as would be expected. In addition, the empirical investigations on the conflict-performance relation are as heterogeneous as the theoretical proclamations. Needless to say, dissensus on the true functional form of the relationship between the two constructs is widespread. It is hoped that an exposition of a "behavioral" model of the relationship between conflict and performance will break ground on establishing some consensus. But first, the "classical" models offered by channel analysts should be considered. Alderson (1965) postulated that the effects of conflict on channel performance are generally negative. Stern (1971c), on the other hand, takes a similar position but also points to possible positive effects as well. Dixon and Layton (1971) advocate that conflict can result in a threat to the survival of the channel. This suggests that conflict and performance can, at times, be negatively related. Also, Assael (1969) has emphasized the possible positive effects on conflict on various aspects of the channel. Figure 2.1 shows the classical model of the conflict-performance assumption and the position taken by various channel analysts concerning the form of the relationship. Rosenbloom (1973) postulates that from a managerial viewpoint, the key question that should be addressed is "how does conflict affect channel efficiency?" Channel efficiency is defined as "the degree to which the total investment in the various inputs necessary to achieve a given distribution objective can be optimized in terms of outputs" (p. 27). Rosenbloom’s position is that optimizing marketing inputs in carrying out the distribution objective will have an impact on efficiency. For example, when there is reduced efficiency with increasing levels of conflict, a negative effect will be observed. Increased efficiency with increased levels of conflict reveals a positive effect. Constant efficiency with increasing levels of conflict renders no impact. Finally, over a range of increased levels of conflict, channel efficiency may not change, increase, and decrease. This, of course, is the curvilinear effect. Imbedded in the curvilinear effect is the "threshold effect;" that is, the breakpoint between the maximum level of channel efficiency and 34 Figure 2.1: A m odified (behavioral) model of channel member c o n flic t and performance. 35 reduced efficiency with increased levels of conflict. Other channel analysts (Boulding, 1965; Lusch, 1976a; Rosenberg & Stern, 1971) have taken similar positions. Thus, the channel literature offers a plethora of conceptual models which capture the essence of the effects of channel conflict on channel efficiency. However, empirical examination of these conceptual models provides more insights on the validity of the relationships. There have been a number of attempts to test models of interorganizational conflict in marketing channel settings (Brown, 1978; Lusch, 1976a; Pearson, 1973; Walker, 1970). Walker (1970) conducted the first of these studies in a laboratory setting. He found that the impact of conflict on channel performance was highly influenced by the distribution of channel power. In situations where power was balanced, conflict resulted in nonoptimum joint profits and dissatisfaction on the part of weaker channel members. Pearson’s (1973) investigation involved a field study among a large regional grocery chain and 55 of its suppliers. The hypothesis tested was that the operational results (service level, inventory turnover, adjusted inventory turnover) associated with channels characterized by cooperation are superior to the operational results associated with channels characterized by conflict. The analysis of the data collected failed to support this hypothesis. Pearson cautioned that conflict and cooperation may not be directly related to operational performance, measured using customer service levels and inventory. 36 Lusch (1976a) tested three alternative models depicting three possible relationships between conflict and retailer operating performance. These models were empirically tested in a franchised channel of distribution for new automobile suppliers and dealers. It was hypothesized that there is a direct and inverse linear relationship between the franchisee’s operating performance and the franchisee’s conflict with the supplier. Furthermore, it was hypothesized that there is a threshold (curvilinear) relationship between the franchisee’s operating performance and the perceived conflict with the supplier. The sample of auto dealers was partitioned into five groups with respect to product/brand of automobile carried. Using return on assets and asset turnover as performance measures, performance was regressed on an index of perceived manifest conflict by dealers that was developed in another study (Lusch, 1976b). The results of the analysis of the data collected indicate that for two of the five dealer groups, a negative linear relationship between the two performance measures and perceived manifest conflict proved to be statistically significant. However, the positive and threshold models were statistically insignificant across all dealer groups. Pearson and Monoky (1976) examined the relationship between conflict, cooperation, and performance in the channel of distribution for grocery products. A multivariate measure for a conflict-cooperation continuum was utilized to analyze the perceptions of 54 channel dyads (grocery wholesaler/retailer or grocery buyer/vendor) regarding the nature of their relationship. 37 Fifteen conflict statements and 15 cooperation statements were used to anchor the conflict-cooperation continuum. Performance was operationally measured as service level--the nonmetric dependent variable. Using discriminant analysis to examine the relationship, the findings suggest that a relationship exists between performance and the channel member’s position on the conflict-cooperation continuum. Channels categorized as low performance channels tended to manifest the conflict dimensions, whereas high performance channels manifested the cooperative dimensions. Furthermore, across all situations, the cooperative dimensions were perceived by members of the dyads as being greater than the conflict dimensions. Thus, this study lends empirical support to the model depicting channel conflict and performance as negatively related; that is, increased conflict leads to reduced performance. In another study to examine the threshold model on the conflict-performance assumption in the automobile franchise channel of distribution, attempts were made to control for the effects of marketing mix variables and environmental factors on performance (Brown, 1977). Eight measure of performance were regressed on an overall index of perceived manifest conflict and indices of conflict over several issue groupings. The marketing mix and environmental variables were also included as exogenous variables. The analysis of the data revealed some support for the threshold relationship, the direct and inverse relationships, and an upright "U-shaped" relationship. In many instances, however, no significant relationships were unveiled. It was concluded that the way in which 38 conflict is related to performance depends upon the conflict issues examined, the performance dimension(s) used, and the particular channel setting. Summarizing the current state of knowledge on the conflict­ performance relation, Brown (1980) concluded that the relationship between channel conflict and performance is not "general" across all types of channel systems. Consequently, various research issues with respect to the conflict-performance assumption were offered, keeping in mind that specific conflict issues and specific performance dimensions should be used in any future empirical investigation. Inasmuch as the conflict-performance assumption is unsettled, future research on this relation is not replete. For all the model types that describe the relationship, there is a noticeable dissensus with respect to which model best describes the "true" functional form of the association between the two constructs in marketing channels. Consequently, any future efforts to investigate this relation should attempt to shed some light on the nature of conflict using the buyer-seller interaction framework. Furthermore, additional investigations should treat channel conflict as an external stimulus that elicits reactionary behaviors by channel intermediaries. The nature of this reaction will affect the channel participants’ operational performance as well as that of the entire channel system. 39 Theoretical premises and empirical investigations of the conflict-performance assumption have created suspicions that there may be other factors that intervene in this relationship. Furthermore, a different methodological approach to examine the relationship between the two constructs may provide a better understanding for channel theory development. A modified model of conflict and performance is introduced. It contains four variables--conflict, involvement, compliance, and performance--which will be the subject of inquiry in this research. The model, a "behavioral" approach, characterizes the relationship between conflict and performance as indirect. Also, it asserts that the interaction of the three major explanatory variables will have a more powerful impact on channel member performance than either one alone. Hypotheses will be specified for the explanatory variables and their interactions with a different methodological approach in mind. A Behavioral Model of Channel Member Conflict and Performance The extent of the channel literature on conflict and performance gives rise to a proposed revised model of the relationship between the two constructs. (See Figure 2.2.) The classical models of channel conflict and performance have been empirically investigated specifying the functional form of the relationship between the two constructs by channel analysts. Their findings have not confirmed the general nature of the relationship. The consensus among channel analysts is that the form of the 40 of the conflict-perform ance assumption. i— o LU LL. LI­ LLI Q z O oc T LU CO I— co POSITI o I— ^2° LLI model u_ Ll_ oc LI­ a?? LLI LU LU LU cla ssica l LU z 7LL|ZLLI —I ZQIZ CE OCOQ—1DC A Illi z । । । । । । o Illi Figure 2.2: 41 relationship is specific to the channel, the industry investigated, and the measures used for conflict and performance. The revised or "behavioral" model is an attempt to uncover factors that may interact with conflict and lead to stronger performance results. The basic premise of the modified model is that conflict and performance are indirectly related. The relation is mediated by intervening variables--the level of channel member involvement in implementing a decision response(s) and the level of channel member compliance with a decision response(s). It is conceivable that conflict can be created by two stimuli--actions taken by a supplier or actions requested of the intermediary by a supplier. A distinction should be made between the two stimuli because the psychological processes and behaviors will be different for each even though they both have the potential to create intrachannel conflict. The contractual arrangement in the franchised channel of distribution for new automobiles lends itself to the essential distinction between the supplier taking action or requesting action of the intermediaries. For example, within the limits of the contractual agreement between the auto supplier and its dealers, the auto supplier takes action such as offering cash rebates or discount financing for new cars to consumers. The dealers react by responding to the supplier’s action willingly or not. The dealers are compelled to respond in a manner (marketing program adjustment) that would generate sales and profits for their firm. Given the contractual agreement, the auto supplier’s option 42 set is limited, as is that of dealers. Dealers respond to the action taken by the suppliers by taking the necessary steps to choose from their set of response options that (those) which will reward their firm through increased sales and profits. The rewards are the incentive for auto suppliers and dealers to offer such sales incentives to consumers. On the other hand, when auto suppliers request actions of their dealers, they (suppliers and dealers) are bounded by the contractual arrangement as well. However, if the supplier requests that dealers increase consumer advertising for new cars, dealers comply by going along with the manufacturer’s request or not going along in the form of requesting a cooperative advertising program. They choose from their set of response options that (those) which will have an impact on their performance results. Regardless of whether a supplier takes actions or requests actions of the intermediaries, intermediaries comply by implementing decision responses that are action oriented and perceived as necessary to satisfy both partners of the exchange. Furthermore, the intermediaries’ complying decision responses will be within the limits of the contractual arrangement of the trade relation. The three variables that will be manipulated include the level of channel conflict, the level of involvement, and the level of compliance to resolve conflict. There are high and low levels of conflict which arise in channel members’ interactions. Involvement in the conflict situation is an "inherent" characteristic. Therefore, the intermediary’s decision response and level of 43 involvement to implement the decision response can lead to varying levels of channel member compliance. Previous research by channel analysts has theoretically and empirically examined the conflict-performance relation (Brown, 1978; Lusch, 1976b; Pearson, 1973; Pearson & Monosky, 1976; Reve, 1977; Rosenberg & Stern, 1970, 1971; Stern & Gorman, 1969; Walker, 1970). If the impact of conflict on channel member performance is considered in an experimental design framework, the studies heretofore would have employed conflict as the treatment variable and performance as the response variable. However, the relationship between conflict and performance in previous studies was investigated by use of regression analysis to detect significant effects. In the present study, conflict is assumed to be a treatment variable that can be manipulated to have different effects on channel members’ performance. Involvement Channel intermediary involvement is an inherent characteristic of a conflict situation. Thus involvement as used by most investigators is concerned with the emotional commitment a dealer has to resolve a particular conflict issue given its potential impact on dealers’ performance results. In the content of channel conflict, the "issues" over which suppliers and dealers disagree can be characterized by high and low involvement. Within the limits of the relationship between conflict and performance, empirically, it 44 has been shown to be a function of specific issues of conflict used and specific dimensions of performance used (Brown, 1977, 1980). Similarly, the level of channel member involvement in conflict situations should depend on the particular issues used and the channel and/or industry investigated. In this research, "issue involvement" not "response involvement" is the scope of the involvement variable. The psychological literature makes a clear distinction between the two (Apsler & Sears, 1968; Miller, 1965; Zimbardo, 1960). Involvement as used by most investigators of psychological phenomena is defined as the concern with a given "issue" because it is intrinsically involving, and is related to the individual’s needs and values. However, this naturalistic kind of issue-involvement often confounds a number of factors, such as initial differences in commitment. On the other hand, response involvement can be viewed as an individual’s concern with the consequences of his/her "response" or with the instrumental meaning of his/her opinion to the achievement of a desired goal. Thus, response involvement can control the contaminating factors usually associated with the different levels of issue-involvement (Zimbardo, 1960, p. 87). In spite of the potential problems in measuring different levels of issue-involvement, it was felt that within the scope of channel conflict and performance, intermediary involvement with conflict issues is more conceptually rich. Channel members’ level of involvement with particular conflict issues should affect the 45 level of effort expended to make resolution response options work. This will lead to certain consequences--positive or negative effects--with respect to performance outcomes. Compliance Compliance is a behavioral concept in the context of manufacturer-dealer relationships. Manufacturers implement deci­ sions and/or request decision actions of dealers. These decisions or requests are capable of generating conflicts with dealers. In their attempts to resolve conflicts, dealers either react or respond in a manner that is beneficial for both members or for itself solely. There are generally multiple ways to respond once conflict is manifested. A dealer’s or manufacturer’s response action will be less (more) impactful on performance with more (less) response options to resolve a particular conflict situation. In the franchised channel of distribution for automobiles, the manufacturer traditionally has been the more powerful of the two. Thus, an unbalanced power structure exists in this type of channel (Brown & Day, 1981; Gaski & Nevin, 1985; Hunt & Nevin, 1974; Lusch, 1976a, 1976b; Walker, 1970). Although there has been a power gain by dealers who have multiple franchises at multiple locations (superdealers), the locus of power still resides with the manufacturer. Consequently, it is expected that most auto dealers will comply with their suppliers’ requests or decisions to resolve conflict rather than not comply by responding with their own choice. The degree of dealer compliance may be related to the conflict 46 situation; that is, compliance varies across conflict issues that arise. Also, the degree of compliance may vary with the level of involvement of dealers in the conflict issue to respond in a manner to beneficially resolve it. Thus, conceptually, when a dealer responds to conflict with a supplier, whether it is the right thing to do or not, is related to its degree of involvement with the issue. Response of any type will have an impact on some dimensions of dealer performance or total channel performance. As channel intermediaries respond to conflict situations, it is conceivable that their responses may indicate a high or low level of compliance with a request or action taken by their suppliers. Under these circumstances, performance may or may not change. Auto dealers respond with various conflict resolution behaviors. These behaviors may represent going along with their suppliers action or request or not going along yet responding. The level of channel member compliance with resolution behaviors to a conflict situation(s) is (are) greatly influenced by the specific conflict situation and the level of involvement in that situation. The performance impact is not expected to change significantly, regardless of whether the response reflects high or low dealer compliance. The greatest interest in this research lies in the potential interaction effects of franchiser-franchisee conflicts, franchisee involvement in the conflict situations, and franchisee compliance response options on performance impact. The basis for this 47 investigation lies in the theoretical and empirical works by selected channel analysts (Brown, 1978; Lusch, 1976b; Rosenberg & Stern, 1970, 1971; Rosenbloom, 1973; Stern & Gorman, 1969). Stern and Gorman (1969), Rosenberg and Stern (1970), and Rosenbloom (1973) have conceptualized that the relationship between conflict and performance changes over time such that it is curvilinear--"inverted U-shaped" or "U-shaped." That is, the relationship can change from negative to positive and the converse. This is referred to as the "threshold" model of conflict and performance. Other channel analysts (Brown, 1978; Lusch, 1976b; Reve, 1977) found empirical evidence of a "U-shaped" and an "inverted U-shaped" model in their investigations of the conflict and performance assumption. In the context of the curvilinear models, it is of interest to investigate if there are other factors involved which interact with conflict that lead to variations in performance as channel intermediaries change their conflict resolution behaviors. Conflict, as a whole, does not change performance. It is how channel members react or respond to conflict that leads to changes in their performance. Consequently, when a supplier takes action that leads to conflict with its distributors, conflict resolution response options for them will be more limited, particularly in contractual relationships. However, the level of perceived manifest conflicts when a supplier requests actions of a distributor could be similar 48 to that when a supplier takes action. It is conceivable that channel intermediaries have more flexibility to assess the necessity of complying to a request in terms of impactful performance results. When actions are taken by a supplier, reactions by distributors are inevitable. These reactions typically stop short of terminating the channel relationship. The line of distinction between the two stimuli is that an action taken by a supplier requires some type to decision(s) via behavioral reactions by a channel intermediary. If the supplier requests actions, an intermediary will have a wider range of resolution response options to comply with the request, if it complies at all. Consequently, the level of intermediary involvement with a particular conflict situation will be manifested in the efforts put forth and in the intermediary’s perceptions of the derived impact as a result of choosing particular resolution behaviors to resolve franchiser-franchisee conflicts. For example, if the channel intermediary thinks that responding with a particular response option will lead to success or reward- increased sales and profits--for the firm, the level of effort put forth to make a decision response option to resolve the conflict should be high. On the other hand, if there is more uncertainty about the payoff from responding with a particular response option, the intermediary still complies with the request, but the effort put forth to make a resolution response option work should be lower, given the exchange relation. Therefore, reluctance to comply with a requested action(s) should show up in the efforts put forth to 49 implement a decision response option to resolve conflict with a supplier. The behavioral model of channel conflict is akin to an original model conceptualized by Rosenberg and Stern (1970). Their model included three major variables--causes of intrachannel conflict, level of conflict, and outcomes of conflict. Cause-effect relationships are suggested among the three major variables in which there are shifting pairs of independent-dependent variables. Intrachannel conflict can be viewed as a process as well as a state. The complexity of the conflict process in distribution channels has received much empirical investigation (Brown, 1978; Dommermuth, 1976; Firat, Tybout, & Stern, 1974; Lusch, 1976a; Pearson, 1973; Rosenberg, 1974; Rosenberg & Stern, 1971; Shuptrine & Foster, 1976; Walker, 1970). A testable model of the conflict process is presented. It identifies significant variables, shows their hypothesized direction of relation, and permits operationali­ zation and measurement of the constructs. Figure 2.2 presents an overview of the behavioral model of the conflict-performance relation. It asserts that channel member A (supplier) takes or requests actions of channel member B (intermediary), which leads to a heightened state of conflict. The level of conflict present will be associated with the structural arrangement of participating firms. The manufacturer’s decision actions or actions requested of its distributors are capable of generating measurable levels of 50 conflict. The level of conflict is characterized by the nature of the conflict situations thought to occur or that has actually occurred. As the model asserts, the stimuli (conflict situations) may create high or low conflict. The impact of different levels of conflict on the level of effort expended to make response options work and on performance results should be different. If not, the level of conflict may be insignificant. Intuitively, high conflict may lead to high or low levels of involvement in the particular situation over which there is a perceived problem. High involvement should have a different effect on the efforts expended by an intermediary to make resolution response options work, thereby leading to different impacts on the intermediary’s performance results than low involvement. Conversely, low conflict may lead to low or high involvement, as well, which affects the impact of the intermediary’s performance results. Admittedly, the performance results of channel intermediaries in time t should influence the original conflict stimuli in time t + 1. However, this study will not examine this recursive relationship. Furthermore, this study will not attempt to explain "why" performance changes (increases or decreases) with various combinations of conflict, involvement, and compliance. The basic premise tested in this research is that the proposed behavioral model explains more of the variation in distributor performance than do the classical models of conflict and performance. 51 A Proposed Theoretical Framework of Channel Member Conflict and Performance Relationships Channel analysts have conceptualized and empirically investigated the assumption that conflict can have a positive, negative, or curvilinear relationship on channel member performance and the entire channel (Brown, 1978; Lusch, 1976b; Pearson, 1973; Reve, 1988; Rosenbloom, 1973; Walker, 1970). Following previous research approaches, it seems reasonable to assume that conflict will have a significant effect on performance, be it negative, positive, or curvilinear. In the context of the general linear model, the current literature on the conf1ict-performance assumption can be stated in the form of two models: Model 1: = Bo + C^ + e Model 2: = 6q + B, C„ ♦ b2 + E where: Cqj = perceived conflict Yj] = objective performance measure i = 1, 2, . . . , n observations Theoretical conceptualizations (Alderson, 1965; Brown, 1980; Rosenbloom, 1973) and empirical investigations (Brown, 1977; Lusch, 1976b; Pearson, 1973; Reve, 1977; Walker, 1970) have asserted that, in Model 1, Bj is positive and significant. Therefore, increasing levels of conflict leads to increased performance. A wholesaler and a retailer may realize that they both can increase their marketing 52 inputs as a result of a conflictual issue which leads to greater outputs for both. On the other hand, Bj can be negative and significant in Model 1, as well. Channel analysts (Brown, 1978, 1980; Lusch, 1976a) have also conceived and tested a curvilinear or an inverted U-shaped relationship between conflict and performance. In Model 2, this means that Bj will be positive and significant and &2 will be negative and significant. The current theoretical framework is an attempt to reveal that channel analysts need to take a "broader" approach to the study of conflict and performance. The empirical investigations in the channel literature have yielded results that lack consensus regarding the relationship between the two constructs. In the context of this conceptual framework, attempts are made to show that there are some omitted variables that interact with conflict which have stronger and more significant performance impacts. Any empirical test on the main effect of conflict on performance solely is "biased," at best. Also, testing a model that includes only main effects, i.e., conflict, involvement, and compliance, as well as any other conceivable variable, will be biased as well. We would expect to find that other factors interact with conflict. Therefore, any future test of the conflict­ performance assumption should take into account these "other" factors. Referring to Model 3, the main effects of conflict, involvement, and compliance may or may not be significant. However, this conceptual framework asserts that it is more important that the 53 interaction effects of conflict-involvement, conflict-compliance, involvement-compliance, and conflict-involvement-compliance are significant; that is, significant Bs (fy to 67). The purpose of this conceptual framework is to provide a foundation for broader conceptualization and empirical investigation of interorganizational relations in channels of distribution. It is an attempt to better understand what already exists among interdependent channel organizations. This is not a framework that provides a model(s) to "explain" what already exists. The model proposed is valid within the context of the perceptual perspective used in this research. This proposed conceptual framework will be examined by evaluating the perceptions of channel participants in a franchised vertical marketing system for new automobiles. Thus, conflict, involvement, compliance, and performance will be examined as perceptual constructs. Building upon the existing literature on conflict and performance, the current research will attempt to show that there are other factors that can affect a channel member’s performance, namely, involvement and compliance. Also, it can be asserted that conflict interacts with involvement and compliance, which leads to an even greater effect on performance. Using the general linear model framework, the following model is proposed and examined: Model 3: y = Bo + 8-| Conf + invol + 63 Compl + 84 CI + 85 CC + 8g IC + 87 CIC + e 54 where: y = dealer perceived expected performance Conf = dealer perceived manifest conflict Invol • dealer perceived involvement with the conflict issue Comp - dealer perceived compliance with the manufacturer’s action or request CI = conflict x involvement interaction CC = conflict x compliance interaction IC = involvement x compliance interaction CIC = conflict x involvement x compliance interaction In examining the relationship between conflict and performance in a broader conceptual and empirical framework, it is of interest to know whether potential nuisance or uncontrollable independent variables, for example, dealership size, experience, competition, and power, may contaminate the influence of the experimental design variables--conflict, involvement, and compliance interaction effects on channel member performance. For instance, with respect to dealership size, it is expected that large dealers would not experience problems with suppliers to the same degree as small dealers. Many of the large-volume dealers indicated during pretest interviews that issues such as vehicle allocation and distribution, purchase of new product technology, and parts availability pose no major problems for them with auto suppliers, primarily because of their high level of market penetration in their trade areas. Thus, they can get better 55 cooperation from suppliers that is very difficult for small-volume dealers to attain. The degree of experience of dealers in resolving conflictual issues might have an effect on the relationship between conflict­ involvement-compliance and performance. It might be expected that dealers who have had auto franchises for a long period of time tend to have more options available to resolve conflict and to have greater knowledge about how to resolve conflict with suppliers. As a result, conflict would be less frequent and/or intense and resolution behaviors employed would be more mutually beneficial for both channel members. The level of competition in a dealership’s local trade area might affect a supplier-dealer relationship in terms of the frequency and intensity of problems with the supplier’s policies and programs, as well as the number and severity of resolution behaviors. It might be expected that the greater the level of competition a dealer faces, particularly for its primary product item(s) or line(s), the greater the frequency and intensity of conflict. Thus, competition may contaminate the "pure" effects of conflict, involvement, and compliance on performance. Channel member power (exercised or unexercised, perceived or self-attributed) has been conceptually offered and empirically found to resolve conflict on the one hand and to create conflict in supplier-dealer relationships on the other hand (Gaski & Nevin, 1985; Hunt & Nevin, 1974; Lusch, 1976b; Stern & El-Ansary, 1972; Stern & Gorman, 1969; Walker, 1970). Within a given channel of 56 distribution, each channel member has some degree of "relative power." In channels where the locus of power resides with the manufacturer, the greater the relative power of a dealer, the greater the likelihood a dealer will get more cooperation from its supplier, which improves its negotiation position with a supplier(s). Thus, a dealer’s relative power base might affect its performance, regardless of the relationship between conflict and performance. In addition, power may moderate the effect of experimental design variables on performance such that the experimental results are biased. Overall, it seems likely that two of the potential nuisance variables might be correlated in that the larger a dealership’s size, the greater its relative power. Thus, the influence of any potentially uncontrolled variables must be considered to purify the results of a research design. Establishment of Hypotheses A channel of distribution that is highly vertically integrated, i.e., there is a defined power structure--balanced or unbalanced-- between members of the channel will be chosen to test the significance of the behavioral model of conflict and performance. This research will examine the conflict-performance relation with power held constant. The rationale for this is that the relationship between conflict and performance varies with the kind of power relationship between channel members. An asymmetric (unbalanced) power relation will yield one type of relationship, 57 whereas a symmetric or balanced power relation will yield a different type of relationship between the two constructs (Lusch, 1976b). In vertical marketing systems, the channel is highly integrated whereby the opportunity exists for greater interdependence among channel members and high channel stability. Given the higher level of interdependence, there tends to be a greater potential for power relationships to develop (Lusch, 1976a). As a consequence of a defined power relation, conflict is likely to exist, although it may exist at a high or low level. Thus, power can or will have a major impact on the level of conflict in the channel, thereby yielding varying relationships between conflict and performance. Nevertheless, this is another reason why power in this study will be held constant. The ideal means of data collection would be direct observation of the supplier-dealer dyad longitudinally. However, temporal and resource constraints limit the research to a cross-sectional study. For the same reasons, other researchers have conducted their studies similarly (Lusch, 1976b; Roeing & Michie, 1978). The setting for this study is one side of the dyadic channel arrangement for new automobiles. Like other studies of channel conflict and performance (Brown, 1978; Lusch, 1976a; Reve, 1977; Walker, 1970), this research setting is a franchised channel. The limitations and unsolved issues of the existing research resulted in a decision to empirically revisit the relationship 58 between retailer perceived channel conflict and retailer perceived operating performance. The current research attempts to uncover the role that conflict resolution behaviors (degree of compliance) of retailers plays in the conflict-performance relation as well as the degree of involvement in each conflict situation. The setting for the research is the contractual vertical marketing system, that is, the franchise channel for the distribution new automobiles. In this channel, the retailer (franchisee) and the manufacturer (franchiser) are very interdependent, and, therefore, conflicts between them are likely to have an impact on retailers’ performance and total channel performance (Lusch, 1976a). With the trend toward superdealers, they are adding to their power base relative to manufacturers. Yet, an unbalanced power relationship still exists since the franchiser typically has been and remains the most powerful member (McCaulay, 1966). Thus, this type of power relationship is expected to result in franchiser-franchisee conflicts, which lead to a negative impact on franchisees’ perceived expected operating performance. Walker (1970) found that in channels characterized by unbalanced power relationships, they led to suboptimum profits and a high level of dissatisfaction to the franchisee. During preliminary interviews, many automobile dealers described situations that reinforced the negative impact of conflicts on performance. The comments made by a Pontiac dealer who had been a dealer for 20 years are reinforcement of conflict’s negative impact: 59 . . . the success of a dealership depends on the relationship it has with the factory zone managers. If a dealer gets on its bad side, it has the power to disrupt your allocation because they control dealer allocation of new vehicles. Most dealers interviewed during the preliminary investigation of the research indicated that it is better for a dealer to work with manufacturers because they have the power to negatively affect a dealer’s performance in the long term, if not the short term. The following hypotheses will be examined as derived from the channel literature and the proposed conceptual framework: Hl: Conflict, involvement, and compliance will interact in their effect on dealer perceived performance impact. H2: Conflict and involvement will interact in their effect on dealer perceived performance impact. H3: Involvement and compliance will interact in their effect on dealer perceived performance impact. H4: Conflict and compliance will interact in their effect on dealer perceived performance. The proposed relationships among performance and conflict, involvement, and compliance are examined using a balanced repeated measure quasi-experimental design. Using this type of design, an investigator only has control over who is measured and when measures are taken (Churchill, 1987). CHAPTER III METHODOLOGY This chapter describes the methodology used to examine the hypotheses developed in this study. It includes a discussion of the scope of the experimental design. Next, it includes a discussion on construction of the measurement instrument. Finally, the analysis methods are discussed. Experimental Design Overview The scope of this research is to examine conflict and performance in a "perceptual context." The objective is to gather perceptual data from members of the retail end of the franchise channel of distribution for new automobiles. The measurement process involves asking respondents their perceptions of "what would happen" when conflict is manifested instead of actually observing "what would happen" when conflict situations arise with their suppliers. The experimental procedure consists of choosing problem areas among channel members which fit treatment levels of three independent variables. Each respondent is exposed to each situation and evaluates questions pertaining thereto. A procedure which allows control over which respondents’ perceptual measurements are 60 61 taken and when they will be measured on the variables used in the study is essential. However, the tradeoff to having control over respondents ’ measurement is the inability to control the scheduling of the treatments and the inability to randomize the respondents’ exposure to the treatments (Campbell & Stanley, 1966). As mentioned above, the same groups of respondents will be exposed to all treatment-level combinations for the independent variables. Thus, the layout of the experimental procedure in the study fits the requirements of a balanced repeated measure design that is quasi- experimental in nature. The hypotheses for this study have been specified in the form of a linear model, and it is essential that critical assumptions are made to ensure an adequate test of the model. Therefore, a 2 x 2 x 2 factorial design was used in this study as shown in Figure 3.1. The factors (independent variables) include channel member conflict, channel member involvement, and channel member compliance. Each variable was manipulated into high and low levels. Operationalization of Conflict and Involvement The experimental stimuli included four different conflict situations. The conflict situations were selected by examining a collective list of 20 problem areas gathered from past studies on conflict in the auto industry, industry trade publications, and preliminary interviews with the subjects. The conflict situations were selected with several criteria in mind. First, the conflict issues had to be those which all subjects were faced with; and the 62 conflict issues had to be ones which the subjects were motivated to resolve. Finally, the situations had to fit high and low levels of conflict, involvement, and compliance. Involvement High Low HDC HDC High LDC LDC Conflict HDC HDC Low LDC LDC Figure 3.1: A 2 x 2 x 2 design of conflict, involvement, and compliance (HDC = high dealer compliance; LDC = low dealer compliance). Preliminary analysis. In the first phase of the research process, an attempt was made to generate an exhaustive list of actual conflict situations which were currently prevalent in supplier-dealer relationships in the automobile industry. First, industry trade publications were examined by analyzing the results of studies done by automobile trade association groups. Second, preliminary interviews were done with 15 new car dealers in the Lansing-East Lansing-Okemos metropolitan areas. They were asked open questions about conflicts or problems with their primary suppliers. Subjects were asked to identify problem areas and/or 63 asked to verbally indicate problem areas from an assembled list of possible conflict areas before the interview. Also, an exhaustive list of conflict resolution behaviors was generated for each conflict situation. Subjects were also asked to reveal the actions they would take to resolve each of the conflict situations mentioned. New car dealers were selected for this preliminary analysis by using the yellow pages in the telephone directory to attain their addresses and telephone numbers. Afterward, each dealer was sent a prenotification letter describing the nature of the research inquiry and requesting one hour of its time for a personal interview. Also, they were told that they would receive a telephone call to confirm their interest and to set a time to conduct the interview. Respondents were contacted by telephone approximately three days after sending out the prenotification letters to allow sufficient time for them to receive and read them. The interviews lasted from 30 minutes to one and one-half hours. To insure that a cross-section of dealers were selected, they were grouped by size and type of new car franchise. Afterward, they were chosen in a manner to ensure both large and small dealers who have either a domestic, foreign, or domestic/foreign franchise were selected for the interviews. The respondents interviewed were mostly owners and/or general managers of the franchise. These subjects were principals in the organization who interacted directly with their auto suppliers’ representative (the district manager and/or the zone manager) in dealer-factory relationships. 64 During the interviews with some of the dealers, other personnel--the auto parts manager, the service manager, and/or the sales manager--participated in the interviews to provide insights on conflict issues raised and resolution behaviors when these issues were prevalent. Each dealer was asked for permission to tape-record the interview. They were assured the interview would remain confidential and anonymous. All respondents granted permission and the sessions were recorded. The purpose for recording the interviews was to facilitate the process of formulating (1) an exhaustive list of current conflict issues in factory-dealer relations and (2) an exhaustive list of resolution behaviors for each conflict issue. Some of the recorded interviews can be found in Appendix A. Considerable knowledge was gained about factory-dealer relationships in the automobile industry from the interview sessions. Dealers were very willing to talk about their experiences in the automobile business. At the conclusion of the interview, some dealers requested a copy of the results of the study. The conflict situations selected were as follows: (1) new vehicle allocation and distribution for the low conflict-high involvement category, (2) adoption of new product technology for the high conflict-low involvement category, (3) parts availability for the low conflict-low involvement category, and (4) use of the customer satisfaction index (CSI) surveys for the high conflict-high involvement category (see Figure 3.2). 65 Involvement High Low Conflict Issue 4 Conflict Issue 2 Customer Satisfaction Purchase of new High Index (CSI) rating product technology survey to service new models Conflict Conflict Issue 1 Conflict Issue 3 Low New vehicle alloca­ Parts availability tion and distribu­ tion Figure 3.2: Conflict situations for levels of conflict and involvement. Unique criteria were used to select the four conflict situations included in this study from exhaustive lists of situations that fit the treatment levels of conflict and involvement. First, the conflict situations occur on a regular basis (annually or bi annually) with the introduction of new and/or improved models by manufacturers. Second, the aforementioned four conflict situations were the most frequently cited by dealers in the preliminary interviews when the lists of problem areas were being assembled. Third, these issues were found to present the most problems in manufacturer-dealer relationships from the results of studies done by auto industry trade groups (e.g., National Automobile Dealers Association). Finally, these conflict areas are 66 most realistic in terms of the potential impact they can have on dealers’ operating performance. For example, new vehicle allocation and distribution issues (low conflict-high involvement) can affect a dealer’s sales volume, trade area market share, and profits. In other words, when a dealer has problems getting allocation of new popular models from the manufacturer, this can reduce its sales volume and market share relative to competitive dealers in the trade area. On the other hand, when they get their allocation of popular models, this usually entails having to accept some slow-selling models as well. As a result, profits can be reduced due to the rise in inventory costs affiliated with storage of these automobiles until they are sold. Thus, this presents a great incentive for dealers to get "involved" with this issue to make it work for them. The purchase of new product technology (high conflict-low involvement conflict situation) can have an impact on a dealer’s performance by potentially reducing its profits. That is, on the one hand, there are the high costs involved with the purchase of new product technology and special tools and equipment needed to use the technology for servicing new model vehicles. On the other hand, many dealers feel that they should buy the new technology to provide their customers with top-quality after-sale service. However, this issue poses problems when the equipment is rarely used and/or needed to service a new model and yet they are pressured by the manufacturer to buy it. Parts availability--low conflict-low involvement conflict situation--can potentially reduce a dealer’s profits as a result of 67 buying all parts from the manufacturer directly when they can be purchased from less expensive sources. However, this issue was chosen from a list of low conflict-low involvement conflict situations, in addition to the aforementioned reasons, because, on the one hand, dealers feel that it is better to purchase parts directly from the manufacturer to ensure consistent quality. On the other hand, some fast-moving parts produced by the manufacturer can be purchased less expensively at local auto parts stores. However, the manufacturer generally requests that dealers purchase parts for their primary make of vehicle(s) from it directly and exclusively. Thus, this is a low-grade problem area between manufacturers and dealers yet it is prevalent through the year. The manufacturer’s use of customer satisfaction index (CSI) surveys was the high conflict-high involvement conflict situation chosen for this study. Low CSI ratings for a dealership can mean low market penetration (sales volume and market share) given that customers tend to blame dealers for low-quality products as a surrogate and/or representative for the manufacturer, on the one hand. This conceivably can include brand switching to a foreign car by customers on their next purchase. On the other hand, the manufacturer blames dealers for low market penetration, which can lead to future allocation and expansion problems. Thus, the dealer is caught in the middle and has to get "involved" with this issue to develop resolution behaviors that are mutually beneficial for itself and the manufacturer. 68 Low conflict-high involvement. For every low conflict-high involvement conflict situation, subjects read the following scenario: "The factory (primary supplier) has an unwritten policy with respect to new vehicle allocation and distribution. That is, to get the desired quantity of a particular new model, you must accept more than you would like of other models." Care was taken not to confuse getting allocation of a particular new model with having to accept slow-sei ling cars. According to many of the dealers, a dealer’s new model allocation in any year is based on the number of cars a dealer sold in the previous year. The more cars you sell, the greater your allocation will be and converse. A dealer will be allocated any new model vehicle that is produced. However, the quantity of a new model allocated is determined by a dealer’s "rate of travel" or "turn and earn." Thus, "turn and earn" or "rate of travel" for a given dealer determines the quantity of a new model allocated, not whether a dealer will or will not receive a particular model. All subjects are well informed about the relationship between allocation and "rate of travel." One exception to this procedure is that for General Motors (GM). GM only produces cars for orders from dealers. Therefore, if a dealer does not order new models, there will be no allocation. One Chevrolet dealer indicated that if a customer comes in to buy a particular version of a new model and the dealer does not have it on order with GM, the dealer cannot get that car. Therefore, the 69 customer will have an extended waiting period before getting possession of the new model after purchase. High conflict-low involvement. The second conflict situation dealers were exposed to was "adoption of new product technology." It addressed the issue of dealers purchasing special tools and equipment required to use the new product technology. For this high involvement-low conflict situation, subjects read the following scenario: "The factory (primary supplier) has informed you that you will not receive your typical allocation of new models unless you purchase the new product technology designed to service them." Low conflict-low involvement. The low conflict-low involvement conflict situation was "parts availability." It attacks the issue of whether dealers have sufficient parts available--factory parts and/or alternative source parts--to provide fast yet quality service for its primary make of car. Subjects were exposed to the following scenario: "The factory (primary supplier) has requested that you purchase all parts for your primary make from it exclusively." This situation is of concern to franchises because the parts can be obtained less expensively from another source. One Cadillac dealer indicated that GM’s fast-moving parts (spark plugs, filters, hoses, etc.) can be purchased at a local auto parts store cheaper than buying them directly from GM. High conflict-high involvement. Finally, subjects were exposed to the fourth conflict situation of high conflict-high involvement: 70 "the Customer Satisfaction Index (CSI) survey." The scenario read by the subjects was as follows: "The factory has informed you that the CSI survey is designed to assess the quality of your sales and service rendered rather than the quality of the manufacturer’s product." This situation was carefully stated after pretesting to insure that subjects understood the nature of the issue being addressed. The CSI survey poses a problem in that customers fill out the questionnaire thinking that they are rating the quality of the manufacturer’s product, instead of the quality of a dealer’s sales and service rendered. Thus, a dealer’s CSI ratings could be very low whereas they would ordinarily be relatively high. Dealers are very concerned because the supplier has the power, as specified in the contractual agreements: (1) to revoke a franchise with consistently low CSI ratings, (2) to not grant an additional store location to a franchisee with consistently low ratings, and (3) to directly affect a franchisee’s new sales and profits by varying its allocation schedule. Operationalization of Compliance To resolve conflict in channels of distribution, channel intermediaries may or may not comply or go along with the actions or requests of their supplier(s). For this research, an attempt was made to investigate conflict resolution behaviors in the franchised channel of distribution for automobiles where certain conflict situations were prevalent. 71 In the preliminary investigation, exploratory research was done to generate an exhaustive list of resolution behaviors--actions that dealers would undertake--to reduce the level of conflict with their suppliers. After asking subjects to identify problem areas with their suppliers and/or to comment on a preassembled list, they were asked, "What would you do?" to resolve each issue. Thus, an exhaustive list of resolution behaviors was generated from subjects during this phase of the research (see Tables 3.1 through 3.4). Selection process. With respect to high and low compliance, all of the conflict resolution behaviors generated in the preliminary stage of this research across all conflict situations presented were written out in no particular order. Afterward, each response option was classified as a high compliance or low compliance resolution behavior. The procedure used to classify the resolution behaviors was to assess each in terms of whether it was a response to a supplier’s action or request (conflict issue) solely or whether it was a response to the conflict yet the subject does not believe that doing what is requested by its supplier will be mutually beneficial. For example, auto suppliers usually request that new car dealers buy parts from them exclusively. This is a potential conflict situation in that some dealers indicated that they can get the supplier’s parts much cheaper from an alternative source than buying them directly from the manufacturer. Other dealers were buying nonfactory parts which were even cheaper in an effort to reduce costs. In this situation, when dealers purchased the parts from the 72 Table 3.1.--High conflict-high involvement CRB (CS = new vehicle allocation distribution).8 High Compliance CRB . Work with the factory to isolate areas for improvement. . Try to understand the factory’s perceptions of the CSI survey’s use. . Use the New Vehicle Inspection Program (NVIP) to correct any errors before it is released to a customer. . Use customer relations personnel to call on customers who recently purchased a new vehicle and those whose vehicles were serviced to discover areas for improvement. . Offer more incentives to customers to fill out the survey to yield positive ratings. Low Compliance CRB . Develop your own interpretations of the CSI survey results. . Request that the factory redesign the CSI survey to allow cus­ tomers to evaluate both the quality of the product and the quality of your sales and services. . Request that the factory solicit from dealers specific areas to improve the quality of the product given that you provide the service for the product when there are problems. . Protest the current use of the CSI survey. aCRB = conflict resolution behaviors; CS = conflict situation. 73 Table 3.2.--High conflict-low involvement CRB (CS » new product technology purchase).3 High Compliance CRB . Purchase the special tools and equipment to service a new model from the factory even though the new product technology is rarely used. . Purchase the special tools and equipment to service a new model from alternative supply sources that are approved and recom­ mended bv the factory. . Purchase the special tools and equipment to service a new model from alternative supply sources of your choice. Low Compliance CRB . Request that the factory use performance ("turn and earn") solely to determine whether or not you received your allocation. . Request that the factory permit you to use existing product technology to service new models if it is appropriate. . Do not purchase the new product technology and thus do not receive typical new models. aCRB « conflict resolution behaviors; CS = conflict situation. 74 Table 3.3.--Low conflict-high involvement CRB (CS = parts availa­ bility).8 High Compliance CRB . Order more than the desired quantity of new cars to ensure that you receive the desired quantity. . Engage in demand forecasting for different models and build inventory for those models which may become popular in the future. . Accept the factory’s typical dealer allocation of new models. Low Compliance CRB . Use a computerized locating system or a conventional locating system to find special vehicles and engage in a dealer trade. . After locating a vehicle you purchase it outright. . Concentrate your efforts on buying and selling used cars when new models are not available. . Appeal to your district manager to increase your allocation. . Protest the factory’s unwritten policy in regards to new vehicle allocation and distribution. aCRB = conflict resolution behaviors; CS = conflict situation. 75 Table 3.4.--Low conflict-low involvement CRB (CS - customer satis­ faction index (CSI) survey use).a High Compliance CRB . Purchase all parts from the factory. . Purchase all parts from the factory except for those which require back-ordering. . Purchase all parts from the factory except those which could be purchased less expensively from an alternative source. Low Compliance CRB . Purchase "slow-moving" parts from the factory and "fast-moving" parts from an alternative source of supply. . Purchase all parts from your primary make alternative source of supply if the parts are less expensive. . Do not purchase all parts for your primary make from the factory exclusively. aCRB = conflict resolution behaviors; CS = conflict situation. 76 manufacturer solely, this would be a high compliance response to resolve conflict. If, on the other hand, a dealer only purchased slow-moving parts from the supplier and fast-moving parts (spark plugs, filters, etc.) from an alternative source, this represents a low compliance response behavior to resolve conflict. Thus, each conflict resolution behavior was subjected to this analysis before categorization. The next stage of categorizing conflict resolution behaviors for the experimental subjects was to associate each with the four conflict situations used as the stimuli. Therefore, a second list was prepared, consisting of each conflict situation followed by high compliance and low compliance resolution behaviors. The final stage of this part of the research consisted of categorizing the resolution behaviors according to high and low conflict and high and low involvement. Response options were classified as requiring high involvement if a dealer had to be emotionally involved with the conflict situation in order to make the resolution response work--that is, make a positive impact on its performance. This analogy holds regardless of whether the behavior is the correct or incorrect thing to do. On the other hand, if a resolution response required little or no emotional involvement in the conflict situation to make it work, the behavior was classified as a low involvement response. For example, an auto manufacturer’s warranty program can pose a problem for dealers in that the manufacturer has the power or discretion to determine "excess charges" by dealers for warranty work performed. In this case, reimbursement for warranty work is a low conflict situation and it requires low involvement by a dealer to make it work. In addition, the dealers are knowledgeable about the manufacturer’s discretionary power with respect to reimbursement costs because it is specified in the franchise contractual agreement. On the other hand, if the manufacturer informed the dealer that a new showroom is needed in a different location, this may be a low conflict situation, but a dealer would have to get highly involved in the building and relocating to make this action work because of the substantial cost involved and potential sales loss due to competitive reaction and lost customer patronage. Consequently, at least three conflict resolution behaviors were used for high and low compliance and at least six responses for each category of high and low conflict with high and low involvement. The next section presents conflict and involvement for high and low compliance conflict resolution responses. Compliance resolution behaviors: High/low conflict and involvement. In the aggregate, for each cell in Figure 3.1, high and low compliance resolution behaviors were evaluated. First, the respondents were asked to allocate 100 points among high compliance resolution responses, and second, to do the same 100-point allocation for low compliance resolution responses. For example, the conflict situation which represented high conflict-high involvement dealt with the manufacturer’s administration and use of the customer satisfaction index (CSI) surveys to evaluate the r W 78 quality of dealers’ sales and service rendered. Respondents were asked to evaluate the relative effectiveness of each conflict resolution response for high and low compliance to the manufacturer’s action (see Table 3.1). The high conflict and low involvement conflict situation which respondents were exposed to centered on dealers’ adoption of new product technology. They were asked to evaluate a series of resolution behaviors concerning the source from which to purchase special tools and equipment to use the new product technology. Table 3.2 reveals the conflict resolution responses for high and low dealer compliance with the manufacturer’s action or request. New vehicle allocation and distribution represented the low conflict-low involvement conflict situation. Dealers allocated 100 points among high and low compliance resolution responses listed in Table 3.3. Finally, parts availability represented the low conflict-low involvement conflict situation. Dealers were asked to allocate 100 points among high and low dealer compliance resolution responses to resolve this conflict situation. Table 3.4 lists the specific responses for high and low compliance. After dealers evaluated the relative effectiveness of the responses for high and low compliance, they were asked to choose the typical or most likely way they would resolve each of the four conflict situations. Among high compliance resolution responses, dealers were asked: 79 "If you go along with the factory’s ____________ , what is the typical (most likely) way you resolve this conflict situation given the responses listed in Case A above?" Among the low compliance resolution responses, dealers were asked: "If you do not go along with the factory’s ____________ , what is the typical (most likely) way you resolve this conflict situation given the responses in Case B above?" After reading the question, dealers were asked to circle the appropriate response number which corresponded with the numbered resolution responses in the two cases before the question asked. The response numbers were listed after each of the questions in the same order as presented when dealers allocated 100 points among them. Measurement of the Dependent Variable Operationalization of performance. The dependent variable used in this research is channel member performance. The operationalization of it is given below. A copy of the questionnaire (Type B) used to collect the data is included in Appendix C. Most measures of performance represent objective, quantitative evaluation of costs, sales, profitability, return on equity, return on investment, input-output analysis, value added, and the strategic profit model of channel or channel member performance that are determined by the behavioral relationship in the channel (Sibley & Michie, 1981). Lusch (1976b) used return on asset and asset turnover in his study of conflict and performance. 80 In this research, performance is treated as a "perceptual" construct because of the difficulty of isolating the exclusive impact of behavioral responses by subjects to objective measures such as sales, costs, profits, etc. Furthermore, the sociopsychological and organizational behavioral literature has developed performance as a sound behavioral construct (Cartwright & Zander, 1968; Child, 1974; McMahon & Perritt, 1973; Tannenbaum, 1968; Woodward, 1965). Sibley and Michie (1981) measured channel member performance as a behavioral construct where dealers were asked to indicate how successful they were at implementing 18 policies in their firms. For each policy, performance was measured on a 5-point scale ranging from very unsuccessful to very successful. An index of dealer performance was computed as the average sum of dealer responses to each policy. The index was found to be valid and reliable. Consequently, the method used to measure performance in this research is a variation from the aforementioned research studies. Channel member performance was operationalized with the following items: total new car sales volume, trade area market share, and profits. These performance items were selected from the domain of performance measures as a result of personal interviews with some of the dealers and experts. Subjects on the panel of experts from the automobile industry were exposed to a series of resolution behavioral responses for each conflict situation. They were asked to rate the expected impact of each conflict resolution response on the three performance measures on a 7-point scale 81 ranging from 1 (extremely negative impact) to 7 (extremely positive impact). Derivation of the dependent variable. Some channel analysts (Brown, 1977; Lusch, 1976b; Pearson, 1973; Walker, 1970) have used objective measures of performance and related them to conflict. In fact, most studies about performance in channels of distribution have used objective measures of performance. The analyses have been examined at the retail, wholesale, or manufacturer levels in the channels (Sibley & Michie, 1981). In this study, perceived expected performance is a derived dependent variable. Performance is directly tied to the conflict resolution behaviors dealers choose to resolve conflicts with their suppliers. Thus, the expected impact of a resolution response on four performance measures represents only part of the information used to compute a composite expected value performance impact score. Data from two dealer groups were used to derive values of the dependent variable. Group 1 was exposed to high and low compliance resolution responses that would resolve each of four conflict situations. These resolution responses were chosen to be unique for each conflict situation. Dealers were asked to allocate 100 points among, first, the high compliance responses, and second, the low compliance responses. The point allocation was designed to reflect a dealer’s perception of the relative effectiveness of each response to resolve the conflict situation relative to other responses in the same set. Also, allocation of points across a response set reflects 82 a dealer’s perception of the probability or likelihood that each response would resolve the conflict situation with a supplier. Group 2 dealers (panel of experts) were exposed to the same set of resolution responses as the first group of dealers. However, they were asked to rate the perceived expected impact of each response on three performance measures: total new car sales volume, trade area market share, and profits. Experts rated the impact of each response across the four conflict situations on a 7-point scale ranging from 1 (extremely negative impact) to 7 (extremely positive impact). The objective of using the panel of experts was to have an independent group evaluate the impact of the responses on performance. This process would eliminate built-in bias by allowing the same group of dealers to evaluate the relative effectiveness of resolution responses to resolve conflict issues and to evaluate their impact on performance. Thus, Group 1’s evaluation will serve as "weights" or "probabilities" which are combined with Group 2’s mean standardized impact scores on the four performance items. The procedure to derive the expected value performance impact scores is a multi-step process described as follows: Step 1: The evaluations by the panel of experts will be standardized for the three performance items by using the standard deviation. It is determined across N resolution responses (high and low compliance) for each of the first three performance items for a particular cell. Then, the impact score for each expert on each performance item is standardized by adjusting the score by the standard deviation for the ith performance item across N responses for the kth conflict situation. That is, 83 ^ijkr ~ ^ijkr / (1) where: ^ijkr “ Standardized impact score for the rth expert on the ith performance item and the jth resolution response for the kth conflict situation Yijkr = the unstandardized impact score for the rth expert on the ith performance item and the jth response for the kth conflict situation S.^ = the standard deviation for the ith performance item across resolution responses for the kth conflict situation = resolution responses for the kth conflict situation r = 1, 2, . . . , R experts j = 1,2,...,^ resolution responses k = 1, 2, . . . , K conflict situations i = 1,2,...,! measured performance items Finally, an average impact score for new car sales volume, market share, and profits is calculated across R experts for each of resolution responses. That is, R 2^ = Zij / R (2) where: Zjj = average standardized impact score for the ith performance item and the jth resolution response across R experts A fourth performance impact score is calculated for each expert by summing its scores across the three measured items. Second, an average summed standardized impact score is calculated across R experts for each of the resolution responses. Each performance item’s mean standardized impact score will have a different weight or probability used from Group 1’s evaluations to determine an expected value performance impact (EVPI) score. 84 Each performance item’s mean standardized impact score will have a different weight or probability used from Group 1’s evaluations to determine an expected value performance impact (EVPI) score. A large standard deviation for a performance item means that the experts are less alike in their perceptions of the impact of the jth response on the ith performance item across K conflict situations. On the other hand, performance items with small standard deviations or variances indicate that the experts are more alike in their perceptions of the impact of the jth resolution response on the ith performance item. The perceived impact scores across the three performance items are standardized to put all scores in the same unit of analysis. These procedures eliminate the differences in the perceived impact of resolution responses across the three performance items as determined by the size of the means and standard deviations. Step 2: Group 1 respondents allocated points across and N2l resolution responses. They are divided by 100 to convert them to weights or probabilities across the four conflict situations. For example, Pjkt ■ Xjkt / 100 (3) where: Pjkt = respondent’s probability weights for the jth response relative effectiveness to resolve the kth conflict situation under the tth compliance condition Xjkt = points allocated out of 100 total for the jth response relative effectiveness to resolve the kth conflict situation under the tth compliance condition » responses for high compliance to resolve the kth con­ flict situation 85 N£k = responses for low compliance to resolve the kth con­ flict situation ^lk+^2k = ^k responses to resolve the kth conflict situation 4 N = z Nb k=l t =1,2, (high, low) compliance situations Step 3: For each respondent in Group 1, its probability vector across Nb resolution responses will be combined with average standardized impact scores (Z^js) for the four performance items calculated in step 2 Jabove. The Z^s are constant across respondents. The expected value per­ formance impact scores are calculated as follows: P N EVPI = Z Z Wil Zij (4) 1=1 j=l J where: EVPI = Expected value performance impact scores across N]and N2k compliance responses and K conflict situations Wji = Probability weight for the 1th respondent and the jth resolution response Zjj = Mean impact score for the ith performance item and the jth resolution response 1 = 1, 2, 3, . . . , P dealers For dealer 1 across high and low compliance responses, there will be four equations across the performance items: one for total new car sales volume, trade area market share, profits, and a summed performance item. For example, 86 Performance Item 1: EVPIllk = ^ij 7ij + ^ij+1 7ij+l + ^ij+2 7ij+2 + ♦ ♦ • + wij+N-j *1j+N-j (5) Performance Item 2: EVPI12k = w2j *2j + w2j+l *2j+l + w2j+2 *2j+2 + • • • + w2j+N-j 72j+N-j W Performance Item 3: EVPI13k = W3j 73j + W3j+1 *3j+l + W3j+2 73j+2 + • • • + w3j+N-j 73j+N-j (7) Summed Performance Item 4: EVPI14k = W4j 74j + W4j+1 74j+l + W4j+2 74j+2 + ♦ ♦ ♦ + W4j+N-j 74j+N-j (8 ) Across the four conflict situations and eight cells in the design, there are 32 equations which yield 32 EVPI values for the 1th dealer. The mean standardized impact scores (Zjqs) in the equations will be the same across P dealers for the same resolution response j, but they will be different across responses and I performance items. On the other hand, the weights (W^s) in the equations may be different across P dealers given the possibility that each dealer will have different likelihood vectors across N^ and N2|< responses and K conflict situations. However, the weights will be the same across I performance items. 87 Validation Marketing Ph.D. students and a small set of auto dealers acted as "judges" in a pretest to independently confirm that the situations elicited appropriate "high or low" conflict and "high or low" involvement. Four marketing Ph.D. students were shown the four conflict situations and asked to categorize each. After explaining the nature of each conflict situation, the Ph.D. students classified them which matched that of the research. In addition, 10 dealers were exposed to each conflict situation and asked to comment on the nature of the issue described. Their comments and questions about the content of each conflict scenario were subjectively used to confirm their classification as high or low conflict and as high or low involvement. Each dealer evaluated all four conflict situations. Thus, the cell sizes in terms of the number of subjects were the same for each of the eight cells in the design. Dealers were instructed to view each conflict situation as an independent event. To ensure that this was possible, each conflict situation was placed on a separate page of the questionnaire. All questions which pertained to that situation were listed on that page. Development of Data-Collection Instruments Two types of questionnaires were used in this study: type A for new car dealers and type B for a panel of experts (new car dealers who had been in business at least 20 years). The questionnaires were different only in the way the dealers and 88 experts evaluated the resolution responses to resolve four generic conflict situations. Dealers were asked to evaluate the relative effectiveness of response options to resolve conflicts in factory-dealer relationships in two cases. In Case A (dealer compliance with the factory’s action or request), they were asked to assume that dealers go along with their supplier’s action or request. The subjects allocated 100 points among resolution responses which represented compliance with their supplier’s action or request. The allocation procedure was designed to indicate the subjects’ perceptions of the most and least effective response to undertake to resolve a specific conflict situation. In Case B (dealer noncompliance with its supplier’s action or request), respondents were asked to assume that they did not go along with their primary supplier’s action or request. They were again asked to allocate 100 points among noncompliance resolution responses to reflect the relative effectiveness of each to resolve each of four conflict situations. At least three resolution response behaviors were provided for compliance and noncompliance with the manufacturer’s action or request. Finally, the respondents were asked to indicate the typical or most likely way they would resolve each conflict situation under compliance and noncompliance presumptions. A panel of experts were asked to evaluate the "expected impact" of dealers’ response behaviors on three performance measures: impact on total new car sales volume, impact on trade area market share, and impact on profits. The subjects performed their 89 evaluations for the four conflict situations and all response options included on questionnaire type A. Pretest of the Questionnaire Both questionnaires (type A for new car dealers and type B for a panel of automobile experts on factory-dealer relationships) were subjected to a pretest. Eight of the 15 dealers used in the preliminary investigations were selected to pretest questionnaire type A, and two dealers were selected as experts to pretest questionnaire type B. The pretest took place approximately one year after the preliminary investigations were completed. Respondents for the pretest were contacted in the same manner as they were when the preliminary interviews were done. If the owner or general manager for a franchise had changed since the first interview, only the owner of the franchise was contacted and interviewed during the pretest. The prenotification letters sent were a little different in that respondents were reminded of the preliminary interviews a year earlier and the purpose of the research. Also, time was requested for a second interview to evaluate the questionnaire for clarity and confusion. Interviews were granted by all dealers contacted. They were asked to complete the questionnaire and to ask questions and/or make comments or suggestions as they filled it out. The pretest interviews lasted on the average 30 minutes. Respondents’ comments, questions, and suggestions increased the length of the interview. It was estimated that it would take a respondent between 20 and 25 minutes to 90 complete the questionnaire once it was received by mail. Only minor changes were necessary for each form of the questionnaire. Data Collection The data to test the hypotheses were collected as part of a census of new car dealers in Michigan. Seven hundred eighty members and 145 nonmembers of the Michigan Automobile Dealers Association (MADA) were selected. The MADA provided the sampling frame (mailing list) for its members. In addition, the association provided the list of nonmembers which it keeps a record of, yet they were not on the current mailing list. All members and nonmembers were chosen as the group from whom to obtain data because the objective was to have a substantial response rate. Furthermore, generalizations can be made about the group of new car dealers from which usable questionnaires were received. Thus, a total of 925 questionnaires were sent out on the first wave. The instrument was sent out the last week of August 1988. A follow-up reminder was mailed approximately four weeks after the initial wave of questionnaires. A postcard was the instrument used to convey the message. They were mailed to all of the original 925 new car dealers, members and nonmembers of MADA. Dealers who had not returned the questionnaire were instructed to call the association toll-free for another copy of the questionnaire if they had not received one or it had been misplaced. Those who had returned the questionnaire were thanked and told to disregard the reminder. Arrangements were made with the MADA to check whether a 91 dealer was a member or nonmember. This was done to keep track of the form of questionnaire which was mailed out. Also, the researcher wrote down numbers on the questionnaire by the dealer’s name and address when they called to request another copy. For nonmembers, the follow-up reminder was mailed one week later due to the unavailability of the mailing list. Questionnaires were sent to both members and nonmembers. This was done to check for differences in response to their perceptions of conflict and involvement. Also, interest centered on how they chose to resolve conflict with their supplier either by complying with (go along with) their supplier’s request or action, or not complying with (do not go along with) their supplier’s action or request. A panel of experts were selected from the list of new car dealers (members and nonmembers) to collect data on measures of the dependent variable, performance. Criteria used to select this group were that they had to be new car dealers who had been a dealer for at least 15 years and/or who had worked in the retail end of new car sales for at least 15 years before purchasing a franchise. In addition, the owner had to have the same name as the franchisee to partially insure that they met the first criterion. This process was done by examining the files on members and nonmembers from the MADA and checking the date at which a dealer became a member of the association and checking the name of the franchisee and its dealership’s name. Therefore, a total of 20 dealers (members and nonmembers) were selected as the panel of experts. The remainder of 92 new car dealers were sent a similar yet different questionnaire. It was designed to obtain data on the relative effectiveness of alternative conflict resolution behaviors under conditions of dealer compliance and noncompliance with its primary supplier’s actions or requests. The respondents used in this research are a cross-section of domestic, foreign, and domestic/foreign new car dealers who retail new cars from a cross-section of U.S. and foreign automobile manufacturers. Manipulation Checks It is expected that the subjects’ involvement level with a conflict situation will vary according to the issue raised. In the context of this research, the subjects’ perception of the level of manifest conflict is related to the level of inherent involvement. Manipulation checks were addressed for conflict, involvement, and compliance. Conflict. With respect to manifest conflict for subjects in Group 1, after subjects read each conflict situation, they were asked two questions. First, they were asked, "How frequently have you faced this particular conflict situation in the past three years?" Then, they answered by writing in the number of times each conflict situation had occurred. This single-item measure of conflict was adapted from studies in the channel literature (Brown & Frazier, 1978; Etgar, 1979; Foster & Shriptrine, 1974, 1976; Hunger & Stern, 1976; Kelley & Peters, 1976; Lusch, 1976a, 1976b). Second, 93 subjects were asked, "How intense would you characterize typical disagreements over this conflict, regardless of whether the disagreements were settled or not?" Then, they revealed their perception of the intensity of the conflict by marking on a 7-point rating scale ranging from "not very intense" to "very intense." This scaled item as a measure of conflict conformed to that used in the literature (Assael, 1968; Etgar, 1979; Hunger & Stern, 1976; Pearson, 1973; Pruden, 1969; Rosenberg & Stern, 1971; Stern, Sternthal, & Craig, 1973b). Involvement. Involvement is an inherent characteristic of the conflict situation faced. It was essential to check for variation across subjects with respect to their perceptions of high and low involved conflict situations. Subjects answered three statements which were adapted from studies in the socialpsychological literature on issue involvement (Apsler & Sears, 1968; Miller, 1965; Sherif, Kelly, Rogers, Sarup, & Littler, 1973; Zimbardo, 1968). A multi-item scale of involvement is desirable to enable the researcher to obtain a reliability estimate. The following statements were developed for this research: "I am very concerned about the factory’s (conflict issue) to any dealership." "I am very motivated to resolve this conflict." "The factory’s (conflict issue) is very important to my dealership’s success." Then they were asked to reveal their perceptions of involvement in each conflict situation by marking on three 7-point rating scales (strongly disagree, neither, strongly agree). 94 Subjects in the panel of experts were asked the same multi-item questions about conflicts, although slightly modified to make them amenable for this group. Also, subjects responded to similar multi­ item statements about involvements with each conflict situation. Compliance. Channel member compliance was manipulated to assess how subjects resolved their conflicts with their suppliers. In this research, an attempt was made to categorize the resolution behaviors, first, according to whether they represented high compliance or low compliance, and second, to classify them in terms of high and low conflict and high and low involvement. Moderator variables. Data were also collected for several moderator or covariate variables from the respondents. With respect to dealership size, respondents were asked, "Approximately, what is your dealership’s average annual turnover (new car sales volume in units) over the past 3 years?" In terms of dealer experience, respondents were asked, "How long has your dealership been in business under its present name?" and "How long has your dealership been in business at its present location?" They answered by writing in the "number of years" and the "number of months" for each question. Trade area competition was measured by asking respondents, "In your trade area, how many other franchised dealers are there for your primary make of vehicle?" They answered by writing in the "number of dealers." Finally, dependence was used as a surrogate measure of power. Respondents were asked to indicate "the approximate percentage of 95 their total new car business (past 3 years average) represented by each of their suppliers." They were instructed to write in the "percent of sales volume" derived from either a single supplier or multiple suppliers. The rationale for using dependence as a surrogate for power is that it is suspected that the greater the number of supply sources for a given product, the less dependent a dealer is on a particular supplier and the greater its "relative power" with that supplier. Analysis Methods The nature of the relationship among channel member conflict, involvement, compliance, and performance was assessed using the univariate approach to multivariate analysis of variance (MANOVA) for a quasi-experimental design. Four separate MANOVA procedures were done, one for each of the four expected value performance impact measures of the dependent variable, to test the effect of the full model. Also, simple, simple models (two-way interaction effects) and main effects models were tested using the aforementioned analytical method. In addition, multivariate analysis of covariance (MANOCOVA) was performed to remove the influence of several covariate or uncontrolled variables on the dependent variable that might contaminate the pure effects of the experimental treatment factors. The 2x2x2 factorial design was assumed to be a fixed-effect model. Thus, a series of hypotheses were tested with respect to examining expected value performance impact for various factor-level combinations. CHAPTER IV RESULTS This chapter presents statistical results showing the validity of the data. Specifically, it includes a discussion of the manipulation checks for the experimental treatments. Next, it includes a discussion of the reliability for measures of the experimental treatments. It also includes a discussion of the perceived relative effectiveness of conflict resolution behaviors. In addition, there is a discussion of the perceived impact of conflict resolution behaviors on performance. Finally, it includes a discussion of the study’s response rate and its representa­ tiveness. Manipulation Checks The theoretical framework proposed in this research asserts that there are other factors that may interact with conflict to affect the performance outcome of a channel member. Conflict, involvement, and compliance are the three factors investigated in terms of their relationship with performance. The research design was set up to include two levels (high and low) for each of the factors. Figure 4.1 shows the conceptual categorization of the research design. 96 97 Involvement High New Product High CSI Survey Technology Use (4) Purchase (2) Conflict New Vehicle Low Allocation and Parts Distribution (1) Availability (3) Figure 4.1: Conceptual categorization of conflict issues across levels of conflict and involvement (CSI = Customer Satisfaction Index). Conflict Table 4.1 reveals the conflict situations, ordered from low to high for conflict and involvement. Two measures of conflict were used (frequency of conflict and intensity of disagreement) to build a case for high and low conflict situations. The mean and median scores are shown for each variable. The data reveal that two of four conflict situations are slightly different from their conceptual categorization with respect to conflict. New vehicle allocation and distribution (NVA-D) is conceptually a low conflict situation. Yet the data show that this situation occurred, on the average, eight times over the past three years. Also, when this issue arises, it leads to more intense disagreements between dealers and manufacturers, with an average score of 3.8. Across the four conflict situations, the average number of times each situation 98 occurs ranges from a low of three times for new product technology purchase to a high of eight times for new vehicle allocation and distribution and Customer Satisfaction Index (CSI) rating surveys. A plausible reason why the data show new vehicle allocation and distribution issues as a high conflict instead of low conflict is the composition of the study’s respondents. The majority of auto dealers who responded (46.9%) are rural or small-town dealerships or perceive themselves as moderate-volume franchises (52.3%). The preliminary investigations and comments reported by some dealers indicate that small dealers have more problems getting their allocation of new models. Manufacturers base dealer allocation on sales volume in the previous year or "turn and earn" or "rate of travel." That is, the more cars a dealership sells, the greater its allocation of new models for the current period and the converse. The purchase of new product technology (NPT) for new models, conceptually, is a high conflict issue. The data reveal that it is perceived as a low conflict issue in that it occurred, on the average, three times over the past three years. Also, the intensity of disagreement is somewhat low, with an average intensity score of 3 on a 7-point scale. NPT purchase has no direct impact on a dealership’s sales volume or trade area market share. Given that it may increase a dealership’s costs and reduce profits, it seems plausible that this issue would be a high conflict situation between dealers and manufacturers. Also, it seems plausible that most dealers comply with the manufacturer’s request and buy the new product technology regardless of whether they believe it is the 99 Table 4J.--Empirical categorization of conflict issues by Group 1 dealers. LCSa HCSa NV A/D PA NPT CSI Mean Median Mean Median Mean Median Mean Median Conflict Frequency of conflict 8.00 3.00 3.30 5.00 3.00 2.00 8.00 3.00 (# of times over 3 yrs.) Intensity of disagreement 3.80 4.00 2.80 2.00 3.00 3.00 3.70 4.00 LI(CS)b HI (CS)b NPT PA NVA/D 1DS I Mean Median Mean Median Mean Median Mean Median Involvement Concern for issue 4.34 4.00 5.30 6.00 5.27 6.00 5.60 6.00 Motivation to resolve 4.88 5.00 5.30 6.00 5.60 6.00 5.60 6.00 issue Importance of issue to 4.60 5.00 6.40 7.00 6.60 7.00 5.60 6.00 dealer success aLCS = low conflict situation; HCS « high conflict situation. ^LI(CS) « low involvement conflict situation; HI(CS) = high involvement conflict situation. 100 right thing to do or not. Issues that relate to parts availability and CSI survey use are conceptually and empirically low conflict and high conflict situations, respectively, as revealed by the results in Table 4.1. The next step in the process of examining the manipulation checks was to determine whether significant differences exist between the mean scores across each conflict situation and each operational measure of conflict and involvement. In this study, the paired comparison t-test was appropriate to use to examine the differences in mean scores given correlated samples of respondents (Churchill, 1987). Table 4.2 shows the results of using a paired t-test for a statistically significant difference in mean scores for Group 1 respondents. It was found that, with respect to conflict, the means’ difference (3.834) for "frequency of disagreement" is statistically significant when the two low conflict situations are compared (t = 5.230, p < .001). Also, the difference in means (-5.059) is statistically significant when the two high conflict situations are compared (t = -5.260, p < .001). When "intensity of disagreement" was used for conflict, the means’ difference (1.089) was significant for the two low conflict situations (t = 7.81, p < .001). In addition, the means’ difference (-0.414) is statistically significant (t = -2.530, p < .02) for the two high conflict situations. The size of the absolute value of the means’ differences is large enough in the case of the low and high conflict situations 101 Table 4.2.--Results of paired t-tests for differences in means (Group 1 subjects). LCSa HCSa NVA-D/PAb NPT/CSI Useb dic T-Value d^c T-Value Conflict Frequency of disagreement 3.834 5.230d -5.059 -5.260d (.733) (.962) Intensity of disagreement 1.089 7.810d -0.414 -2.530e (•139) (.164) LI(CS)a HI(CS)a NPT/PAb NVA-D/CSI Useb djC T-Value dic T-Value Involvement Concern for issue -0.052 -0.360 -1.280 -8.390d (-147) (.153) Motivation to resolve 0.324 2.530e -0.754 -6.110d issue (-128) (.124) Importance of issue to 0.157 1.810f -0.996 -5.900d dealer success (.087) (.169) aLCS = low conflict situations; HCS = high conflict situations; LI(CS) = low involvement conflict situations; HI(CS) = high involve­ ment conflict situations. bNVA-D = new vehicle allocation and distribution; NPT = new product technology; PA = parts availability; CSI « customer satis­ faction index ratings. cValues in parentheses represent the standard error of mean difference (d^). dp < .001. ep < .02. fp < .08. 102 when frequency of disagreement is used for conflict to conclude that they are operationally significant as well. This means that categorizing the conflict issues as "low" and "high" conflict situations reinforces the notion that they are classified on a "relative" basis. As for "intensity of disagreement," the size of the difference in means for the two low conflict situations is borderline to being operationally significant although it is statistically significant. For the two high conflict situations, the means’ difference is not operationally significant because it is less than the absolute value of one. It was concluded that these findings are somewhat beyond expectations in that it was anticipated that there would be little or no perceived difference between the two low conflict issues and the two high conflict issues. Nevertheless, these findings were not so antithetical as to render testing this study’s hypotheses meaningless. The panel of experts responded to a slightly different form of the questionnaire. Their perceptions of the four conflict situations with respect to conflict provide a better match to the conceptual categorization (see Table 4.3). The average and median scores range from a low of 3.7 and 4.0 for new vehicle allocation and distribution to a high of 4.7 and 6.0 for CSI survey use with respect to the "frequency" of conflict. In terms of the "intensity of disagreement" over the four conflict issues, the mean and median 103 Table 4.3.--Empirical categorization of conflict issues by Group 2 dealers. LCSa HCSa 1MVA/D PA NPT CSI Mean Median Mean Median Mean Median Mean Median Conflict Frequency of conflict 3.7 4.0 3.9 4.0 4.0 4.0 4.7 6.0 (# of times over 3 yrs.) Intensity of disagreement 3.7 4.0 4.9 5.0 4.5 4.5 5.4 6.0 LI (CS)b HI (CS)b 1MPT 1PA NVA/D CSI Mean Median Mean Median Mean Median Mean Median Involvement Concern for issue 5.0 5.5 6.3 6.0 6.6 7.0 6.1 7.0 Motivation to resolve 5.8 6.0 5.6 5.0 6.3 7.0 6.1 6.0 issue Importance of issue to 6.2 6.5 5.9 6.0 6.7 7.0 5.3 6.0 dealer success aLCS = low conflict situation; HCS = high conflict situation. ^LI(CS) = low involvement conflict situation; HI(CS) « high involvement conflict situation. 104 scores range from 3.7 and 4.0 for new vehicle allocation and distribution problems to 5.4 and 6.0 for CSI survey problems. Table 4.4 reveals the results of using a paired t-test to examine the statistical significance of the differences in means for conflict and involvement for the low and high conflict issues for Group 2 respondents. It was expected that there would be little or no perceived difference in conflict and involvement for within-low and within-high conflict situations. The most important criterion for this group of respondents was consistency in evaluating the issues presented. When "frequency of disagreement" and "intensity of disagreement" were used for conflict, no statistically significant results were found in the differences in means for low and high conflict situations. However, the size of the difference (dj > 1) is large enough to render the results operationally significant except in one case. Involvement Three items were used to measure respondents’ level of involvement in conflict issues: (1) concern about the issue, (2) motivation to resolve the issue, and (3) the importance of the issue to dealer success. As the results show, these items were reliable measures of involvement. Table 4.1 reveals that the conceptual and empirical categorizations of the four conflict situations have little to no discrepancy across the three items measuring involvement for Group 1 dealers. A 7-point scale ranging from 1 105 Table 4.4.--Results of paired t-tests for differences in means (Group 2 subjects). LCSa HCSa NVA-D/PAb NPT/CSI Useb d^c T-Value «ic T-Value Conflict Frequency of disagreement -0.143 -0.240 -1.333 -1.230 (.595) (1.085) Intensity of disagreement -1.143 -1.550 -1.167 -1.340 (.738) (.872) LI(CS)a HI(CS)a NPT/PAb NVA-D/CSI Useb dic T-Value «ic T-Value Involvement Concern for issue -1.500 -2.090d 0.429 0.630 (.719) (.685) Motivation to resolve 0.167 0.420 0.143 0.240 issue (.401) (.595) Importance of issue to 0.667 0.420 1.429 1.470 dealer success (.401) (.972) aLCS = low conflict situations; HCS = high conflict situations; LI(CS) » low involvement conflict situations; HI(CS) = high involve­ ment conflict situations. bNVA-D = new vehicle allocation and distribution; NPT = new product technology; PA = parts availability; CSI = customer satis­ faction index ratings. cValues in parentheses represent the standard error of mean difference (d^). dp < .10. 106 (strongly disagree) to 7 (strongly agree) was used across the three items. The mean and median scores for "concern for the conflict issue" range from 4.34 and 4.00 for new product technology purchase to 5.60 and 6.00 for CSI survey use. The mean and median scores for "motivation to resolve the conflict issue" range from 4.88 and 5.00 for new product technology purchase to 5.60 and 6.00 for new vehicle allocation and distribution issues and CSI survey issues. The mean and median scores for "importance of the conflict issue to a dealer’s success" range from 4.60 and 5.00 for new product technology purchase to 6.60 and 7.00 for new vehicle allocation and distribution issues and 6.40 and 7.00 for parts availability. A discrepancy between the conceptual and empirical categorizations occurs with parts availability issues and CSI survey issues with respect to high and low involvement on one of the three items. The mean and median scores for "importance of the conflict issue" are 6.40 and 7.00 and 5.60 and 6.00 for parts availability issues (high involvement) and CSI issues (low involvement), respectively. The conceptual classifications of the two issues are low and high involvement, respectively. The data reveal that there may be a third level of involvement (moderate involvement) which would capture these two conflict situations, at least for Item 3. Involvement in a conflict issue was measured by using three items (see Table 4.5). The four conflict situations were also categorized with respect to high and low treatment levels of this factor. New product technology (NPT) and parts availability (PA) 03 Table 4 .5 . --Standardized r e lia b ility c o e ffic ie n ts fo r involvement item s. 107 108 represented the low involvement issues, whereas new vehicle allocation and distribution (NVA-D) and customer satisfaction index (CSI) survey use represented the high involvement issues. Table 4.5 reveals that the differences in means are statistically significant for all three measures of involvement for the low and high involvement issues except for one case. However, in terms of operational significance, the results are not significant except for "concern for the conflict issue" for the high involvement issues. The means’ difference is greater than the absolute value of one (1.280). The remaining differences in means are less than one (the absolute value). Overall, these findings reveal that there are little or no perceived differences between the two low involvement issues and between the two high involvement issues except for one measure of involvement. Although the results are statistically significant, operationally, the differences in means are small enough to make it worthwhile to proceed with testing the hypotheses in this study. In Group 2 (the panel of experts), the conceptual and empirical categorizations of the four conflict situations are consistent except for one of the three items measuring involvement: "importance of the issue to a dealer’s success" (see Table 4.3). The discrepancy occurs for new product technology purchase and CSI survey use, which, conceptually, are low and high involvement conflict issues, respectively. The mean and median scores are 6.20 and 6.50 and 5.30 and 6.00 for new product technology purchase and CSI survey use, respectively. 109 Table 4.4 shows that there is no statistically significant difference in the means for the three measures of involvement across the low and high conflict situations except for one case. That is, the means’ difference for "concern for the issue" is significant when new product technology and parts availability are compared (t = -1.500, p < .10). The sizes of the differences in the means are greater than the absolute value of one in only two cases. These findings provide stronger support for prior expectations that there would be little or no perceived difference for within-low and within-high conflict situations. As a result, testing the hypotheses in this study is worthwhile, subject to some disparity in the results. Compliance The conflict resolution behaviors which were categorized for high and low compliance for each of the four conflict situations were preexamined with dealers to determine whether they felt each response fit the preestablished classification. They were asked whether they felt a particular resolution behavior was a high or low compliance response for each conflict situation as a pretest manipulation check. The pretest results were used to alter or to maintain the a priori list of high and low compliance resolution behaviors. Thus, compliance was not subjected to the same type of manipulation check as were conflict and involvement due to the nature of this construct. Also, it was thought that the variance among the results from the respondents with respect to whether or no not a particular resolution behavior was a high or low compliance response would be so small that it would be meaningless. High and low compliance conflict resolution behaviors were examined in a "relative" context. That is, the "low" compliance resolution behaviors all consist of actions that the dealer took that were either mutually beneficial to the manufacturer and itself or ones that dealers mostly would benefit from or those in which the manufacturer’s requests or action to resolve the conflict were protested by dealers. On the other hand, "high" compliance resolution behaviors were ones in which dealers thought the manufacturer had a legitimate right to request to resolve the conflict or they were ones which dealers thought were warranted to be mutually beneficial or they were ones which dealers implemented because of little to no choice because of the manufacturer’s potential to exercise power--supplier benefits mostly. Thus, the level of compliance suffers, in particular, in that there are "degrees" of high compliance actions and "degrees" of low compliance actions as well as "degrees" of high or low compliance actions to resolve conflict between manufacturers and dealers. Reliability It appears that "importance of the conflict issue" is not a dimension that measures involvement. The discrepancy on this item occurs for both groups of dealers. Furthermore, reliability coefficients are higher across both groups using the first two items solely as indicators of involvement across the four conflict Ill situations (see Table 4.5). Except for a few cases, the reliability coefficients are higher when "importance of the conflict issue" (Item 3) is removed. The coefficients range from .2953 to .6740 when Item 3 is included for Group 1 respondents and .7649 to .9571 for Group 2 respondents. With Item 3 excluded from the set, the reliability coefficients range from .5099 to .6891 for Group 1 dealers and .7143 to .9708 for Group 2 dealers across the four conflict situations. Also, Table 3 reveals that the reliability coefficients are higher for Group 2 across the four conflict situations with and without Item 3 removed from the analysis except in a few cases. This may be attributed to the fact that Group 2 contains only 7 dealers, whereas Group 1 consists of 217 to 243 dealers. Two items were used to measure conflict across the four conflict situations. The data reveal that "frequency of conflict" and "intensity of disagreement" are each distinct measures of conflict. All reliability coefficients are less than .40, and the correlation between the two items is less than .25 across the four conflict situations and across both groups of dealers. Although the data clearly reflect some unique problems with respect to the directionality when the conflict situations are ordered from low to high conflict and low to high involvement, the study still shows some promise such that the hypotheses can be tested. The mean and median scores for measures of conflict and involvement are slightly different for the two low conflict 112 situations and the two high conflict situations (see Tables 4.1 and 4.3). The greatest difference occurs with respect to the mean and median scores between the low and high conflict situations. Even though the scores are different and reveal the expected direction (low to high) except for one case for conflict, they probably would reveal insignificant differences in a statistical test. It is necessary to point out that the levels of conflict and involvement were developed in a "relative" context. That is, each of the four conflict situations used was chosen to "fit" high and low levels of conflict and involvement. Thus, each of the low perceived conflict situations is low relative to each other. The same rationale holds for the two high conflict situations as well. Given an exhaustive list of high and low conflict situations in manufacturer-dealer relationships, all low conflict situations would be perceived relative to each other and all high conflict situations would be perceived relative to each other. Also, high and low conflict situations are evaluated relative to each other as well. The manipulation checks for each of the treatment factors are somewhat weak, yet they are strong enough to render it meaningful to test the hypotheses. Consequently, this study seems to be "valid" although not as strong as one would hope. Thus, the empirical results more closely match the conceptual categorization of the four conflict situations. Therefore, it seems appropriate to run the factorial experiment to test the hypotheses of this research. r 113 Perceived Relative Effectiveness of Conflict Resolution Behaviors Dealer compliance centers on whether a dealer goes along with a manufacturer’s action or request or chooses not to comply by responding independently of a manufacturer’s request to resolve a conflict issue. In this research, at least three resolution responses were generated across the four conflict situations under conditions of high and low compliance. Dealers were asked to allocate 100 points across the resolution responses under both conditions. This process is akin to assigning a probability to a given response in terms of the perceived likelihood that it is (will be) effective at resolving a particular conflict situation relative to other resolution responses. The greater the number of points out of 100 total a dealer allocates to a particular response, the greater the perceived likelihood that a response is (will be) effective at resolving a particular conflict situation relative to other responses. This applies for high and low dealer compliance. Tables 4.6 through 4.9 show the mean and median points allocated per resolution response. The most typical way dealers chose to resolve conflict is indicated by the percentage who choose a particular response. Across the four conflict situations, the responses are arranged from the highest mean and median points allocated to the lowest for high and low dealer compliance. For example, given new vehicle allocation and distribution problems, most dealers perceive that the most effective way to resolve this conflict is to (1) accept the manufacturer’s typical allocation to 114 Table 4.6.--Univariate statistics for resolution responses (CS • new vehicle allocation and distribution conflict).8 Mean& Median^ nc Percentd ne High Dealer Compliance 1. Accept the manufacturer’s typical 37.5 30 219 34.3 230 allocation 2. Engage in demand forecasting 37.0 33 212 34.8 230 3. Order more than the desired quantity 31.6 30 217 30.9 230 to ensure getting the desired quantity Low Dealer Compliance 4. Dealer trade after vehicle location 28.5 25 219 33.6 232 5. Appeal to district manager 27.0 22 219 31.5 232 6. Concentrate effort on buying and 23.0 20 220 21.1 232 selling used cars 7. After vehicle location, purchase it 16.0 15 204 5.6 232 outright 8. Protest the manufacturer’s unwritten 14.9 10 183 8.2 232 policy aCS = conflict situation. fylean and median points allocated per resolution response out of 100 total points. cSample size used to compute the mean and median. ^Percentage of respondents who chose a particular response as the "most typical" way to resolve the conflict situation. eSample size used to compute the percentage. 115 Table 4.7.--Univariate statistics for resolution responses (CS « new product technology purchase).a Mean^ Median^ nc Percent^ ne High Dealer Compliance 1. Purchase STE to use NPT from the factory 56.8 50 217 61.8 228 2. Purchase STE to use NPT from alternate source approved and recommended by the 33.8 33 178 24.1 228 manufacturer 3. Purchase STE to use NPT from alternate 26.0 20 168 14.0 228 source of dealer choice Low Dealer Compliance 4. Request that manufacturer permit dealer to use existing product technology if 58.0 50 210 54.5 220 it is appropriate 5. Request that the manufacturer use per­ formance (turn and earn) to determine 43.0 40 188 39.1 220 allocation 6. Do not purchase new product technology 12.0 5 139 6.4 220 aCS = conflict situation. ^Mean and median points allocated per resolution response out of 100 total points. cSample size used to compute the mean and median. ^Percentage of respondents who chose a particular response as the "most typical" way to resolve the conflict situation. eSample size used to compute the percentage. 116 Table 4.8.--Univariate statistics for resolution responses (CS = parts availability).3 Mean*5 Median*5 nc Percent^ ne High Dealer Compliance 1. Purchase all parts for primary make 46.8 40 194 40.4 225 from the manufacturer 2. Purchase all parts for primary make from the manufacturer unless they are 41.0 34 205 33.8 225 cheaper from an alternative source 3. Purchase all parts from the manufac­ turer except those that require back- 27.6 25 184 25.8 225 ordering Low Dealer Compliance 4. Do not purchase all parts for primary 55.0 50 175 50.9 212 make from the manufacturer 5. Purchase all parts for primary make 36.7 33 165 24.1 212 from alternate sources 6. Purchase slow-moving parts from the manufacturer and fast-moving parts 33.7 30 165 25.0 212 from alternate sources 3CS = conflict situation. fylean and median points allocated per resolution response out of 100 total points. cSample size used to compute the mean and median. ^Percentage of respondents who chose a particular response as the "most typical" way to resolve the conflict situation. eSample size used to compute the percentage. 117 Table 4.9.--Univariate statistics for resolution responses (CS « Customer Satisfaction Index (CSI) rating survey).3 Mean*5 Median*5 nc Percent^ ne ‘ High Dealer Compliance 1. Use customer relations personnel to call on customers who recently purchased a 29.6 25 212 30.9 230 new vehicle or had it serviced under warranty for areas of improvements 2. Work with the manufacturer to find 29.0 25 210 32.2 230 areas for improvement 3. Use the new vehicle inspection program (NVIP) to find errors before the 28.6 25 204 22.6 230 vehicle is released to a customer 4. Try to understand the manufacturer’s 18.8 20 195 9.1 230 perceptions of dealer CSI ratings 5. Offer customer incentives to yield 8.0 154 5.2 230 positive ratings Low Dealer Compliance 6. Request that the manufacturer redesign the CSI survey to allow customer to 46.0 40 202 50.2 215 evaluate the quality of the product and quality of dealer sales and service 7. Request that the factory solicit from dealers areas for improvement of new 32.0 30 181 25.1 215 vehicles 8. Dealer develop its own interpretations 26.0 20 161 18.1 215 of its CSI ratings 9. Protest the manufacturer’s use of the 16 10 150 6.5 215 CSI survey results aCS = conflict situation. ^Mean and median points allocated per resolution response out of 100 total points. cSample size used to compute the mean and median. ^Percentage of respondents who chose a particular response as the "most typical" way to resolve the conflict situation. eSample size used to compute the percentage. 118 dealers (mean and median points are 37.5 and 30.0, respectively) or (2) engage in demand forecasting for models that may become popular --mean and median points allocated are 37 and 33. Also, the majority of the dealers typically resolve this conflict situation with their suppliers with response 1 (34.3%) and response 2 (34.8%). The data also reveal that most dealers tend to comply with their suppliers’ requests or actions, given the higher mean and median points allocated among high compliance responses relative to low compliance resolution responses. Across the four conflict situations, the resolution responses that received the highest mean and median allocated points were the most typical way dealers chose to resolve these conflict issues with their suppliers. Except for new vehicle allocation and distribution issues, a low compliance resolution response was perceived to be more effective at resolving parts availability problems, new product technology issues, and CSI survey problems in terms of the mean and median points allocated. As for new vehicle allocation and distribution, dealers are aware of the difficulty of getting the quantity of new vehicles desired and the fact that they must accept some "slow selling" vehicles if they are to get the desired ones. Many dealers revealed that the relationship with the district manager and the zone manager can affect a dealer’s allocation. This may be a plausible explanation of why dealers are more likely to comply with their supplier’s action or request to resolve this issue. 119 Thus, the resolution response analysis served as the basis for derivation of values for the dependent variable in this study. Impact on Performance Group 2 respondents rated the perceived impact of the set of resolution responses evaluated by Group 1 respondents on three performance items: total new car sales volume, trade area market share, and profits. The respondents rated each item on a 7-point scale ranging from 1 (extremely negative impact) to 7 (extremely positive impact). Across the four conflict situations, the data reveal that resolution responses have similar impact. The mean and median scores are more similar for two of the three performance items--new car sales volume and trade area market share. For new vehicle allocation and distribution problems, Table 4.10 shows that the mean and median impact scores for new car sales volume and trade area market share range from 2.71 and 2.00 to 5.27 and 5.00, respectively. Also, the variances across these two performance items are smaller than those for profit impact except in two cases. This seems to indicate that the experts are more similar in their perceptions of the impact of high and low compliance resolution responses on performance. The data also show that the variances are generally larger and the mean and median impact scores lower than the other performance measures. Consequently, it appears that there are greater perceptual differences among the panel of experts. In general, the data reveal that the resolution responses have somewhat 120 of a positive impact on new car sales volume and trade area market i share. As for profit impact, the perceptions among the experts are that the responses to resolve new vehicle allocation and distribution conflict issues have, in general, somewhat negative and positive impacts on profits--mean and median impact scores range between 3.33 and 3.00 and 4.71 and 5.00, respectively. Table 4.10.--Univariate statistics for the performance impact of CRR: CS = New vehicle allocation/distribution.a Mean Median Variance High Compliance Responses Dealer accepts the factory’s typical allocation of new models . New car sales volume 4.71 5.00 1.57 . Trade area market share 4.57 4.00 1.62 . Profits 3.57 4.00 3.91 Dealer orders more than the desired quantity . New car sales volume 5.29 5.00 .91 . Trade area market share 5.27 5.00 .91 . Profits 3.57 3.00 3.27 Dealer engages in demand forecasting . New car sales volume 5.00 5.00 1.00 . Trade area market share 5.00 5.00 1.00 . Profits 4.43 5.00 1.29 121 Table 4.10.--Continued. Mean Median Variance Low Compliance Responses Dealer engages in a trade after locating special models . New car sales volume 5.14 5.00 2.81 . Trade area market share 4.17 4.00 2.91 . Profits 4.57 5.00 3.95 After locating a vehicle, dealer purchases outright . New car sales volume 4.86 5.00 3.14 . Trade area market share 4.67 4.50 2.68 . Profits 4.00 5.00 5.67 Dealer concentrates on buying and selling used cars when new orders are not available . New car sales volume 2.71 2.00 4.57 . Trade area market share 2.71 2.00 4.91 . Profits 4.50 4.50 3.50 Dealer appeals to the district manager to increase allocation . Total new car sales volume 4.57 5.00 2.29 . Trade area market share 4.71 5.00 2.23 . Profits 4.71 5.00 1.57 Dealer protests the manufacturer’s unwritten policy in terms of NA/D . Total new car sales volume 3.16 5.00 .57 . Trade area market share 3.33 3.50 .67 . Profits 3.33 3.00 .27 aNo perceived impact scores are represented by 4.00 on the scale. CRR - conflict resolution responses; CS = conflict situa­ tion. 122 As for new product technology purchase conflict issues, there are very small differences in the experts’ perceptions of the impact of resolution responses on performance (see Table 4.11). The mean, median, and variance are nearly equal across the three performance items for low compliance resolution responses. Overall, these findings seem to indicate the experts perceive low compliance resolution responses to have a very positive impact on performance with the exception of not complying at all. The mean and median impact scores across the three performance items range between 1.50 and 1.67. The variances on these items are less than one. These findings seem to indicate that the experts perceive noncompliance with the manufacturer’s request or action will have an extremely negative or very negative impact on a dealer’s performance. Some dealers explained in the preliminary phases of the study that if they fail to purchase the new product technology needed to service new-model vehicles, the manufacturer can retaliate in a future period when it comes to their allocation. Manufacturers prefer and request that dealers purchase parts for servicing cars from them. Many dealers feel that they should use parts supplied by the manufacturer particularly for their primary make of vehicle in their efforts to provide high-quality auto repair service. However, Table 4.12 shows that when there are problems between manufacturers and dealers on the source of parts purchase, behavioral resolutions have no impact or a slightly negative impact on new car sales volume and trade area market share. The mean and median impact scores range from 3.71 and 4.00 to 4.17, 123 Table 4.11.--Univariate statistics for the performance impact of CRR: CS - New product technology purchase.® Mean Median Variance High Compliance Responses Dealer purchases special tools and equipment to service new models from the factory . Total new car sales volume 3.29 4.00 1.24 . Trade area market share 3.71 4.00 .24 . Profits 2.71 3.00 1.24 Dealer purchases special tools and equipment to service new models from alternate sources approved and recommended by the manufacturer . Total new car sales volume 4.00 4.00 .40 . Trade area market share 4.00 4.00 .40 . Profits 4.33 4.00 1.07 Low Compliance Responses Dealer requests that the manufacturer use performance (turn and earn) to determine allocation . Total new car sales volume 5.86 5.00 1.14 . Trade area market share 5.86 5.00 1.14 . Profits 5.86 5.00 1.14 Dealer requests that the manufacturer permit it to use existing product technology . Total new car sales volume 5.50 5.50 1.90 . Trade area market share 5.67 6.00 1.87 . Profits 5.67 6.00 1.87 Dealer does not purchase new product technology . Total new car sales volume 1.67 1.50 .67 . Trade area market share 1.67 1.50 .67 . Profits 1.50 1.50 .30 aNo perceived impact scores are represented by 4.00 on the scale. CRR = conflict resolution response; CS » conflict situation. 124 Table 4.12.--Univariate statistics for the performance impact of CRR: CS » Part availability.3 Mean Median Variance High Compliance Responses Dealer purchases all parts from the factory . Total new car sales volume 4.00 4.00 2.33 . Trade area market share 4.00 4.00 2.33 . Profits 4.00 4.00 1.33 Dealer purchases all parts from the factory except for those that require back-ordering . Total new car sales volume 4.17 4.00 .17 . Trade area market share 4.17 4.00 .17 . Profits 4.57 4.00 1.62 Low Compliance Responses Dealer purchases "slow-moving" parts from the manufacturer and "fast-moving" parts from an alternative source . Total new car sales volume 3.71 4.00 1.57 . Trade area market share 3.71 4.00 1.57 . Profits 5.00 5.00 1.00 Dealer purchases all parts from alter­ nate source if they are less expensive . Total new car sales volume 4.71 4.00 . Trade area market share 3.83 4.00 .17 . Profits 4.57 4.00 2.29 Dealer does not purchase all parts from the factory . Total new car sales volume 4.00 4.00 . Trade area market share 4.00 4.00 . Profits 4.57 4.00 1.62 aNo perceived impact scores are represented by 4.00 on the scale. CRR = conflict resolution response; CS = conflict situation. 125 respectively. The pattern is consistent across high and low compliance resolution responses. Noteworthy is the perception by experts of the impact of resolution responses (high and low) on profits. It appears that actions by dealers to resolve this conflict have a somewhat positive impact on profits--mean and median impact scores range between 4.00 and 5.00. A major source of conflict between auto manufacturers and dealers is the manufacturer’s use of Customer Satisfaction Index (CSI) surveys. In the early phase of this study, many dealers expressed concern and feel that the survey is designed to allow customers to blame them for low-quality products. Yet the contractual relationship permits the manufacturer to "push" a dealer to improve its sales and service rendered. Table 4.13 shows that high and low compliance resolution responses have a perceived positive impact on performance except in one case of noncompl iance-- a dealer protests the use of the CSI survey. The mean and median impact scores are 3.67 and 4.00, respectively, across the three performance items. The variances are the same (.27) for the three performance items, as well. The data seem to indicate that noncompliance will have a somewhat negative impact or no impact on a dealer’s performance results. 126 Table 4.13.--Univariate statistics for the performance impact of CRR: CS = CSI survey use.a Mean Median Variance High Compliance Responses Work with the manufacturer . Total new car sales volume 5.67 5.00 .67 . Trade area market share 5.50 5.50 1.10 . Profits 5.16 5.00 .97 Dealer tries to understand the manu­ facturer’s perceptions of the CSI survey’s use . Total new car sales volume 4.43 4.00 .62 . Trade area market share 4.43 4.00 .62 . Profits 4.28 4.00 .24 Dealer uses the new vehicle inspection program to correct error before a vehicle is released to a customer . Total new car sales volume 5.50 5.50 1.10 . Trade area market share 5.33 5.50 1.47 . Profits 5.50 5.00 1.50 Dealer uses customer relations personnel to call on customers who recently pur­ chased a car or had it serviced to discover areas for improvement . Total new car sales volume 5.86 6.00 1.48 . Trade area market share 5.86 6.00 1.48 . Profits 5.71 7.00 2.91 Dealer offers some incentives to cus­ tomers to yield positive rating . Total new car sales volume 4.00 5.00 5.00 . Trade area market share 4.00 5.00 5.00 . Profits 3.86 4.00 4.81 Low Compliance Responses Dealer develops its own interpreta- tions of the CSI survey results and makes changes accordingly . Total new car sales volume 5.14 4.00 3.14 . Trade area market share 5.14 4.00 3.14 . Profits 5.33 5.50 3.47 r 127 Table 4.13.--Continued. Mean Median Variance Dealer requests that the manufacturer redesign the CSI survey so that it reflects evaluation of the quality of the product and the quality of its sales and service rendered . Total new car sales volume 5.14 4.00 3.14 . Trade area market share 5.14 4.00 3.14 . Profits 5.27 4.00 2.57 Dealer requests that the manufacturer solicit from dealers areas to improve the quality of the product . Total new car sales volume 6.17 7.00 1.77 . Trade area market share 6.17 7.00 1.77 . Profits 6.17 7.00 1.77 Dealer protests the current use of the CSI survey . Total new car sales volume 3.67 4.00 .27 . Trade area market share 3.67 4.00 .27 . Profits 3.67 4.00 .27 aNo perceived impact scores are represented by 4.00 on the scale. CRR = conflict resolution response; CS = conflict situation. Also, it is perceived by the experts that offering customers incentives to fill out the survey to yield positive ratings (a high compliance response) has no impact or a somewhat positive impact on new car sales volume and trade area market share--mean and median impact scores are 4.00 and 5.00, respectively. However, the variances of impact scores are very large on both items. These findings seem to indicate that this resolution response is perceived to have no impact on profits; mean and median scores are 3.86 and 4.00, respectively. The variance of the profit impact score is 128 large (4.81), as well. The results are consistent with expectations of this response action to resolve CSI conflict issues. Given the large variances across respondents on the three performance items, it appears that the experts are very different in their perceptions of the impact of offering customers incentives to get good ratings on performance. Also, this large disparity in perceptions on this resolution response might indicate that some dealers are determined to get good ratings by any means relative to others. The potential consequences that a manufacturer can invoke for consistently low ratings can be severe, i.e., possibly revoke the franchise or deny a dealer’s request for an additional store location. Overall, the data seem to indicate that dealers are better off complying outright or at least to some degree with their suppliers in the actions they take to resolve conflictual issues. The worst that can happen by complying is that the actions to resolve conflict will have "no impact" on a dealer’s performance. Sample Representativeness The current mailing list for members of the Michigan Automobile Dealers Association (MADA) served as the sampling frame for this study. The MADA also provided the most current list of new car dealers who were nonmembers, as well. The sampling frame for current members of the MADA contained 780 new car dealers. The mailing list for nonmembers contained approximately 145 dealers in Michigan. 129 A census was taken of new car dealerships in the state of Michigan. The mailing list for association members was very current; thus, sampling error was minimum for this group. The mailing list for nonmembers was not as up-to-date, for many of the instruments were returned either because the dealership had moved or gone out of business. Thus, the frame error was somewhat higher with this group. Overall, frame error has minimum effects on the results of the study. In taking a census, the problems of sampling error, nonresponse error, and sampling frame error associated with random sampling are eliminated or kept to a minimum. Two forms of the questionnaires were used in this study. Form A was mailed to approximately 925 members and nonmembers of the MADA. A total of 252 instruments were returned. However, the number of usable questionnaires was 243. The response rate for Form A of the research instrument was 26.3%. The number of usable questionnaires received from MADA members was 233, for a response rate of 29.9%. Nineteen usable instruments were returned by nonmembers, for a response rate of 13%. Group 2 (panel of experts) was mailed Form B of the questionnaire. Twenty-five instruments were mailed to this group, of which seven were usable. This resulted in a response rate of 28%. This group consisted of 15 association members and 10 nonmembers. Five usable instruments were returned by members and two usable ones by nonmembers. The response rates for members and nonmembers of this dealer group were 33% and 20%, respectively. 130 Overall, across both groups of dealers, 250 usable questionnaires were returned after one wave of the questionnaire was sent out and one wave of follow-up postcards. Therefore, the overall response rate for this survey was 25%. This is an acceptable response when compared to studies that take surveys of business firms. Although the response rate was lower than expected, it is within the range of others reported in the channel literature where studies were national or regional surveys. That is, this study’s response rate (26.4%) compares favorably with those reported for previous channel studies in terms of a low response rate (Butaney & Wartzell, 1988, 11.08%; Etgar, 1976, 19%; Michie, 1978, 16%; Roering & Michie, 1978, 16%); a moderate response rate (Brown & Day, 1981, 21%; Heide & John, 1988, 25%; Hunt & Nevin, 1974, 26%; Robbins, Speh, & Mayer, 1982, 26%; Rosenberg & Stern, 1971, 35%; Schul, Little, & Pride, 1985, 33%; Schul, Pride, & Little, 1983, 33%); and a high response rate (Dwyer & Oh, 1987, 69%; 1988, 71%; Frazier, 1983b, 46.1%; Frazier, Gill, & Kale, 1989, 100%; Frazier & Summers, 1984, 46%; Lusch, 1976a, 1976b, 49.4%; Skinner & Guiltinar, 1985, 88.8%). Nonrespondents were not analyzed; however, an assessment of the study’s external validity was ascertained through comparison of the number of respondents who are members of the MADA and the population with data obtained from the MADA’s records. Three dealer demographic variables were considered: (1) type of franchise, (2) franchise orientation, and (3) association membership. 131 Table 4.14 shows that in terms of the type of franchise, all three groups were overrepresented in the sample. With respect to franchise orientation, the percentage of domestic and foreign franchises in the survey group was slightly less than their representation in the population, on the one hand. On the other hand, the percentage of respondents with domestic-foreign franchises was greater than that in the population of MADA members. There was a small difference in the percentage of MADA members in the population (84.3%) and in the survey group (88.5%). These results are not so discrepant to conclude that the survey group is vastly dissimilar relative to the population. As a result, it was decided that the data were "clean" enough to proceed with testing the research hypotheses in this study. Table 4.14.--Percentage comparison of the sample and population on selected demographic variables. Variable Sample3 Population^ Type of Franchise General Motors 43.6 39.3 Ford 28.8 22.6 Chrysler-Jeep-Eagle 25.9 22.3 Imports 26.3 10.4 Franchise Orientation Domestic 75.7 82.7 Foreign 8.2 10.5 Domestic/foreign 9.1 6.8 MADA Members 88.5 84.3 aN = 243 survey respondents. bN = 780 (MADA members). CHAPTER V TEST OF HYPOTHESES The purpose of this chapter is fivefold. First, the purpose of the theoretical framework presented in Chapter II is briefly revisited. Second, a discussion of the results of the statistical tests of the hypotheses is presented. Third, a discussion of the magnitude of treatment effects for the full models and reduced models is presented. Fourth, a discussion of the power of the statistical tests used to examine the hypotheses is presented. Finally, a discussion is presented on the results of tests for the effect of several moderator variables on the influence of the experimental treatments on the dependent variable. Theoretical Framework Revisited A theoretical framework was proposed in Chapter II as a paradigm to begin broader conceptual and empirical investigations of the effect of channel conflict on channel performance. It is an approach that might offer some explanations for the disparity in the empirical results in the channel literature on the form of the relationship between conflict and performance. Yet no attempt is made to develop a model to explain what has been found concerning the relationship between these two constructs. Given the proposed 132 133 theoretical framework, a series of hypotheses based on the literature review were presented. These hypotheses address the issue of whether there are other factors that interact with conflict that yield a stronger impact on channel performance rather than conflict taken alone. This section is devoted to discussing the results of an analytical test of the hypotheses. The analytical method most suited for this research is multivariate analysis of variance (MANOVA). Recall that there are four derived dependent measures (expected value impact on total new car sales volume, trade area market share, profits, and a summed performance measure) and three independent variables--conflict, involvement, and compliance. Test of the Full Effect Model A test of the full model--three-way interactions--involved examining the effect of conflict-involvement-compliance interaction on performance impact for new car sales volume (NCSV), trade area market share (TAMS), profits, and summed performance. The hypotheses stated in Chapter II indicate that the main effects of conflict, involvement, and compliance would be insignificant in a test of the full model. To test these hypotheses, the univariate approach to MANOVA was used to examine the full model, which contains all main effects, as well as two-way and three-way interaction effects. With the univariate approach to MANOVA, each response is treated as a case, yielding eight cases per respondent in this experiment. Each of the four performance measures was analyzed separately for the full model to detect whether the effects 134 of the treatment and their interactions are consistent across the four measures. The results of the analysis are shown in Tables 5.1 through 5.4. As shown in Table 5.1, when new car sales volume impact was the dependent variable, the conflict-involvement-compliance interaction effect is significant (F = 6.57, p = .01). However, the main effects of conflict, involvement, and compliance, as well as their two-way interactions, are also significant (F > 5.0, p < .02, respectively). These findings support the hypothesis that the conflict-involvement-compliance interaction has a significant effect on performance (Hl). Also, these findings provide tentative evidence that a broader approach to the study of conflict and performance is warranted. Table 5.1.--MANOVA results of the full model3 (dependent variable « new car sales volume). Source df SS MS F PR>F Conflict 1 526.11 526.11 271.69 .000 Involvement 1 349.26 349.26 180.36 .000 Compliance 1 11.45 11.45 5.91 .015 Conflict x involvement 1 2480.85 2480.85 1281.16 .000 Conflict x compliance 1 11.23 11.23 5.80 .016 Involvement x compliance 1 42.60 42.60 22.00 .000 Conflict x involvement x 1 12.72 12.72 6.57 .010 compli ance Error* 1936 3784.90 1.94 Total 1943 7219.20 aUnivariate approach used, ktfithin-cel 1s error. 135 When trade area market share (TAMS) impact was employed for performance, the conflict x involvement x compliance interaction effect is statistically significant. Table 5.2 shows that conflict, involvement, and compliance effects are significant, as well as their two-way interactions. More important, the three-way interaction effect of conflict, involvement, and compliance is significant (F = 7.97, p < .01). These findings support Hl. Table 5.2.--MANOVA results of the full model3 (dependent variable = market share). Source df SS MS F PR>F Conflict 1 158.47 158.47 90.11 .000 Involvement 1 11.00 11.00 6.26 .012 Compliance 1 16.71 16.71 9.50 .002 Conflict x involvement 1 1334.64 1334.64 758.88 .000 Conflict x compliance 1 20.59 20.59 11.71 .001 Involvement x compliance 1 27.72 27.72 15.76 .000 Conflict x involvement x 1 14.02 14.02 7.97 .005 compliance Error1’ 1936 3404.82 1.76 Total 1943 4987.97 aUnivariate approach used. ^Within-cells error. Performance impact was not measured only by NCSV and TAMS but also by profits. The MANOVA results for a test of the full model are shown in Table 5.3. Table 5.3 reveals that the three-way 136 interaction effect of conf1ict-involvement-compliance is significant (F = 13.44, p < .01). This finding supports Hl. Table 5.3.--MANOVA results of the full modela (dependent variable = profits). Source df SS MS F PR>F Conflict 1 2926.03 2926.03 189.92 .000 Involvement 1 283.07 283.07 18.37 .000 Compli ance 1 .01 .01 .00 .975 Conflict x involvement 1 15613.59 15613.59 1013.45 .000 Conflict x compliance 1 45.54 45.54 2.96 .086 Involvement x compliance 1 271.84 271.84 17.64 .000 Conflict x involvement x 1 207.03 207.03 13.44 .000 compliance Errorb 1936 29826.79 15.41 Total 1943 19347.11 aUnivariate approach used. bWithin-cells error. A fourth performance impact measure was developed by averaging respondents’ scores on NCSV, TAMS, and profits. This procedure yielded an expected averaged summed performance impact measure. Table 5.4 shows the results of the univariate approach to MANOVA to test the full effect model when this fourth performance measure was used. As Table 5.4 reveals, the compliance effect on performance is insignificant (F = .01, p > .90). Thus, H3 is supported by these findings. Again, the conflict-involvement-compliance effect is significant (F = 13.44, p < .01). This finding supports Hl, which is the primary contention of the study. Finally, conflict and 137 involvement effects as well as their two-way interaction are significant although they are uninterpretable. Table 5.4.--MANOVA results of the full modela (dependent variable = overall performance). Source df SS MS F PR>F Conflict 1 2926.03 2926.03 189.92 .000 Involvement 1 283.07 283.07 18.37 .000 Compliance 1 .01 .01 .00 .975 Conflict x involvement 1 15613.59 15613.59 1013.45 .000 Conflict x compliance 1 45.54 45.54 2.96 .086 Involvement x compliance 1 271.84 271.84 17.64 .000 Conflict x involvement x 1 207.03 207.03 13.44 .000 compliance Errorb 1936 29826.79 15.41 Total 1943 49173.90 aUnivariate approach used. ^Within-cells error. In summary, the results of a test of the full model lend support to the hypothesis of a significant three-way interaction effect of conflict, involvement, and compliance. This hypothesis is supported across all four performance impact measures. While the main effects and two-way interaction effects were also significant, they are not interpretable in a model that has significant three-way and higher-order interaction effects. 138 Test of Simple, Simple Models (Two-Way Interaction Effects) Reduced models were tested using the univariate approach to MANOVA. The purpose was to examine the two-way interaction effects of conflict, involvement, and compliance on performance as perceived by auto dealers. Given that the results from a test of the full model were mostly consistent across the four performance measures, only new car sales volume (NCSV) was used in the analysis of two-way interaction effect models. As further evidence of the consistency of the results across the performance measures, selected two-way interaction effect analyses were performed using the profit performance measure as the dependent measure. The results were mostly identical to the results, to be presented, found when NCSV was the dependent variable. When NCSV was employed for performance, the two-way interaction effects of conflict, involvement, and compliance are significant except for two cases (see Tables 5.8 and 5.10). Table 5.5 shows that the conflict x involvement interaction effect is significant for high compliance conflict resolution behaviors (F = 993.65, p < .01). Also, Table 5.6 reveals that for low compliance conflict resolution behaviors, the conflict x involvement interaction effect is also significant (F = 438.30, p < .01). These findings support H2. They indicate that when the effect of compliance is neutralized, the interaction effect of conflict and involvement has a significant impact on performance. 139 Table 5.5.--MANOVA results of two-way interaction model, adjusting for high compliance (dependent variable = new car sales volume). Source df SS MS F PR>F Conflict 1 345.53 345.53 241.03 .000 Involvement 1 73.95 73.95 51.59 .000 Conflict x involvement 1 1424.42 1424.42 993.65 .000 Errora 968 1387.65 1.43 Total 971 3231.55 aWithin-cells error. Table 5.6.--MANOVA results of two-way interaction model, adjusting for low compliance (dependent variable = new car sales volume). Source df SS MS F PR>F Conf1ict 1 191.80 191.80 78.63 .000 Involvement 1 317.91 317.91 130.33 .000 Conflict x involvement 1 1069.14 1069.14 438.30 .000 Errora 968 2361.24 2.44 Total 971 3940.09 aWithin-cells error. The interaction effect of involvement x compliance on performance is significant (F = 38.85, p < .01) when conflict is high between dealers and their suppliers (see Table 5.7). On the other hand, Table 5.8 shows that when conflict is low, the interaction effect of involvement x compliance is insignificant (F = 1.71, p = .191). These findings lend support to the hypothesis (H3) 140 of a significant involvement x compliance effect on performance when conflict between manufacturers and dealers is high. Yet this hypothesis is not supported when conflict is low. Table 5.7.--MANOVA results of two-way interaction model, adjusting for high conflict (dependent variable = new car sales volume). Source df SS MS F PR>F Involvement 1 484.22 484.22 369.26 .000 Compliance 1 .00 .00 .00 .984 Involvement x compliance 1 50.94 50.94 38.85 .000 Error3 968 1269.25 Total 971 1804.41 aWithin-cells error. Table 5.8.--MANOVA results of two-way interaction model, adjusting for low conflict (dependent variable = new car sales volume). Source df SS MS F PR>F Involvement 1 2345.88 2345.88 915.82 .000 Compli ance 1 22.68 22.68 8.85 .003 Involvement x compliance 1 4.38 4.38 1.71 .191 Error3 968 2479.55 2.56 Total 971 4852.49 aWithin-cells error. 141 The analysis results of the interaction effect of conflict x compliance for low and high involvement are shown in Tables 5.9 and 5.10. On the one hand, when there is low involvement in conflictual issues between manufacturers and dealers as perceived by dealers, the conflict x compliance interaction effect on performance is significant (F = 8.80, p < .01) (see Table 5.9). Yet, on the other hand, the conflict x compliance effect on performance is insignificant (F = .02, p « .887) when dealer involvement is high with conflictual issues between auto manufacturers and auto dealers (see Table 5.10). These findings lend support to the hypothesis (H4) of a significant interaction effect of conflict x compliance when dealer involvement is low with conflictual issues in channel relations. However, when there is low involvement as perceived by auto dealers, the hypothesis is not supported. Table 5.9.--MANOVA results of two-way interaction model, adjusting for low involvement (dependent variable = new car sales volume). Source df SS MS F PR>F Conflict 1 2645.92 2645.92 973.02 .000 Compliance 1 4.94 4.94 1.82 .178 Conflict x compliance 1 23.93 23.93 8.80 .003 Errora 968 2632.26 2.72 Total 971 5307.05 aWithin-cells error. 142 Table 5.10.--MANOVA results of two-way interaction model, adjusting for high involvement (dependent variable = new car sales volume). Source df SS MS F PR>F Conflict 1 361.03 361.03 312.97 .000 Compli ance 1 49.11 49.11 42.57 .000 Conflict x compliance 1 .02 .02 .02 .887 Errora 968 1116.64 1.15 Total 971 1526.80 aWithin-cells error. In summary, the results of two-way interactions among conflict, involvement, and compliance, except in two cases, provide further evidence that a broader approach to the conceptualization and empirical investigation of the conflict-performance relationship might be enriching. Test of Main Effects It was hypothesized and examined in the literature that conflict is expected to significantly affect performance as an explanatory variable on the one hand. Yet, on the other hand, its effect on performance is hypothesized to be insignificant in models that contain interaction effects as a fundamental premise of this study. Furthermore, involvement and compliance are proposed in this study as additional factors that significantly affect channel member performance. However, the interaction between them and their 143 interactions with conflict are proposed as a broader attempt to examine the relationship between conflict and performance. NCSV performance measure was the single performance item used to examine the main effects model for the same rationale as that used to examine the two-way interaction models. Table 5.11 shows that conflict has a significant effect on performance impact (F « 810.49, p < .01) when there is high compliance by dealers to resolve conflictual issues with auto manufacturers. Yet these issues elicit low involvement by dealers to resolve them. An examination of the results in Table 5.13 reveals that conflict has a significant effect on expected performance (F = 127.10, p < .01) for highly involved dealers with conflict issues who do not completely comply with their suppliers’ requests to resolve conflict. These findings support the results of previous studies (Brown, 1978; Lusch, 1976a; Pearson, 1973; Reve, 1977; Walker, 1970). Table 5.11.--MANOVA results of main effect model given high compliance-low involvement (dependent variable = new car sales volume). Source df SS MS F PR>F Conflict 1 1586.53 1586.53 810.49 .000 Error3 484 947.42 1.96 Total 485 2533.95 aWithin-cells error. 144 The results shown in Tables 5.12 and 5.14 reveal that conflict has a significant effect on expected performance (F - 201.66, p < .01) when dealers are highly involved in the conflict issue and they highly comply with their suppliers’ request or action to resolve the conflict. In addition, when there is low involvement in the conflict issue and low compliance by dealers to resolve the issue, conflict again has a significant effect on expected performance (F = 311.20, p < .01). Again, these findings support the results of previous studies of the effect of conflict on performance. Table 5.12.--MANOVA results of main effect model given high involvement-high compliance (dependent variable = new car sales volume). Source df SS MS F PR>F Conflict 1 183.42 183.42 201.66 .000 Errora 484 440.23 .91 Total 485 623.65 aWithin-cells error. Table 5.13.--MANOVA results of main effect model given high involvement-low compliance (dependent variable = new car sales volume). Source df SS MS F PR>F Conflict 1 177.63 177.63 127.10 .000 Errora 484 676.41 1.40 Total 485 854.04 aWithin-cells error. 145 Table 5.14.--MANOVA results of main effect model given low involvement-low compliance (dependent variable = new car sales volume). Source df SS MS F PR>F Conf1ict 1 1083.31 1083.31 311.20 .000 Errora 484 1684.84 3.48 Total 485 2768.15 aWithin-cells error. When the involvement effect model was tested, the results were mostly consistent with the results of the conflict effect model. Table 5.15 shows that involvement has a significant effect on expected performance (F = 508.73, p < .01) when there is high conflict between auto manufacturers and dealers yet dealers highly comply with their suppliers’ requests on resolution behaviors to resolve the conflict. In addition, involvement has a significant effect on expected performance for high conflict-low compliance dealers (F = 61.82, p < .01), for low conflict-high compliance dealers (F = 528.32, p < .01), and for low conflict-low compliance dealers (F = 413.03, p < .01) (see Tables 5.16, 5.17, and 5.18). These findings, however, would not support a hypothesis that involvement has an insignificant effect on performance. The results also show that involvement has the greatest significant effect on expected performance for low conflict-high compliance situations and the least significant effect on expected performance for high 146 conflict-low compliance situations using the size of the F-ratio criterion. Table 5.15.--MANOVA results of main effect model given high conflict-high compliance (dependent variable - new car sales volume). Source df SS MS F PR>F Involvement 1 424.63 424.63 508.73 .000 Errora 484 403.99 .83 Total 485 828.62 aWithin-cells error. Table 5.16.--MANOVA results of main effect model given high conflict-low compliance (dependent variable - new car sales volume). Source df SS MS F PR>F Involvement 1 110.53 110.53 61.82 .000 Errora 484 865.36 1.79 Total 485 966.89 aWithin-cells error. 147 Table 5.17.--MANOVA results of main effect model given low conflict- high compliance (dependent variable = new car sales volume). Source df SS MS F PR>F Involvement 1 1073.74 1073.74 528.32 .000 Error3 484 983.67 2.03 Total 485 2057.41 aWithin-cells error. Table 5.18.--MANOVA results of main effect model given low conflict- low compliance (dependent variable = new car sales volume). Source df SS MS F PR>F Involvement 1 1276.53 1276.53 413.03 .000 Error3 484 1495.88 3.09 Total 485 2772.41 aWithin-cells error. When the compliance effect model was tested, the results reveal that compliance has a significant effect on expected performance except for one case. Table 5.22 shows that when there is low perceived conflict and low dealer involvement with a conflictual issue, compliance has an insignificant effect on expected performance (F = .79, p « .374). On the other hand, Tables 5.19, 5.20, and 5.21 reveal that compliance has a significant effect on expected performance when conflict and involvement are high (F - 148 15.23, p < .01), when conflict is high and involvement is low (F = 26.94, p < .01), and when conflict is low and involvement is high (F = 37.67, p < .01). Also, closer analysis of the results reveals that compliance has the least significant effect on expected performance when conflict and involvement are high and the greatest significant effect when conflict is low yet involvement in the conflictual issue is high. Table 5.19.--MANOVA results of main effect model given high conflict-high involvement (dependent variable = new car sales volume). Source df SS MS F PR>F Compliance 1 25.63 25.63 15.23 .000 Errora 484 814.70 1.68 Total 485 840.33 aWithin-cells error. Table 5.20.--MANOVA results of main effect model given high conflict-low involvement (dependent variable = new car sales volume). Source df SS MS F PR>F Compli ance 1 25.31 25.31 26.94 .000 Errora 484 454.65 .94 Total 485 479.96 aWithin-cells error. 149 Table 5.21.--MANOVA results of main effect model given low conflict- high involvement (dependent variable - new car sales volume). Source df SS MS F PR>F Compli ance 1 23.50 23.50 37.67 .000 Errora 484 301.94 .62 Total 485 325.44 aWithin-cells error. Table 5.22.--MANOVA results of main effect model given low conflict- low involvement (dependent variable = new car sales volume). Source df SS MS F PR>F Compliance 1 3.56 3.56 .79 .374 Errora 484 2177.61 4.50 Total 485 2181.17 aWithin-cells error. In summary, the main effects of conflict, involvement, and compliance on performance are significant as expected when their interaction effects are controlled for except for one case. Furthermore, using the size of the F-ratio criterion, the results show that conflict has the greatest significant effect on performance, while compliance has the least significant effect on performance. Also, the findings would support a hypothesis of a F 150 significant compliance effect except in the case of the compliance effect when conflict and involvement are low. Relative Magnitude of Experimental Effects and Power The analytical results of the data did not support the hypotheses of significant main effects for conflict, involvement, and compliance in the full models tested. Yet the hypotheses of significant three-way interaction effects for the four performance measures were supported. Most of the two-way interaction effects were significant in the simple, simple effects models, and the main effects were significant in a test of each main effect model except for compliance. However, it is noteworthy that in this balanced repeated measure design, the number of observations is 243 for each of the eight cells or a total of 1,944 cases. The F-tests used to test the significance of the full effect, simple, simple effect, and main effect models can be overly sensitive to a large number of observations used. An examination of the "relative magnitude" of treatment effects on performance was the next step in the analytical process. The most popular index used to measure experimental effect is omega squared (u>^). it is responsive to the strength of the association between experimental manipulations and changes in behavior and independent of the sample size (Keppel, 1982). Omega squared (w^) reflects the proportional amount of the total population variance "accounted for" by experimental treatments. This index reflects the proportion of "explained variance" by 151 manipulating the treatment in an experiment (Cohen, 1977; Keppel, 1982; Sawyer & Ball, 1981; Winer, 1971). A biased, but consistent estimator for “2 is ft2 (yjner, 1971, p. 429). This estimator was determined for the main effects and interaction effects in Models I, II, and III in this study. Table 5.23 shows the estimated magnitude of effect for conflict, involvement, and compliance for the full effect model across the four performance measures used. The proportion of explained variance for the three-way interaction effect ranged from a low of .0015 when NCSV was used for performance to a high of .0108 for profits. In each case, less than 1% of the total variance in performance is accounted for by this three-way interaction effect. In addition, the two-way interaction effect of conflict and involvement consistently accounted for the largest proportion of the explained variance across the four performance measures--.3450, .2671, .2884, and .3170, respectively. The percentage of the total variance that is accounted for by all effects is greatest when NCSV was used for performance (47.61%) and lowest when TAMS was used (31.48%). While the values of &2 presented in Table 5.23 are relatively small, large values are unlikely to be observed in most behavioral and social science research because of the relatively large contribution of error variance to total variance (Keppel, 1982, p. 92). For behavioral and social science research, Cohen (1977) suggests that a "large" effect in an experiment yields a value of .15 or greater, a "medium" effect is .06, and a "small" effect is 152 .01 (pp. 284-288). As shown in Table 5.23, the three-way interaction of conflict x involvement x compliance produces a less than small effect size by this criterion except when profits was used for performance--!.08% explained variance. Among the effects produced by the full model, only conflict x involvement interaction reveals a "large" effect size across the four performance measures. Table 5.23.--Magnitude of effect on performance (full models). Estimated Omega Squared (<4)a Effect NCSVb TAMSb Profits Summed Perf. Conflict .0730 .0314 .0228 .0592 Involvement .0483 .0019 .0070 .0054 Compliance .0013 .0030 .0077 .0000 Conflict x involvement .3450 .2671 .2884 .3170 Conflict x compliance .0013 .0038 .0000c .0006 Involvement x compliance .0057 .0025 .0090 .0052 Conflict x involvement x .0015 .0052 .0108 .0039 compliance All effects .4761 .3148 .3456 .3913 aThe numerator for $ is p - 1 or (q - 1) or (r - 1) (F^- 1) MSE for each main effect; (p - 1) (q - 1) (F-j j ik jk - 1) MSEJfor each two-way interaction effect; and (p - 1) -’1) (r - 1) (Fjjk - 1) MSE for the three-way interaction effect. The denominator isJ(p - 1) (Fi - 1) MSE + (q - 1) (F-j - 1) MSE + (v - 1) (Fk - 1) MSE + (p - 1) (q - 1) (Fjj - 1) MSE + (J -1) (r - 1) (Fjk - 1J MSE + (q - 1) (r * 1) (Fik - 1)JMSE + (p - 1) (q - 1) (r - 1) (?ijk - 1) MSE + MSE, which contains all effects (Winer, 1971). bNCSV = new car sales volume; TAMS = trade area market share. cfi]3 = -.0003; fif = -.0003. 153 The "significance" of jS assessed by the regular F-test--a significant F implies that $ is significantly greater than zero, as well. And yet, the significance of the F-test is affected by small sample sizes and lower power (Keppel, 1982, p. 92). The results presented above revealed significant F-ratios for the three- way interaction effects for each performance measure used (see Tables 5.1 through 5.4). The relative magnitude of treatment effect for the simple, simple models (two-way interaction effects) is shown in Table 5.24. Under high and low conflict, there is a "small" effect (.0275) and less than a "small" effect (.0004) for involvement x compliance interaction on performance. The proportion of the total variance in performance that is accounted for by the conflict x compliance interaction effect is .0007 when involvement is high and .0040 with low involvement in a conflict situation. These findings reveal a less than "small" effect. The strongest effect is produced by the conflict x involvement interaction. The percentage of explained variance of the total variance in performance is .4936 for high compliance and .2706 for low compliance. Using Cohen’s (1977) criteria, a "large" effect size is produced by this interaction term. Table 5.24 also shows that the main effects of conflict and involvement produce "large" effect size, .2352 and .4978 for high and low involvement and .2674 and .4827 for high and low conflict, respectively. The proportion of explained variance in performance when "all effects" are considered in each model ranges from .2942 (high conflict) to .5696 (high compliance). 154 Table 5.24.--Magnitude of effect on NCSV (simple, simple models). 2 a Estimated Omega Squared (&i)d Effect HCb LCb HIb LIb HCOb LCOb Conflict .2352 .4978 .1063 .0480 Involvement .2674 .4827 .0224 .0800 Compliance .0000c .0041 .0314 .0009 Conflict x involvement .4396 .2706 Conflict x compliance .0007 .0040 Involvement x .0275 .0004 compliance All effects .2942 .4872 .2662 .5022 .5696 .3986 aThe numerator for is (p - 1) or (q - 1) or (r - 1) (F^k - 1) MSE for main effect; (p - 1) (q - 1) or (p - 1) (r - 1) or (q - 1) (r - 1) (Fji ik jk ‘ 1) MSE for two-way interaction effects; and (p - 1) (q - i)’(r - i) (F44l - 1) MSE for three-way interaction. The denominator is (p - 1) (Fi - 1) MSE + (q - 1) (F^ - 1) MSE + (r - 1) (Fk - 1) MSE + (p - 1) (q - 1) (F^j - 1) MSE + (tf - 1) (r - 1) (Fik - 1) MSE + (q - 1) (r - 1) (Fik - 1)JMSE + MSE, which contains all effects (Winer, 1971). bHC = high conflict, LC = low conflict, HI = high involvement, LI = low involvement, HCO = high compliance, LCO = low compliance. = -.0007. The results above reveal that the F-test for the model examining the conflict x involvement interaction effect under high and low compliance is significant for small significance levels. Admittedly, power is a function of the significance level, the precision of the estimate (sample size), and effect size (Sawyer & Ball, 1981, p. 1). Thus, the power of the F-test for the effect of conflict x involvement interaction proves to be high given that the findings reveal a "large" effect size; a large sample was used, and 155 the significance level can be large as well. The F-test for the model examining the involvement x compliance interaction effect is insignificant at large significance levels for low conflict (p = .191) and significant at small significance levels under high conflict (p < .01). The power of the F-test can be lower given "small" effect sizes and small significance levels despite a large sample size. The model testing the conflict x compliance interaction effect is significant under low involvement (F - 8.80, p = .003) and insignificant under high involvement (F = .02, p = .887). The power of the F-test may be low given "small" effect sizes. Table 5.25 shows the relative magnitude of treatment effects for conflict, involvement, and compliance for the main effect models. The proportion of explained variance in performance when conflict is the effect ranges between .6249 for low involvement-high compliance and .2922 for high conflict-high compliance. When involvement is the treatment effect, the percentage of explained variance in performance ranges between .1112 under high conflict-low compliance conditions to .5204 under low conflict-high compliance conditions. When compliance is the main effect, the results reveal that the proportion of explained variance ranges between .0284 under high conflict-high involvement and .0702 under low conflict-high involvement conditions. Also, the results reveal that conflict has no effect on performance under conditions of low conflict-low involvement situations. These findings reveal that conflict and involvement produce "large" effects on performance; the percentage 156 of explained variance is .15 or greater. These findings also reveal that compliance produces a "small" to "moderate" effect on performance; the percentage of explained variance is between .02 and .07. Table 5.25.--Magnitude of effects on NCSV (main effect models). 2 Estimated Omega Squared (fi>j)a Conflict Involvement Compliance HC-HIb .0284 HC-LIb .0507 LC-HIb .0702 LC-LIb .0000c HI-HC0j> .2922 HI-LCOb .2060 LI-HC0b .6249 LI-LCOb .3896 HC-HC0b .5104 HC-LCOb .1112 LC-HCOb .5204 LC-LCOb .4588 aThe numerator for G? is (p - 1) (Fj - 1) MSE or (q - 1) (Fj - 1) MSE or (r - 1) (Fl - 1) MSE for each main effect. The denominator is [(p - 1) (Fj - 1) MSE or (q - 1) (Fj - 1) MSE or (r - 1) (Fk - 1)] + MSE, which contains all effects (Winer, 1971). bHC, HI, HCO = high conflict, high involvement, high compliance; LC, LI, LCO = low conflict, low involvement, low compliance. cfi>3 = -.0004. The results of the F-test on the significance of each main effect reveal that conflict, involvement, and compliance (except 157 under high conflict-low involvement) are significant at small significance levels (see Tables 5.17 through 5.22). Given the "small" to "large" effect (percentage of explained variance) and large sample sizes used, the power of the F-test on these main effects models might be high. Test for the Effect of Moderator Variables on the Experimental Treatment Effects MANOCOVA was used to determine if the potential influence of several extraneous or moderator variables (e.g., dealership size, experience, competition, power) can be removed or partialled out ex post facto from performance (Hair, Anderson, Tatham, & Grablowsky, 1984). The primary purpose for using MANOCOVA is to address the question, "Will the interaction effects (three-way and two-way) of conflict, involvement, and compliance remain significant after the influence of the four dealer demographic and psychological characteristics (covariates) on the dependent variable is removed?" Due to time restraints, power was not analyzed. Table 5.26 shows the MANOCOVA results for the full factorial models for each of the four dependent measures. In each case, the conflict x involvement x compliance interaction effect remained statistically significant even when the effects of dealership size, experience, and competition were removed. Note that dealership size and experience had a very significant influence on all four measures of performance--NCSV, TAMS, profits, and overall performance. Competition had an insignificant influence on each performance measure. In addition, removal of the effects of the covariates on 158 the performance measures slightly increased the conflict x involvement x compliance interaction effects by reducing the size of the residual sum of squares. Table 5.26.--MANOCOVA results in the full effect models.3 Covariates NCSV TAMS Profits Overall Perf. Competition B - .0303 - .0317 - .0376 - .0340 (.0049) (.0046) (.0040) (.0137) T-value -1.2591 -1.3149 -1.5600 -1.4106 PR > |T| .208 .189 .119 .159 Experience B - .0043 - .0587 - .0660 - .0623 (.0017) (.0016) (.0014) (.0048) T-value -2.5223 -2.5889 -2.9122 -2.7468 PR > |T| .012 .010 .004 .006 Size B .0540 .0607 .0532 .0577 (.0004) (.0004) (.0003) (.0001) T-value 2.2743 2.5190 2.2101 2.3941 PR > |T| .023 .012 .027 .017 aValues in parentheses represent the standard error of B. When MANOCOVA was applied to the simple, simple models (two-way interaction effects), the results remained unchanged when the effects of dealership size, experience, and competition on performance were partialled out. That is, the joint effects of conflict x involvement remained significant for high and low compliance; the conflict x compliance effect remained significant for low involvement and insignificant for high involvement; the 159 involvement x compliance effect remained significant under high conflict and insignificant under low conflict. There was no consistent pattern to the effect of the covariates on performance in the two-way interaction models. For example, under low conflict, dealership size, experience, and competition had an insignificant effect on performance, whereas size and experience had a significant effect under high conflict in the models testing the effect of the involvement x compliance interaction. When the conflict x compliance effect was evaluated removing the influence of the covariates, only experience had a significant effect under low involvement. Under high involvement, dealership size had a significant effect whereas competition and experience had an insignificant effect. For low compliance, the three covariates had an insignificant effect on performance in the model testing the effect of the conflict x involvement interaction on the one hand. On the other hand, dealership size had a significant effect whereas competition and experience had an insignificant effect in the model testing the effect of the conflict x involvement interaction under the experimental condition of high compliance. Test for Parallel Regression Slopes When analyzing experimental data using MANOCOVA, it is assumed that in a general linear model, the treatment effects and the covariates have homogeneous (parallel) slopes. That is, all factor 160 x covariate interactions are equal to zero (Hair, Anderson, Tatham, & Grablowsky, 1984; Keppel, 1982; Winer, 1971). The assumption of parallel slopes was tested in this study for the full factor models and simple, simple models. All factor x covariate interactions were insignificant in the full factorial models except in a few cases. The compliance x experience interaction was significant (F ■ 3.85, p = .05) in the model when profit was used for performance and when TAMS was used for performance (F = 2.93, p = .087). These findings do not support the hypothesis of homogeneous regression slopes. When the assumption of parallel slopes was tested with the simple, simple (two-way interaction) models, most factor x covariate interactions were insignificant except for a few cases. New car sales volume (NCSV) was the sole measure used for performance in the simple, simple models given the similarity of results across the four measures of performance used in this study. In the model depicting the effect of the involvement x compliance interaction on performance, the results revealed that involvement x experience was significant (F = 3.36, p = .067) when conflict was low. Under conditions of high conflict, involvement x size (F - 5.09, p = .024) and compliance x experience (F = 3.51, p - .061) were significant. In the model testing the effect of the conflict x compliance interaction, all factor x covariate interactions were insignificant under conditions of low involvement on the one hand. On the other hand, when involvement was high, conflict x size (F » 3.36, p = .067) and conflict x experience (F - 2.62, p = .106) were 161 significant, although the latter was only marginally significant. The results also revealed that there were no significant factor x covariate interactions in the model depicting the effect of the conflict x involvement interaction under conditions of high and low compliance. Thus, these findings support the assumption of homogeneous slopes for the experimental treatments and covariates with the exception of a few cases. Fit of Separate Regression Slopes When a test of factor x covariate interactions reveals significant results, the next step in the analysis of the data is to fit the separate regression slopes for different treatment levels of the factors. In this study, the analysis was limited to the full effect models because in the simple, simple models there were too many empty cells. When TAMS and profits were used for performance, the results revealed that experience with high and low compliance was significant--F « 3.60, p = .027 and F = 4.54, p = .011, respectively. On the other hand, compliance was insignificant when TAMS was used for performance and significant (F = 16.77, p < .001) when profit was used. CHAPTER VI CONCLUSIONS Chapters IV and V discussed the results of the data analysis in relation to the theoretical framework proposed to examine conflict and performance and the hypotheses developed. This final chapter begins with a discussion of the results. It is followed by a discussion of the limitations of the study. Next, the theoretical contributions and managerial implications of the findings are discussed. Finally, a summary of the research is provided. Discussion of Results The first set of hypotheses was established to investigate the significance of conflict x involvement x compliance interaction effect on performance. The purpose was to show that a broader approach may be needed to examine the impact of channel conflict on performance. To summarize the results of the data analysis, Hl (significant three-way interaction) was supported when NCSV, TAMS, profits, and summed performance were used for performance (see Tables 5.1 through 5.4). However, when the relative magnitude of effect of conflict x involvement x compliance interaction was determined, the results revealed only a "small" effect. That is, the proportion of explained variance of the total variance in 162 163 performance was approximately 1% and less than 1%. This was consistent for each measure of performance impact (see Table 5.23). It is suspected that the conflict x involvement x compliance interaction may be significant, in part, because of the large number of observations in each cell. It is known that a significant F-ratio can result from a large sample size or a large treatment effect size or both, among other factors (Keppel, 1982, p. 89). The findings in relation to Hl appear to support the notion that channel analysts must begin to study the relationship between conflict and performance in a much broader context conceptually and empirically. The nature of the conflictual issue, the level of channel member involvement with the issue, and the resolution responses chosen in terms of the level of dealer compliance with the manufacturer’s actions or requests may be more closely related to performance outcomes than each taken separately. These findings are exploratory in nature, for they shed some light on the obscure premise that channel conflict affects channel performance. Given that the conflict x involvement x compliance interaction effect is statistically significant, the joint effects--conflict x involvement, conflict x compliance, involvement x compliance--and the main effects are uninterpretable even though they have significant F-ratios. Hypotheses 2, 3, and 4 were established to investigate whether the joint effects would have a significant impact on performance. The results were calculated for only one of the four performance measures given the consistency in results across all four measures. 164 When new car sales volume (NCSV) was used for performance, the joint effects were statistically significant except for two cases. These findings support H2 and partially support H3 and H4. The conflict x involvement interaction was significant for high and low compliance. Also, this joint effect produced a "large" effect on performance in terms of the percentage of explained variance. This means that dealers perceive that they can make a significant impact on their performance outcome by either doing what the manufacturer requests of them solely to resolve conflictual issues (high compliance) and/or doing what they feel will work for their franchise to resolve the conflict (low compliance). The joint effect of involvement and compliance was perceived to make a significant impact on performance when conflict between manufacturers and dealers is high, on the one hand. On the other hand, the results reveal that this effect is perceived to make an insignificant impact on performance when conflict is low (see Tables 5.5 and 5.10). These findings support H3 for high-conflict situations and do not support H3 for low-conflict situations. The conflict x compliance interaction, on the one hand, was found to have a significant effect on performance when dealer involvement is low and a insignificant effect when involvement is high, on the other hand. These findings provide further evidence that conflict may have a "nonlinear" effect on performance. In other words, in the context of an experimental design, the total variance in performance is not only composed of error variance and treatment 165 variance but also variance attributable to the interaction between (among) conflict, involvement, compliance, and, possibly, other behavioral factors. These findings provide evidence that there are other behavioral factors that are related to conflict which support the notion that a "nonlinear" functional relationship exists between conflict and performance. The conflict x involvement x compliance interaction significantly affects channel member performance according to the results of the data analysis. This indicates that the relationship between conflict and performance might not be linear as previous conceptualizations and empirical investigations have proposed (Brown, 1978; Lusch, 1976a; Pearson, 1973; Rosenbloom, 1973). For example, Lusch (1976a) hypothesized and found a significant negative relationship between channel conflict and dealer operating performance. In the same study, the hypotheses of a significant positive and curvilinear relationship between the two constructs were not supported. Brown (1978, 1980) found only weak evidence of a significant negative relationship between conflict and performance. Thus, any model that asserts a "linear" relationship between these two constructs will represent a misspecification of the functional form of the relationship. Several extraneous independent variables (covariates) were analyzed to partial out or remove their influence from performance. It was thought that dealership size, experience, competition, and power might contaminate the effects of the treatment interactions by rendering them statistically insignificant once they were taken into 166 account. In the case of the full factor models, the conflict x involvement x compliance interaction effect remained statistically significant when the effect of size, experience, and competition was removed from performance. The results were consistent when NCSV, TAMS, profits, and summed performance were each used for performance. There was a unique pattern to the effect of the covariates on performance. For each measure of performance, dealership size and experience had a statistically significant effect whereas competition was statistically insignificant. In the case of the joint effects of conflict, involvement, and compliance, the results revealed that after removing the effects of the three covariates, there were no changes in the conflict x involvement, conflict x compliance, and involvement x compliance interaction effects (statistically significant or insignificant), adjusting for high and low conditions of each factor. There were no consistent patterns as to significant or insignificant effects of size, experience, and competition on performance in the joint effect models. At least one covariate was statistically significant or insignificant except for conditions of low conflict (involvement x compliance effect) and low compliance (conflict x involvement effect), in which case all covariates were statistically insignificant. The assumption of parallel regression slopes in a covariance design was violated in a few cases when the factor x covariate interactions were tested for statistical significance. In the case 167 of the full factor model, the compliance x experience interaction was the only significant effect when profits and TAMS were used for performance. This indicates that dealer experience has a different effect on performance for high and low compliance. It is suspected that the longer a dealer has had a franchise(s), the greater the likelihood it will implement conflict-resolution behaviors that are mutually beneficial or self-sustaining because of its knowledge base from past supplier-dealer problems and resolutions and its greater range of resolution options. In the case of the simple, simple models testing the joint effects of conflict, involvement, and compliance, there was one consistent pattern in the significance testing of factor x covariate interaction. Dealership size and experience significantly interacted with one or more of the factors except for conditions of high and low compliance (conflict x involvement model) and low involvement (conflict x compliance model). It is suspected that dealership size and franchise longevity (experience) significantly affect dealers’ performance, although they did not significantly change the joint interaction effects of the treatments in this ; study. The one minor effect that the covariates had in this study was to reduce the size of the unexplained or residual error in the full and simple, simple models. The covariance analysis provided further support to the premise that a broader approach to the study of the relationship between conflict and performance is essential. Based on the findings in this study, it seems more appropriate to specify a model that hypothesizes a "nonlinear" relationship 168 between conflict and performance. The relationship may be specified as curvilinear--quadratic, or cubic or a higher-order curve. But channel analysts must begin to look beyond the linearity assumption if they are to begin to understand "how" conflict affects performance. There is some conceptual and empirical evidence that supports the results found in this study. Rosenbloom (1973) alludes to a possible nonlinear relationship between conflict and performance in his exposition of a "general curve," which is shown when there is no effect, a positive effect, and a negative effect of conflict on channel efficiency. Brown (1978, and later, 1980) found evidence of a curvilinear effect of conflict on performance manifested as a "U-shaped" and an "inverted U-shaped" function. Pearson (1973) cautioned that conflict and cooperation might not be "directly" related to operational performance, as measured by customer service levels and inventory turnover. Although there is evidence in the channel literature that conflict affects performance negatively (Brown, 1978; Lusch, 1976a; Pearson, 1973) and positively (Brown, 1978), there is stronger evidence that channel conflict has a nonlinear effect on performance. Limitations of the Study Since this research was conducted in an experimental setting, it has a limitation in external validity like any other experimental study. Unlike most experimental designs, the design used in this study is a repeated measure design where the same experimental units are measured for all factor-treatment levels. The generalizability 169 of the results is limited to a totally perceptual framework in which the conflict-performance assumption is empirically investigated. However, no attempts are made to generalize the findings in this study in a nonperceptual context. Every attempt was made to examine the unit of analysis (auto dealers) perceptions of "what they thought" about the effect of conflict situations on the manufacturer-dealer relationship. In addition, the emphasis was on what respondents thought was the effect of things that had happened or currently was happening on their performance. Also, the type of channel (franchised) and the type of product (new automobiles) used in this study are real phenomena. This renders the study’s results similar to other studies on channel issues more externally valid. Another limitation in this study is centered on measurement of the dependent variable--performance. In this research, respondents rated the "perceived impact" of conflict resolution responses on three performance measures--new car sales volume, trade area market share, and profits. Thus, the validity of the results is limited to a perceptual framework that evaluates the effect of channel conflict on channel performance. Like other studies on conflict and performance, this study suffers from the inability to tie episodes of conflict between channel members to "objective" measures of performance (individual or total). In addition, even though other studies (Brown, 1978; Lusch, 1976a; Pearson, 1973) have used objective performance measures, the results have been less than promising or insignificant when the two constructs are examined. 170 Until conflict can be tied to objective performance measures, research findings will have limited managerial implications yet provide some theoretical insights in uncovering the appropriate functional form of the effects of conflict in the channel. Another limitation of this study is that measurement of the constructs and manipulation of the treatments were done with respondents from one side of the channel dyad--retail automobile dealers--due to limited resources. However, this is not unlike other channel-related studies. The results found in this study may be different if channel dyads were the unit of analysis. Another limitation in this study is that the conflict­ resolution behaviors that were generated may not be an exhaustive list. Although the preliminary investigation and pretest of the questionnaire permitted a representative list of resolution responses per conflict situation, it appears that there are different resolution behaviors based on the size and type of franchise (domestic, foreign, domestic and foreign). For example, it appears that small franchises (sale volume units) spend more time and effort buying and selling used cars because of the greater propensity of new vehicle allocation problems. Another limitation in this study is that the research design was tested using only new car dealers in the state of Michigan. General characteristics of new car dealers in the Midwest and the U.S. may be different from Michigan auto dealers. Due to time and resource constraints and the unavailability of data on regional and 171 national dealer demographic and perceptual profiles, the study was confined to the automobile industry in Michigan. Another slight limitation in this study is that the compliance treatment levels were not subjected to the same level of rigorous pretest manipulation checks as were conflict and involvement. However, compliance is a special type of behavioral factor in that it centers on which resolution behaviors channel members chose to resolve specific episodes of conflict with their suppliers. In addition, compliance focuses on whether or not channel intermediaries comply with the manufacturer’s request solely or respond by implementing decisions that are mutually beneficial or are for their individual self-interest to resolve conflict. Given that the locus of power resides with auto manufacturers, dealers usually work with the manufacturers entirely or to a degree to resolve conflictual issues. Theoretical Contributions and Managerial Implications Theoretical contributions can be discussed from the findings in relation to the hypotheses. For Hypothesis 1, it was found that the conflict x involvement x compliance interaction significantly affected performance when NCSV, TAMS, and profits were used as alternative measures. Although some conceptual and empirical analyses have asserted that conflict has a nonlinear effect on performance, no previous research has investigated the nature of the relationship between the two constructs in the context of a within- subject, repeated measure, quasi-experimental design. This research 172 found these phenomena in an experimental context. These findings also provide some insights for theoretical model specification which represents the effect that channel conflict has on performance. In light of these findings, it seems appropriate that channel analysts begin to develop a much "broader conceptual framework" for empirical investigations of the behavioral ramifications of conflict’s effect on performance. In relation to Hypotheses 2, 3, and 4, it was found that the joint effects of conflict x involvement, conflict x compliance, and involvement x compliance have significant effects on performance except in a few cases. Specifically, it was found that conflict x involvement interaction significantly affects performance regardless of the level of dealer compliance with the manufacturer ’s request to resolve conflictual issues. In addition, it was found that involvement x compliance interaction significantly affects performance only when conflict is high. This finding provides some support for the proposition that compared to low perceived conflict situations, high-conflict situations lead dealers to implement conflict-resolution behaviors that have a significant impact on their performance. It was also found that conflict and compliance joint effect has a significant effect on performance only for low involvement conflict situations. This finding provides support for the proposition that compared to high involvement conflict issues, low involvement issues lead dealers to implement conflict-resolution responses that have significant effects on their performance. 173 Overall, these findings with respect to the joint effects also ; provide further evidence that a "nonlinear" model specification is the more appropriate characterization of the effect of conflict on performance. Other behavioral factors must be considered in any conceptual paradigm and empirical investigation of this channel \ phenomenon. । Managerially, this research indicates that conflict can be managed and that it is important to manage conflict in supplier- j dealer relationships instead of avoiding it as if it will "magically" disappear. Manufacturers in franchise channels of distribution must attempt to understand the perceptions of small franchisees with respect to their policies and programs if they are i to effectively exercise control in the channel. For example, two i \ problem areas in particular deserve quality attention by auto i suppliers. Dealers, particularly smaller ones, reveal that the t manufacturer’s policies on new vehicle allocation and the use of s customer satisfaction index (CSI) surveys are very sensitive to . dealers. There is a general perception among many dealers that auto ‘ manufacturers are insensitive to their inputs as far as adapting t their programs and policies to fit the dealer’s trade area and the ji nature of competition therein. if j In order for manufacturers to exercise control in the channel, I they must effectively manage conflict and understand that the effect i of conflict on their performance and that of franchisees is not necessarily a "direct" one. They must begin to understand dealers’ ^involvement level with conflictual issues and the process of 174 selecting and implementing resolution responses that are beyond the suppliers’ requests or expectations. This is essential in order to maintain or regain channel control and competitiveness. In addition, the findings in this research indicate that the majority of franchisees are willing to work with their supplier(s) in choosing response actions to resolve conflict. However, dealers appear more receptive to resolution responses that are mutually beneficial for themselves and their supplier(s). Franchisees in the channel must understand that various levels of involvement with conflict issues and a broader range of options to respond to conflict might provide a better understanding of the impact of these responses on their performance results. The severity of conflict resolution options may be inversely related to the range of options a dealer has to resolve conflict and affect performance. Each conflict situation offers a franchisee a different range of resolution response options. The complexity of the behavioral approach is situation specific. There are some factors that have a bearing on the conflict-performance relationship, which casts it in a difficult light in terms of understanding the scope of specific behavioral factors that are tied to it. A systematic approach to the study of conflict and performance is essential in order to gain a richer understanding of which specific behavioral factors render the relationship much broader than the conventional linearity assumption. 175 Several covariates--dealership size, experience, competition-- were found to affect dealer performance (real or perceived). However, they do not seem to alter the form of the relationship between conflict and performance. Future Research Directions In this study, it was found that conflict x involvement x compliance interaction has a significant effect on channel member performance. Based on this finding, it is suspected that channel conflict has a "nonlinear" relationship with performance. The experimental design used to test the hypotheses of interaction effects included only "within-subject" factor treatment variables. All respondents used in this design were exposed to all experimental conditions. Future research should investigate the effect of conflict x involvement x compliance interaction in a design that includes both between- and within-subjects independent variables. This should be done in repeated and nonrepeated measure experimental designs that include both between- and within-subject factor treatment independent variables. The full models and simple, simple models tested in this study produced significant conflict x involvement x compliance interaction effects and significant joint interaction effects of conflict x involvement, conflict x compliance, and involvement x compliance. As mentioned above, these findings provide evidence of a nonlinear relationship between conflict and performance as a more appropriate functional specification. Future research should investigate the 176 appropriateness of alternative nonlinear model specifications of the functional relation between these two constructs. For example, future research should evaluate the "trend" components of the main and interaction effects of conflict, involvement, and compliance. These trend components may consist of linear, quadratic, or cubic or higher-order dimensions. The purpose of isolating the trend components is to try to detect the "best-fitting curve" which demonstrates the functional form of the relationship between conflict and performance. The findings in this study indicate that the relationship between conflict and performance is "nonlinear," given significant interaction effects of conflict, involvement, and compliance on performance. In so doing, channel analysts must include other behavioral factors (e.g., power, satisfaction) within the scope of empirical investigations of this channel phenomenon. Other research directions focus on the limitations of this study. In the experiment, respondents’ perceptions were gathered from the retail side of the franchise channel dyad for automobiles. In reality, respondents from both sides of the channel dyad should be investigated when behavioral dimensions are explored in future research. Differences in perceptions of reality between exchange partners have been found to be a major cause of conflict (Alderson, 1965; Little, 1965; Stern & Gorman, 1969; Stern & Heskett, 1969). Future research should investigate the perceptions of key informants on the retailer and manufacturer sides of the channel dyads with respect to issues that lead to conflict, resolution behaviors, and the impact of these responses on each party’s performance. In 177 addition, it was pointed out that performance was treated as a perceptual construct in this study. Future research should attempt to isolate conflictual issues between channel members that have measurable impact on objective performance measures. In this sense, the results of studies of the effect of behavioral dimensions (e.g., conflict) on performance will provide usable managerial implications. Summary of the Research The objectives of this research were to (1) provide a theoretical framework that might explain the results of previous studies on the effect of channel conflict on channel member performance and (2) provide some empirical evidence that may show that a broader approach is needed which includes other behavioral factors that interact with conflict to affect performance outcome in channels of distribution. Several basic hypotheses were developed for both objectives. Hypothesis 1 stated that conflict x involvement x compliance interaction will interact to affect channel member performance. Hypotheses 2, 3, and 4 stated that conflict and involvement, conflict and compliance, and involvement and compliance will interact to affect channel member performance. To test these hypotheses, a 2 x 2 x 2 factorial experiment (repeated measure quasi-experimental design) was used, which included unique independent variables--high and low levels of conflict, involvement, and compliance. In the preliminary investigation of the research, four conflict situations were 178 identified that fit the treatment levels of conflict and involvement. In addition, conflict-resolution behaviors were generated for each conflict situation such that they fit the treatment levels of compliance. New car dealers in the state of Michigan were chosen as the respondents from whom perceptual data on the variables used were gathered. Two groups of dealers (243 total) were used, one of which was a panel of experts who evaluated the expected impact of conflict-resolution behaviors on three performance measures: new car sales volume, trade area market share, and profits. Group 1 dealers rated the perceived relative effectiveness of conflict-resolution responses by allocating 100 total points among them. The evaluations by both groups were combined to permit derivation of values for the dependent variable in this study--expected value performance impact. In the analyses, it was found that the empirical categorization of the four conflict situations was a near perfect match to the theoretical categorization. The reliabilities of the scales used to measure conflict, involvement, and performance impact were high. The univariate approach to MANOVA was used to test the full and simple, simple models relating conflict, involvement, and compliance to performance. Hypothesis 1 of the full model was well supported for all four performance measures used. Hypotheses 2, 3, and 4 were supported except for the joint effect of involvement x compliance under high conflict and conflict x compliance interaction effect under low involvement conditions. 179 MANOCOVA was used to examine and remove the effects of three covariates (dealership size, experience, competition) on performance and to see if the effects of factor treatment interactions would change. The results revealed that the three-way interactions of conflict, involvement, and compliance when NCSV, TAMS, profits, and summed performance were each used for performance remained statistically significant. In addition, all joint effects of the factors remained statistically significant or insignificant after removing the effects of the covariates. However, at least one covariate variable had a significant influence on performance in the full and simple, simple models except in a few cases. APPENDICES APPENDIX A SELECTED INTERVIEWS 180 General Motors Dealership Ques: Now with respect to some of the issues that you face as a dealer related to GM, what are some of the areas of discrepancy that arise that lead to a difference of opinion between yourself and GM? Resp: You mean an actual difference of opinion? Usually GM predominates, I mean I can have my opinion, but as far as them ever changing GM I really don’t ever intend to do that. Day-to-day basis, obviously we have difference of opinion but it only stays in this house. I guess if I was invited to a seminar obviously I could be verbal on some of the things. Probably the biggest issue with this dealership would be car distribution, the way cars are sold, the way I have to take them, or some that I can’t take. I think car distribution is a national problem with every car dealer. We all feel that we are unjustly being picked on. If we’re big we feel so, if we’re little we feel so, so I don’t think it’s anything that’s going to be carried out. Personally at this store I have a good relationship with my manager and a Detroit car. And in 4 years I’ve learned not to fret about things I can’t do anything about, just go on and do some­ thing I can do something about. As far as a dealer attack­ ing, or discussing perhaps is a better word, with GM some of his local problems, that can be done, but as far as changing any national I don’t think it can be done. I didn’t answer you. I guess some of the things that I think should be addressed would be better product. Model proliferation is too great. I think they’re on their way to doing something about this. GM is a very large, unwieldy corporation. I think they’re improving; they do do things in 9 months, and it used to take them 3 years. But they for some reason can’t do too much instantaneous; as a businessman I couldn’t operate that way; I’ve never been successful at it. Ques: Have there been or is there a discrepancy between yourself and GM with respect to high sale projection? In other words, . . . you only buy so many? Resp: They say this market should absorb so many based on national statistics; I think I have in this case a particular market situation. The fact that ____ Oldsmobile for many, many years. Olds manufacturers and Olds people tended to buy their local product; that situation is being cured because this is a BOC town, Buick, Olds, Chevrolet town. So I find more and more merchants are willing to say I can buy a Buick now because they’re being made locally, but for years they have always beaten me on the head and you don’t sell enough product. They recognize I had a unique situation, but the 181 whole thing is you don’t sell enough product. I think every dealer in the world hears this. As long as this is a confidential thing, they can tell you about wanting CSI (is that a familiar term to you? Yes.); they want good NVI and they want a good service report, they want all this, but the big [thing] that counts is how many damn cars did you sell? And maybe that’s going to turn around in the corporation. I attended a meeting yesterday where they had scaled down their production for next year. In the 40 years that I’ve been in this business, GM kept saying, bigger, bigger, we got to get bigger. I always felt that the minute you didn’t go along with their theory they say, well that’s not the GM theory; you must get bigger. Never mention particularly better, that’s what’s always bothered me. My family has been in this business well since 1923. I was brought up in it. And I do have a loyalty to Buick and I don’t know any other business. But I have loyalty to my customers, and I know in the U.S. you got to grow to be successful, but is there a happy medium between? And I do think that the corporation in this last statement, Roger Smith said, well we’re not going to go not for numbers but we’re going to go back for quality, I think we deserve that. I think in the last 15 years, it was how many of these damn products we could get out. And I don’t say it was always shoved down our throat, but that we built ourself to a fever pitch. We drove the salesman, we did this, we’d do anything. Anything that we could to sell a car, and I’m not saying we still don’t, but we got to go back to being, I consider myself an honorable merchant, and I think you people as a car-buying public are beginning to look up at us as buffoons. I object to the way some dealers run their business as buffoons, but of course that’s obviously their privilege, but I think I object to merchandising in this country; it’s turning into the K-Mart syndrome. You know, turn the blue light on. I have no idea how old you are, but I see a whole generation of people growing up not buying anything until it’s on sale. And that’s stupid. This is a legitimate business. If you need a car, go buy one. If you need a TV, go buy one. But we do and we are; I have a daughter that’s 33 years that won’t buy anything until it’s on sale. And then sometimes, maybe this is the way the market is going to be, I don’t know. You’re going to get more from me than you wanted. Ques: With respect to the factory incentive program, say the rebates and the discount financing, have people had any discrepancy there with GM with respect to that packaging? Resp: No question about it. I think it’s unfair, unethical and unjust that we participate in it. That they take certain monies away from us or certain credits away from us. I think that traditionally over the years these monies that 182 we’ve relied upon we’ve known how to use these monies, rebates and that type of thing. But ____ with the factory rebate is absolutely ridiculous. The paper work, the room for error. The fact that we make our customer aware that we’re doing something that he doesn’t know of, dealer participation may alter price, what does that reek the guy’s a crook. He may do something. Again, back to my former statement, we’ve got to do something to make ourselves look respectable; we are respectable, but we’ve got to stop being the SOBs of sales. Ques: With respect to using these rebates, will they present a problem for you? What are some of the things you do or decisions that you make in terms of operationalizing them that’s beneficial for ____ Buick, but at the same time give GM some satisfaction as well? What are some of the things you do? Resp: Well, to implement for my own concern, the things I do. I guess I don’t quite understand the question. Ques: Well, in terms of operationalizing the rebate program or the discount financing, sure you use it but the way in which you use it, things that you do with it in terms of satisfying your customers, for example, giving them X dollars in the form of a rebate? Resp: On the first basis, we have an option of doing this or not doing this, that is still my option. In fact, the corporation when they would listen to you would say, well, hell, you don’t have to do this. I also don’t have to breathe, so we have to go along with these things when the dealership decides I have no choice. So how do I implement it? We probably go along with the mass. We have to join the mass media. But many times I instruct the salesman just to level with the customer. To tell him that we have to participate. Sometimes our customers want to know how exactly we participate. When we get down to a tight deal and the guy wants another couple of hundred dollars, I just say I can’t do it. The factory is making me, and they literally are, contribute to your rebate. We’re going to give you back $1,200. Nothing unique about this, but when all else fails, tell the truth. A good example is a small town dealer close to me who decided to buck the system, and he didn’t sign up for the program, and he went along very fine for about 60, 80, 90 days. Then GM sweetened the pot; in other words, the program got better. He caved in. He had to cave in. You can sit there with all this _ .The man was a man of integrity, and he said, hell I can sell my product by simply being honest. Edging your initial question, we just do what every other dealer can, we start 183 advertising it. We tell our salesmen to treat it as best they can, be as honest as they can with the customer. But let’s face it, I’d be a fool not to join. You people, pardon me, but I’m going to lump you all together, but you guys out there like it. I think sometimes the public likes to be deceived. I really believe that. Rebates are not deceptive, except how they’re used can be deceptive. Per se they are very good. For instance, there is a rebate that is out now that isn’t even talked about; it’s a Riviera. Riviera, as you know, the El Dorado, the Riviera, and the Toronado are good products, but, hell, they don’t sell. There is a $2,000 rebate to me that we don’t advertise that but automatically it comes along; when you can buy a $20,000 car for $17,000 you’ll recognize it; what makes the world go around [is] price. Ques: Have you encountered any kind of discrepancy with GM with respect to getting your service technician trained; the support that they give you in getting them trained? Resp: No, I think it’s been very good. The last 4 years I think it’s been excellent. I’m a dealer that believes in service training. I’m a service-oriented dealer, as I told you earlier; anything that is presented to me in the line of service I’ll buy. I have the latest equipment that’s being sold. There was a period about 4 years ago when GM was attempting to charge us to train the technicians. And that lasted about 9 months because we refused to send our people in that case. It got very expensive because obviously I paid the technicians while they’re gone. We pay their time and travel. We pay their hotel. We do all this and then we were. . . . Buick was trying to say, well, it’s worth about $200 or $300 a man per day to train. I was one dealer to rebel; I think most dealers rebelled. We refused to pay our people trained who were paid to repair their product. But no, it’s been very limited. I have a district manager who is good; recognized there are problems in the dealership, their size, and they’ll step up to it. Ques: Now when new product technology comes out to service these cars, particularly these new model cars coming out, and you need special tools in order to be able to use these new products. Do you encounter any problems with respect to getting the tools you need from GM to do that? Resp: No, that’s always been a sore subject with some dealers, but not with me. We have to buy the special tools. The packages can run anywhere from $300 to $5,000 depending on what they are, and I guess it’s a fact of life. No, I’ve never experienced any problem. If we just step up to the job and order them and they come filing in and we go from 184 there. I do think there’s duplication; a dealership again such as this size every year buys a specialized tool. We do get some duplication, but I’m pretty satisfied with the system the way it is. Ques: I believe GM used the CSI that you mentioned earlier. Is this something that has led to a difference of opinion with respect to how they are used? Resp: No question about it. There’s some I think they’re in the process of carrying some of the questions I’d like to refer back to NVI, which is New Vehicle Inspection. Would you like for me to elaborate on it, or do you know what it means? Ques: No, I don’t; please go ahead. Resp: O.K. You purchase a new Buick and in about 8 weeks later you’ll get a questionnaire regarding your delivery experience. Questions such as: Did the salesman deliver it? Did you get a tour of the dealership? Was the car clean? and this type of thing. And there was one question in there that I know every dealer objected to it very strongly because when you make a questionnaire for the public you think you can have it absolutely perfect that the layman can read it, and there was one question that said, "Did you get a tour of the dealership?" and it said, "Yes, No, or Not Related." Well, for instance, one of my old customers that had been buying cars from me for 20 years didn’t get a tour of the dealership. After all, he’s been coming here for 20 years, so he put down "No." Well, these questionnaires are read by a scanner, not by a human being, and the scanner had no choice except to say that the customer was dissatisfied; he didn’t get the full show. Well, that was an example of a bad question. They have since remodeled that question, but obviously that gave the whole dealership a bad reputation. We are concerned with CSI. I pay my salesmen a bonus when they reach a certain plateau; with a certain number of returned questionnaires they get $100. I do believe in new vehicle inspection and CSI. And answering your question, has it changed the way I do business, Yes. It’s given me a tool to go to all of my people from the lowest quarter to the best technician and say, "Hey we’re being graded on this," and they see these grades. At this point it’s not just the guy upstairs saying, "Hey, we got to get better." It’s somebody else saying we got to get better. I like CSI. I think it’s good. It’s good for me. It’s good for you. And I think it’s going to be good for GM when we get right down to it. It’s again a sign, I think, that the corporation is trying to do something. 185 Ques: In terms of increasing the CSI, let’s say there’s a discrepancy there and you’re trying to increase it. What would become of the things you would try to do to get the CSI higher? Resp: There’s several things. There’s things that you and I know as business people that we should always be doing. I don’t think there’s any magic. It’s putting a different man on to deliver a car, it’s wiping a window, it’s saying "Good morning," it’s offering free coffee, it’s a TV in the waiting room, who the hell knows what it is, but they’re no one big thing. This is one of the good things about CSI is that it helps me and my managers do things that we know should be done, but it makes us do them. I think the dealership that gains in CSI is gonna just do common courtesy; it’s just gonna make us stop and think. Sorry to over-simplify, but that’s all there is to it. There is nothing else out there. Ques: Well, from having talked to some of the other dealers, they’re basically saying the same thing--that you need to do the basics. Resp: I have a very strange thing, and if you don’t mind I’ll put it on tape. I recently took a man out of my organization and moved into a sales manager position. And I asked him to write a format on what he would do, and the format happens to be on this pad right here, and I read the whole thing. And my reaction when I got through with it, I said, "Hell, that’s what I learned in 1948," and I know what I expected; I expected in 1988 to get something new and different. And he and I both sat down and decided that there was nothing new in selling. And it was courtesy, it was hard work, it was prospecting, it was doing everything that we were taught to do as salespeople years ago. And we’re putting it back to work and it’s working. So, believe me, I don’t know what you’re writing in this case, but there’s nothing new; I’m sorry to do that to you. It’s the same damn crap my father taught me. It’s the same stuff that I went to school at GMI for, except it’s got some new names, but that’s all there is to it. If you’re writing a thesis on how great the future is in the car business, it’s got a good future to it, but we’re going to use the same principles my father used in 1920, and I really believe it; I couldn’t tell you enough about it. Ques: Have you encountered any discrepancies with respect to GM’s warranty program when you have to service people’s cars? 186 Resp: No, I really haven’t. My dealership and my service manager is self-warranty. He is big enough, and the dealership is big enough, and he is honest enough that GM lets me self- warranty. And I don’t know and maybe I’m using the wrong terms, but my service manager (and this is not for publication), but if you drove in here and you were my good customer or if you just own the product and my service manager felt it had failed you and was out of warranty, he is allowed to override that _____ . Where did we leave off? Well, I don’t have any problems; I really can’t say ____ , but if you drove a Buick into my store this afternoon with a 50,000 mile problem and my service manager thought you actually had something you’re way out of warranty, we could still do something. And for that I’m very grateful. I can keep my customers happy. My service manager does not abuse that privilege; they watch it, when they see my per car warranty cost go up and down, but actually I think I’m about $6 less than zone average right now and I have this ____ . I can literally pick up the phone right now and I say, "I want you to take care of Mr. Perry." Well, Jesus Christ, he’s got 40,000 miles on it, I’ll say I want you to take care of him. And we don’t do any cheating. We don’t roll any speedometers back or anything. He assigned it a number, my district manager will come through and say, "What the hell is this?" with Perry getting a transmission repaired with 40,000 miles and there’s an explanation saying let the guy do this, and this, and this. He’s got 6 Buicks. He has a fleet. He’s just a good guy, he’s not a bitcher; this shouldn’t have happened, that’s why I did it. I hope I didn’t disappoint you. I know you’re tired of hearing me cut GM apart. . . . Ques: No, it’s not a matter of that. I’m just really concerned about the nature of problem areas in. . . . Resp: In this store, warranty is not a problem area. Ques: Is there a discrepancy with respect to reimbursement for warranty work? Resp: No. If you’ve had dealers that tell you that, you want to realize that number one, Buick is my best customer. Absolutely my best customer. So I have people that do nothing but work on Buick paper work. That woman who was here was the office manager has two girls that do nothing but all day long make sure this claim is paid, make sure that claim isn’t paid. There again, let’s go back to our opening statement: you’re going to do it Buick’s way or you aren’t going to do it at all. And when it’s concerning money, I’m going to do it Buick’s way. And it’s up to me to do it Buick’s way. If in your research you find dealers 187 that are mad, perhaps the reason is the fact that they don’t work at it hard enough. Sorry about that. I know little town dealers that will tell you that the damn warranty ain’t worth a damn and da, da, da . . . well, number one, they’re not geared up to do the work, they’re not geared up to do the paper work, they don’t have the equipment to do the paper work and many times they’ll say it’s only worth 8 bucks, I’m not going to worry about it. It only takes about 6, 10, 8 bucks and that’s 800 bucks or that’s 80 bucks and you look up and you’re out of business, so warranty is something that’s got to be worked at. I’m pleased with warranty. I’m not too sure ... I think with extended warranty ... Oh, my God, there’s going to be a lot of paper work; you go from 6 to 60 I’m married to you for a long, long time. But there’s a good side to that as long as I can keep married to you you’re going to keep coming into my store. See, before when the warranty came out in 1 year, then you friendly Joe your gas station. Or you went some place else, but now you’re going to stay married to me. I’m going to make a few bucks off you, and I’m going to make a few bucks off of Buick, but it’s going to take a lot and lot of work, because you want to remember that Buick and GM have all of these sources that they’re in to say no. Until you say hell and reapply, you’re not going to get your money. I feel that their object is to say no, and as long as I know they’re getting ready to say no, I don’t gave any damn excuses they’ll say no. Ques: Have you encountered any differences of opinion or discrep­ ancy with respect to recalls? Resp: There again, as briefly as 5 days ago I had my recalls at 100%. So they’re tough but they put enough carrots in front of my service managers, some trips, and he aggressively went and did it; it’s a job that has to be done. No, I don’t have any discrepancy. I don’t think they want to do it any more than we do. Last Thursday morning my service manager took a technician and went out to ____ where an old Buick was jacked up with no tires on it. It’s kind of a rat trap house out there, and he went out and he crawled underneath the car, replaced the brake cables on this junk car, and marked it done and went back to the dealership. The car will never roll in its life again, but that’s what it takes. And I guess if the government’s on GM’s fanny, then I guess we got to help. Ques: What about vehicle allocation, in terms of you getting cars you want, getting them on time? Resp: Well, that’s back to my original statement, with the car distribution, we never can. I don’t think that’s a problem 188 that can be solved, but I think they’re working on it all the time. I think I get my fair share. I think they get here as fast as they can. I learned to live with the program; again, I don’t fight it. Maybe that’s a laissez- faire attitude, but it’s something I don’t worry about. I have the ability, and I think my district manager knows me well enough on a 1 to 1 basis, that I call him up and say, "Damn it, I need this car and I need it in a hurry." Certain period of time, he’ll do his best to get it here. I think it’s still a little human event in there, and of course obviously after 40 years I know enough people that I think I can make a few phone calls. I can’t get a car in 3 days, some models I can. A local car I can get in 5 days. I got to bring in a few favors to do that, but I think the distribution is not too bad. Ques: If you have a car that you need and want to sell and a customer comes in and wants to buy that car, do you engage in a deal or trade? Resp: Oh, certainly, that’s probably done daily. I don’t trade daily, but all of that’s obvious. Yes, I probably in the month of May I think we traded nearly at least 15 units, that would be, we sold 72, so that’s what, one-third of them? Yeah, we did one-third of them for a year. Do you want me to elaborate on that or what? Ques: For the nature of how that’s operationalizing in terms of making a trade, I presume that in terms of getting a car or some car from another dealer and they want the same car that you have which they don’t have; is that the way it’s done? Resp: No, not exactly. It’s done like this. We assume that the dealer calls us and wants X car. We try to in this way, we try to oblige him. We have certain dealers that we trade with. Usually the larger dealers because they have larger inventories. We try to oblige them because we know it’s going to happen to us. It’s the old golden rule; however, if a dealer wants your very, very fine product and you only got one of them, you just say, "Hey, we know you got one and we don’t want to sell it." And he says, "All right, I can understand that." Many times I call up a dealer and ask for a certain car, and he’ll say, "Fine, I got it," and "Yes, you can have it." And I’ll say, "Will you trade?" And he said, "What have you got?" So we go through our whole damn inventory, and he says, "You got nothing I want; come and buy it." But the trading here, we have to sort of grovel sometimes. I have given away very good products to get a particular car because there was some profit in that car deal and you can only sell them once, but it’s a negotiable thing, not price. It’s a negotiated thing. I try not to, I 189 don’t have any policies, but usually we try to keep the dealers within our 100-mile radius happy because we know that some day he’s gonna want a particular trade back. A little honesty among friends. Ques: So, the negotiation, it’s not only just among dealers, but you’re in a situation that you face with GM, for example, that’s a negotiation going on. For example, when we talked about the issue of GM making factory projections, and they want you to buy so many cars, you may want to buy only a certain number that’s less than what they think you should be buying, and the nature of how you handle that type of situation comes to some type of agreement more or less. Is there some sort of negotiation? Resp: We negotiate. My district manager calls me up and say you haven’t gotten in orders for days. It goes like this, Well I don’t want an order from them. I got 15 on the ground. You got to help me out. I need 5 orders from you. I’ll give you 2, I’ll take them, and we negotiate on that basis. We understand that he and his so-called fanny is in a bind, too. He pleads and cries, well let’s negotiate, we’ll negotiate with the factory. I had even gotten to the point that many times they’ll say if you’ll take the 5 dogs, I got one hot cat for you here. And we’ve done that. Depends on how bad you want that hot cat. So we’ll take 5 dogs out there, but and then sometimes when you’re _____ just told you to say no. And then he has to go and sell it to somebody else and. . . . Ques: Is there generally a discrepancy with respect to parts availability? Resp: No, not really. I can get parts in 72 hours. I get parts daily. There’s a truck that leaves and arrives at my store every day at about 4:00 p.m. in the afternoon. I think our parts situation is very good. I really do. We have in this dealership, I belong to a system that allows me to look into either 15 or 20 other dealers’ parts departments, and see if they’ve got the item without them knowing. We’ve all agreed to sign up for it, and then they will sell it to me, and then. ... I think parts is a very good situation, considering the number of parts we got. I think GM has done a real good job on that, I really do. So if they don’t, let’s say there is some discrepancy there, and you can always, for the most part, get that part from another dealer. There again depends on how much and how sophisticated your equipment is and your personnel. Hell, we do stuff today that 5 years ago I wouldn’t believe I could do it, I mean with computers and with phone lines. And it’s just a lot easier. I can pick up that phone today 190 and talk to my parts department and he’ll be back in 6 minutes and say there’s one in Toledo, I’ve got it on the way, and it takes just about as long as it takes me to telemarket it. I couldn’t have done that 5 years ago. The process would be I had to guess that he had it; we had to call Toledo. He had to go look for it. And he had to come back and say he had it. And all this stuff, we do it without even talking to people. The process goes like this. He looks for the part number, he puts it into the computer, the computer tells him that there’s one in Toledo, and he calls me back and says, "Yes, there is one in Toledo; do you want it?" Then he picks up the phone, he calls Toledo, he says to the parts man, "You got a number 9-0736? Will you ship it to me?" And the guy says, "Yeah," and that’s all there is to it. And it doesn’t take much longer than I’m talking to tell about it. I don’t think parts are a problem. Ques: Now what about factory advertising? National, local, or regional advertising in terms of dealer participation? Resp: That’s bad, you finally found one. GM seems to be trying to pass all the advertising over onto their dealers. They’re doing some national, but very little. Their theory is, when I go to a meeting that you know your market better than we do, and therefore you spend some money and we’ll match it. And I don’t think it’s working that well. Because as a little dealer I can’t really afford to spend the kind of money that I think they should spend. Again, we’ll turn around from this K-Mart syndrome. The public today will buy anything that is advertised. Advertising has got out of hand. I don’t like advertising. I have to do it, but it’s got out of hand so badly and people can be bought and sold with advertising. Obviously, the political situation today, we’re being bought and sold by advertising. It’s a little hard for me to conceive that a little 49-cent bottle of air freshener can be advertised as much as it is, yet we can’t afford to advertise a $20,000 car on that same basis. There’s an inequity there, but we’re never going to correct it. I think the factory has got to put more money into their products to advertise more. You and I both know that the guy that advertises is going to be successful. You take everything I told you about being honest and good service and you do everything you can to be Mr. Niceguy, and you’ll fall right on your ass if you don’t have _____ . And I didn’t start it, and I can’t stop it. And I think part of the problem is and Buick as you know is way down, we’re at the bottom of the barrel on the GM hierarchy right now; we’re just not selling. And if you’ve been reading any of this stuff about getting rid of Hal Holbrook, trying to get rid of that old image, they’re flobbing around; they don’t 191 know what to do. And this is really Buick’s fault. I think it’s the agency’s fault, that they haven’t come with it--the heartbeat of America or something like that. But you watch, sure as hell all of a sudden Buick’s going to find an inch and they’re going to run and they’ve got to do more advertising as far as I’m concerned. I’m damn tired of them giving it back to me and say you advertise the product. It’s their product. Does the grocery store buy television or too much television, going to sell a Continental or going to sell Tidy Bowl; you never saw a grocery store advertise. Damn it, Proctor and Gamble advertises. I think a good example of good advertising is Chrysler. Chrysler looks at it as a whole unit. Now they’re a pretty small operation; I’m not mad at lacocca, but they’re a little bitty operation. But their ads lately have been superb. They push Chrysler products. I’m not too sure that GM couldn’t go back to talking about GM products. Have you seen this Buick, have you seen this Cadillac, or have you seen this? I tell you, the biggest fiasco that they’re pulling is the advertising on the Alanti. Super ads, but if you got $54,000 go try and buy one; there aren’t any. Now who the hell ever thought they should spend the kind of money they’re spending on an Alanti, when you can’t get the product. They know damn well that product . . . but there again, let’s give them some ____ . Cadillac wants that prestige image, so what are they doing? They’re advertising their prestige car. So it’s not all wrong in this case, but I just feel that GM should be putting more money in advertising. Ques: Is there a discrepancy between yourself and GM with respect to the amount of advertising that you do in your own local market? In other words, your combination of radio, TV, and billboard advertising? Resp: No, I’m a free agent so they pay none of it. Ques: In other words, do they feel that you’re not doing enough or maybe you should be using more local television advertising as opposed to billboard advertising or as opposed to local radio advertising? Resp: No, they don’t dictate what media I use, but immediately when my sales fall off, when I need a little help, and I say, "What do you think?" Well, have you looked at your advertising lately? They know everything that I think I know what to do, but they don’t step up and say, "Would you like to have some money to advertise more?" No, they don’t do that. It’s two things you want to find out that a GM dealer goes against GM policy, he becomes an independent businessman. Up until then he’s a partner, but he gets his 192 neck in a sling he becomes a very independent businessman. Somebody starts suing me, unless it’s for a bad product, and I go to them, well we really can’t engage in litigation terms. And I’ve known this all my life. If it’s what the corporation wants, then we’re partners; if it’s what the corporation don’t want, then I’m an independent businessman. You make your own decisions. And I really mean that. And it’s not being a bitch, it’s just part of the game. Ques: Recalls, if I haven’t addressed that already. Resp: You have. And I’m doing 100%. I’m not in favor of them, there’s something that has to be done. That’s a government thing. There again, you can sit back home and do all the complaining and bitching about it that you want to, but you’re going to have to get the job done, so you get it done. Ques: Do they make any kind of evaluation of the physical facility that you have in terms of whether it should be larger, smaller, kept cleaner or these kinds of things? Have there been any discrepancies there? Resp: Not recently because this plant is fairly new. This plant is only 15 years old. They would like about another 5 repair bays on this plant; besides that, it’s pretty adequate. You obviously know they dictate the size of the plant, the square feet, the amount of _____ on the black top. I kind of like the way they’re going around ____ my image of excellence. I’ve ____ it more than 3 years in a row, but I kind of like the way they come around and look at your dealership and keep it clean, keep the grass mowed. I think some dealers need that. I think sometimes I need it. That somebody from the outside has to come in and say, "Have you ever driven on the street and looked at your store?" You come in every morning and, see, you don’t look at it, you just drive into it. I think as a whole GM does a good job on that, on keeping their dealers clean. Ques: Is that really a problem area? Resp: Not for me, I think it’s good. Ques: Has there been any discrepancy with respect to GM specifying that you need X number of mechanics versus how many you feel you need to operate your business? Resp: No, I’ve never had any discrepancies. 193 Ford Dealership Ques: First of all, I’d like to start with the nature of looking at some of the issues or problems that you’ve encountered within the last year, let’s say, that have led to a difference of opinion between yourself and your primary supplier, which would be Ford. Issues on the order of these factory sales incentives that are so prevalent nowadays; last fall, for example, when they came out with a big program. Were there differences of opinion between [your dealership] and Ford with respect to the operation of that program? Resp: We have a mutual interest and that is how to spur the market, how to keep the market running. And recognizing that mutual interest, the only question is "How do you get to the objective? What’s gonna be the means? What’s gonna be the path that you’re gonna follow to get that objective?" And there are three alternatives that generally the factory turns to. That’s some kind of cash incentive, an interest rate incentive, or some kind of dealer incentive, whether it’s dealer cash or some kind of incentive for the salespeople. And it’s all--it’s one of the three, or a combination either/or arrangement for the customer. So I guess, in that we agree that the objective is to sell a volume of cars, our disagreements have only been with regard to what, which ones of those methods of incentive to use in the marketplace, work the best. And the disagreements come with what has worked well for us versus what has worked well maybe in other markets. So how to spur the market has created some differences of opinion. Now, the funding of those incentives with what has traditionally been dealer money, we sometimes refer to it as funny money, got all kinds of different, holdback, carryover money and that kind of thing, but traditionally considered dealer money funding has created some concern and uncertainty. But that area of concern is largely lodged in the old dealer. The dealer that is more tuned to doing battle with the factory, and has an adversarial relationship. I think the new age thinking with regard to factory relations is that we’re both here for the same thing. And that a dealership that creates high customer satisfaction, and is able to penetrate the marketplace in sales penetration at levels that are acceptable, you know that dealer’s the best thing that the factory’s got going. Now if you’re creating customer ill- will, or maybe not the levels of satisfaction that you should, or you’re not penetrating the market as you could or should and you need to be more aggressive in promotions of your dealership, then you have a reason to be defensive about that kind of thing. But, when they come out with a 194 2.9% interest program, you know I think most of the dealers recognize that there is mutual benefit there and are willing to forego some of the funny money. Ques: As you look at this package of incentives, the part that the dealer plays, and the rebates and the discount financing, what do you do as a dealer in terms of trying to rectify that kind of a situation when there’s a difference of opinion on which one of those is best? What do you do as a dealer? Resp: Well, we communicate with the factory what we think would work best in our market. I think, I believe that in this business, those that are successful are those that are willing to take a chance and make a decision to do something. And one of the things that gave me a great deal of pleasure this last year was that Ford, rather than sit back and wait for GM to announce some of their programs, on a couple of occasions, stepped up and said, "This is what we’re gonna do." Now, that’s a bit unusual, especially in that the Ford sales had been going so well. And it’s GM that’s got the problem. So this year that has changed in that there’s no need--Ford doesn’t need to put out any incentives whatsoever. They could keep on going. They’ve got the best-quality product on the marketplace and the best dealer network tuned in to customer satisfaction, so I mean, it’s a situation that doesn’t prompt them to spent a lot of extra money coming up with these incentives. Other than simply being competitive in the marketplace, competing with GM. Your direction here is toward adversarial relationships between the dealer and the factory, and I may not be a good test case in that scenario, because we get along super with the factory because of our higher customer satisfaction and we’re doing a terrific job penetrating the market. Ques: Have you had any problem with respect to the factory support for training your service technicians? Resp: We have; last year we were within 1% of becoming a gold medallion, which is the highest level of training, like a Mercury gold medallion dealer. And we have people attending all of the training schools. Now there’s no question, some of the training schools are far better than others. But that’s just like at the university, some professors are outstanding and some are mediocre and some are poor. It’s primarily a reflection of the enthusiasm that they still have for whatever the topic is that they’re trying to convey. The same thing happens at Ford’s. It’s not as protected as, say, a tenured professor, but it’s still a protected bureaucracy kind of thing. Their schools 195 generally have been timely in terms of the frequency and the need and, for the most part, they’ve been very informative. Ques: Have you had any problems or do you have any problems with the way Ford uses the CSI numbers to more or less evaluate the dealer’s performance? Resp: I, you know when they first came out with the, they call it QCP, when they first came out with the QCP report, one of the first things that they said was that this was not gonna be the old stick coming at you. To use it, recognize the importance, the mutual benefit, and let’s get on the bandwagon and pull together and make this high-quality perception, desire, a reality. No question right now the two things are of equal importance. You can be doing a crummy job in market penetration and be an outstanding QCP dealer, and be in great shape in the perception of the factory. You can be doing an outstanding job in market penetration and a crummy job in QCP and they’ll be all over you. They’ll be in to see you, every hour on the hour, trying to figure out how can you keep selling so many cars and be aggravating your customers. I think both areas are very important and there’s no question that in the future that, in order to expand or acquire other dealerships that those two factors are always gonna be the two predominant factors. We’re looking to expand and acquire some other stores, and every franchise you contact, a major line, I mean you’re talking about Yugo, you know, every franchise, two questions: "What’s your present market penetration as compared to the district and national numbers?" and likewise with your customer satisfaction score. Ques: Then QCP, I’m not familiar with the terminology. Resp: QCP is what they call--this is a tracking that I use during our staff meetings to identify where we are, where we’re going, where we’ve been. We started off, matter of fact, Lincoln-Mercury just this last winter did a filmstrip. It’s the third filmstrip they’ve done. They call it "Champions of Excellence." The dealers who have done an outstanding job in customer satisfaction. That’s what this topic was about. In prior years, it was about sales performance or. . . . This was the third cassette that they’ve. . . . And the--we were included. There were four dealers nationwide that were included in this cassette, "Champions of Excellence," and we were included in it as a result of the outstanding job that we’ve done in customer satisfaction. When we started off a year ago, on a scale of 1-10, we had a 5.07. Horrible. 5.07. You think, well that’s half-way. That’s like a "C." It’s not. It’s a horrible score. We were the lowest, absolutely the worst-rated dealership in 196 the entire state of Michigan. Horrible. It came as a real shock to us. So we worked with that all last year. A lot of it had to do with--not that we’re bad people, but just things, and a lot of little things that we were doing that really were aggravating our customers. So we figured out what they were and we got them fixed and we’re doing a lot of things lately. We went to a 5.07 and built that up through last year. When we finished last year, we ended up with a year-to-date score of 6.31. Well, that doesn’t look like a lot of movement, but a tenth of a percent like on market penetration is movement, a heck of a movement. And you’re talking about hundreds of respondees on surveys. So, anyway, last year, we finished out with a 6.31. That was a year-to-date, calendar year-to-date, average. It came through in January with a phenomenal 8.73 out of a possible 10. And this is a combination not--this is a combination of 30-day people, who had just taken delivery of a car within the last 30 days and people who had owned their car for a year. Now, to try and keep a customer completely satisfied, which is what the 10 score reads, that’s how it’s described, completely satisfied for an entire year’s time. Their first year they come back and say, "I was completely satisfied with the product and with the service delivered by that dealership," is phenomenal. 8.73 is what we got in January. We’ve dropped all ____ . That’s just, that’s so good as to be, they came to town and bought steak dinners for everybody in the store and took everybody out to dinner. I mean, it was just unbelievable. We have--the Detroit district is the best district in the country in customer satisfaction, so, you know, when we were at the bottom with 5.07, there are some districts around the country where we would have been half-way ____ . You know, when we got 8.73, we weren’t even No. 1. 8.73 out of a possible 10, we were No. 2. I mean, so, you have to be 8 to 9, somewhere in there to be really way at the top. We’re really pleased. It’s certainly more fun when you’re doing business and delivering that level of satisfaction and our market penetration numbers are climbing as well, and the whole thing kind of fits together. Ques: Have you experienced any discrepancy with respect to new product technology that comes up to service vehicles and needing certain tools, for an example, to be used for the new technology to service vehicles? Have you encountered any problems there? Resp: Not with Ford. I really haven’t. Ford lets us know in advance what tools we need. Some dealers that ____ on a program and don’t buy the tools then try to fix the cars and you can’t. Two years ago, they told me I needed an electronic diagnostic scope. It was gonna cost me $25,000- $30,000. I was gonna die. But I bought it, and I could 197 plug you up to it right now and take your temperature and take your pulse. I don’t know if you have any emissions that you’re exhaling, but if you did, it would test your emissions. I mean, just a great device, mini-computer, hunts it down and has a built-in printer, so when a customer comes in and their car is running poorly, I can hook up the car to the scope, I can run a print-out as to what the computer says is wrong with the car, it has far greater credibility--in fact, I really don’t--they don’t hesitate to ask you to spend a dollar. But, usually, down the road, you’ll agree that it’s a great investment. Ques: OK, well, I think I’ll have to cut it off here because I have to be someplace else at 4:00. Resp: Do you? I cut you a little short today, didn’t I? Ques: That’s quite all right. Resp: I’ll tell you--the one area that they need to improve on is processing their claims, and the simplicity of their claims- processing procedure. That’s the one area that I would say that, as a dealer, that they need work. Ques: OK. Good enough. 198 Import Dealership Ques: What were some of the major problem areas that you’ve had more or less a difference of opinion, let’s say, with your primary supplier? Resp: Well, primary suppliers are very simply the manufacturers that we deal with, and in our franchise, have franchise agreements with and that would be Volkswagen and Mazda and Volvo. Historically, the dealer-factory relationships are not the very best. It’s the attitude that the factories take toward the dealer that I believe creates the resentment, and it stems primarily from the fact that the factory feels that we’re, a lot of us, are shady operators and we’re not very sophisticated businessmen and we certainly don’t really know what goes on in the marketplace. The attitude a lot of times is so ludicrous that it’s to the point of jokes. The factory comes in here and tries to tell us how to sell cars, when the majority of people from the factory have never been out in the retail market to begin with to understand what it takes to sell a car. They wholesale automobiles, and the joke about that is the factory retails cars to the dealer and we wholesale them to the public because the factory gets their price come hell or high water. We’re the ones that have to go out there and merchandise our product and sell it at the best profit margin we possibly can. It’s a very difficult situation. It’s a Catch 22 in a lot of instances because you can’t come back on the factory, fight a battle with them, because essentially you’ll lose the war if you do. You might win the battle, but you will lose the way; they do control. They can squeeze you and they can do it even though there are laws to protect us; they can do it in very subtle ways. The classic example is, well, they’re not supposed to force us to take a specific mix of automobiles; we’re supposed to be able to freely choose our wholesale allocations but that never happens. You know, they’re going to tell you that you’re not going to get your hot-selling cars unless you take about a dozen of these dogs out there that we can’t sell. That’s just the way it is. You want 6 of the hottest-selling coupes, well, you’re going to have to take 6 of the trucks that we can’t push out the door. It’s that kind of situation. Prove it. You want to prove it, fine. You prove it and what do you win, well, you get your allocations straight right, fine! What about the next year, when you really need something; it just doesn’t seem to happen. Let’s face it; when you’ve got as much money invested in an operation like a car dealership, which is extremely expensive in terms of capital investment, you can’t afford to be waging war with the manufacturer; you 199 have to expect in a lot of instances things that are not really fair. I think that it’s very unfortunate because I don’t think it has to be that way. The attitude on the surface is improving, but a lot of the people that are in positions of power in the manufacturer are people that don’t quite understand the mentality of the dealer, and in a way probably resent the dealer because they see us as making a lot of money and they figure they could do it, too. So the people that you deal with in a field level always feel like they can, they could be doing the same darn thing and you want to say to them, if you think that you’re that sharp, why don’t you go out there and try yourself, let’s see how you deal with taking this kind of risk. The volatility of the marketplace and the so-called dynamics of the marketplace are things that the manufacturer tracks with numbers. And statistics certainly give you some information, some of which is quantitative and informative, but it doesn’t really tell you what’s going on with the human being that walks in your door to buy an automobile from you. The feeling, the intangible aspect of selling an automobile and servicing a customer is something that the manufacturer really does not understand. They track it again with numbers; they have statistical data to support their arguments why we should be doing certain things, and it’s not to say that they’re not right; I mean, there certainly are things that could be done and haven’t been done in the past to give better service and to do better follow-up with the customer. Make it more efficient to sell the consumer a product. I’ll be the first to tell you that the majority of the dealers out there figure that they’ve been on a gravy train for a long time, and all you have to do is just pump more money in the advertising, and you work by the numbers, you know, it’s by virtue of volume. Well, it’s getting to the point that the public is not, doesn’t feel like they want to be treated like a number anymore, they want to be treated like a real customer, and of course in our business that’s where we differ. We don’t sell volume. We don’t sell a product and not care about the customer coming back in the door. We sell service; it’s imperative, absolutely imperative, that we take care of people. I’ve built my business on service, and I think a majority of the import dealers had to sell their cars, their products, their dealerships on service because for a long time it wasn’t a matter of making a lot of money, either by volume or by unit sale; you just didn’t do it. You didn’t have the margins and you didn’t have the volume. Obviously, some of the big importers today, the Toyota, the Nissan, and even Mazda, obviously were playing a numbers game. Now we do have bigger volume, but the strength of what is going to happen in the future is going to be based on how well you service people. How well you take care of them. I don’t 200 know that there’s a great deal of loyalty because of the diverse product that’s out there; people can select from literally hundreds of models today. So what does make the difference, what makes the person come into your dealership and buy from you a second time, it’s how you treat them, and we feel very strongly about that, and the manufacturer recognizes that, as well. But they’ve got the band wagon and what they’ve done with that is that they call their CSI, their Consumer Service Index, our Customer Service Index, and they feel that now that is a sword that they can use. If you don’t have a high CSI rating, you’re not a good dealer, so now that’s their barometer to use when they want to gauge a dealership. It’s a real sensitive area because you could be a very aggressive dealer and be a very good dealer, and you could literally by virtue of having the volume and having a very strict policy alienate a fair number of people but yet do a tremendous job on the majority of the people, and those few people who complain, and that’s usually the case, it’s the people that complain, not the people that are really happy, that respond, so you could end up with a poor CSI factor but you could really be a very good dealer. There’s also an example of a dealer that will go out there and set a price on a car and never service a customer, stealing business from other dealers that customer goes back to the dealer for service; he can buy his car there, but he complains about the things that are happening to his automobile. It’s not the dealer’s fault, it’s the product’s fault. He didn’t buy the car there, but he gets the bad reports. Again, the instrument that the manufacturer uses is not entirely fair, and it makes it very difficult for the dealer to come back to that kind of situation. I guess I wouldn’t be entirely fair if I didn’t tell you that I think there’s a lot of dealers out there still who don’t recognize the necessity to change and to become more professional in your business. Well, the old timers who have made a lot of money in the past and they did it by the old hook and crook method and that was it. But that’s an interesting side to that--what came first, the cart or the horse? I’ve been in this business now 17 years, and I would have to say that in my experience, and I’ve been primarily on the high end, I’ve been with Mercedes Benz, Porsche, Volvo, Volkswagen, I’ve been with all the high products. The public actually does more cheating than the dealer. And that may be a very tough statement, but I’ll tell you the majority of the people that say they have a problem with the dealer usually start the problem themselves, and then if they’re caught in it, they’re embarrassed, so they defend themselves very vigorously and they make it the fact that it’s the dealer and everybody knows that car dealers cheat, so that’s the image. I think if there was a way to educate people to make them realize 201 that it’s not in our best interest to cheat people. I mean, people come in here and say, "You sabotaged my car in the service department, or you’ve taken advantage of me." I throw it right back, and I say, "Now wait a second; what do I have to gain by doing that to the point that you discover it? I mean, why would I, I mean I’m going to make a lot of money doing this, I’m going to make money one time at the expense of you going out there and telling a hundred people what a terrible organization I have. I mean, let’s be sensible. I’d have to be an idiot first of all." But it’s usually the situation where people feel that they can get away with it, they can justify their cause by virtue of what everyone perceives as the way things are, and that’s the fact that dealers--you know they are a cheat organization. So it perpetuates the whole thing. Of course, dealers fight it by playing very hard ball with a lot of people. So you protect yourself, and I guess I can use myself as an example; we deal very straight with people, I mean extremely, very straight to the point that it’s almost a fault. People don’t believe you’re being straight with them; it’s almost like you have to cheat them or make them believe that that’s the way the game is being played before something happens. I mean, I’m astounded. I’m absolutely astounded about the way things happen sometimes. You can tell someone the absolute truth about what the value of their car is, what kind of a deal they’re getting, and they somehow still think you’re screwing them. So, you reassure them. You give them every possible way of showing them that this is a straight-up deal. They go to another dealership and they get jerked, they get bounced around. They get every game, every game in the book is played with them. Now they’re really confused; they don’t know what the truth really is. I don’t know whether it’s out of embarrassment that they, you know because they’ve played so hard with you and you’ve been truthful with them, that they don’t want to or you probably won’t allow them to come back and humble themselves to deal with you, so they end up buying their car at some other place and they get taken and they know they’ve been taken, but they’ll never tell anyone. People like to brag about the fact that they got a good deal somewhere, but they’ll never tell you they’ve been taken. They usually don’t like to say that I’ve paid too much for this car. We see a lot of deals like that when we’ve laid down a very good deal and people claim that they got a better deal somewhere else, and somehow or other they end up in our service department, and you know you’re looking through their service folder to mark their warranty book or something, you find the original sales slip in there, it’s happened hundreds of times. The deal comes back and you say, "You dummy," you know that’s what you want to tell them, you got screwed, you know especially on leases. 202 Leases is one area that you can really take advantage of people. I know thousands of cases of people going out there thinking they got a better deal and really getting taken. But to get back to the relationship between the factories. The factories claim that they don’t endorse this type of activity, and I think for all good reasons they wouldn’t, makes sense that they don’t want someone out there representing their product in immoral, unethical fashion. But the real reality of the situation is the numbers again; you push cars out the door, they kind of look the other way. It’s the numbers. Who’s doing the job? Yeah, I know he’s a sleazebag, he’s a son-of-a-bitch, he’ll fuck you left and right, but the fact of the matter is, you know they don’t tell you that, the fact of the matter is that that’s exactly what happens. So, yeah, they might slap the guy on the wrist and do a lot of little things, but let’s face it, that goes on. Ques: With respect to the CSI, if you have a low CSI, what would be some of the things that you would do or decisions you would make in order to try to get it up? Resp: Well, the biggest thing is communications. I think that’s the absolute essence of the problem. If you do not clearly communicate with your customer, if there is a misunderstanding on either side, I mean it can be a misunderstanding on our side to begin with, so if we don’t ask the right questions we don’t really know what their problem is, and we really don’t know that they might not feel that we’re dealing fairly. So we have to act like in­ house psychologist by asking all of the questions to get the feeling that when that customer leaves here that we’ve really covered all the bases. If we don’t do that, we stand the risk of that guy walking out the door and saying, that son-of-a-bitch, that so-and-so, well we didn’t know, we really didn’t know. We did what we thought was right, but maybe we didn’t ask enough questions to get all the information about what their problem really was. So the customer rather than tell us, that is frustrated and leaves and we don’t know, so that’s where the whole problem starts. If in the follow-up, and we do follow-up calls on all our customers, that’s the one area that you start with. You have to communicate, you have to find out if you’re happy with our service, yes, no, well we were happy, is the car running fine and has everything been taken care of, so you ask another form of that question to get to the clarification of what exactly are they happy or not happy about. And I’d have to say the majority of the people first of all are surprised that we called. They’re not quite sure that you’re asking them the question for the reasons that you are. OK, you know it’s kind of like why are you asking 203 me this; is there some ulterior motive to all of this, but [interrupted by phone call] . . . I’m sorry I lost my train of thought, oh communications, so I guess that’s the starting point is making sure that first of all that when they come in the door that you’ve asked all the questions to get to the root of the problem, and I’m talking specifically of the service department. We try to educate our people while or rather after they purchase the car, while they’re purchasing the car, we have what we call preferred customer program where the booklet they get, inside of the booklet is a step-by-step instruction sheet if you will, to help them help us with their service. So what it does is outline some of the steps they can take if they’re having a specific problem, how to schedule the service, how to identify the problems, because I think the whole problem with the customer today is that people know they have a problem with their car but they don’t understand enough about an automobile to be able to explain it. Then they’re frustrated when they come in because they, and especially women, women feel like they’re not being treated like they have an ounce of sense or understand what is going on, so they don’t really give all the details, and of course if the service department isn’t smart enough to recognize the fact that you have to ask some of these specific questions and get to identifying what the symptoms are of the problem, well, then, what are you doing? You’re wasting a lot of time. You might get lucky, and that’s essentially what happens a lot of the time; you get lucky, you fix a problem, or you end up fixing a lot of things that don’t need to be fixed in a trial-and-error method that eventually gets the problem resolved, but you get everybody pissed off in the process. So again it all starts with educating the person while he’s buying the car: this is how we work, and here’s how you can help us and help yourself. When you do have a service that’s only a normal service this is very simply what happens; if you have a problem, here’s what you can do to help identify the problem and then when you’re in here help us identify what that problem is. Because I don’t care how good the mechanic is today and that’s the part that people don’t understand, they’re just human beings. There’s no machine that’s going to tell them how to fix something; they have to know how to fix it. R & R, meaning removal and replacing of a part, is relatively simple, but it doesn’t necessarily track the problem with the complexity of the automobile today. You might get lucky again and get the right part, and then again it might be in a series or it may be a combination of things. So, identifying the problem, communicating, asking questions and then following up with the customer is the best way to accomplish your CSI, getting your people satisfied. Now you’re not going to satisfy everyone, we know that. I had a lady in here yesterday, a 204 woman reasonably educated driving a Volvo. The image of the car, the way that car was sold to her, and I don’t mean by us, but by the manufacturer, that this Volvo was a long-term automobile, it’s a high-class car, you pay a lot of money for it, it’s dependable, it’s reliable, etc. ... In her mind she shouldn’t have some of these problems, and she has a legitimate gripe. A couple of these things she didn’t. But combined with the normal service items and things that happen to a car over a period of time compounding everything. Then she felt that we weren’t looking far enough to resolve the problem. It’s a "damned if you do and damned if you don’t" situation. I said, "Ma’am, if I had gone further and fixed the things that we now discovered, I don’t know if we could have identified it or not, but if we had done that and it really wasn’t the problem I said, you still would have been mad at us, right? Because we would have caused you a $200 repair bill and not fixed the problem. Because what we would have found is that there is exactly two problems. So one way or the other, we would have lost. We just happened to find the other problem first and not identify this problem." Again a situation of how do you communicate what’s wrong and how far do you go. I feel that if there were some way to communicate with the public, educate the public to understand the complexity of what we’re dealing with, that we are human beings, that we do make mistakes. The process that we have to take sometimes is not a purely scientific one, it’s essentially a trial and error. You take a good educated guess, you do your analysis and your trouble-shooting process to the best of your abilities you can follow, and this is the other one where the manufacturer gets back in and they go, "Well, did you look at the checklist?" Yeah, the checklist tells you to do this and this and this, and I said, well if you do it 14 times we haven’t found a problem, you know you go through this process of their check list and by their engineering a field that they’re going to identify a problem with, I said, well that’s why you got this beautiful car sitting in a nice test lab somewhere with everything hooked up. I said, we’re talking about a car that’s coming in here and that’s got 30,000 miles on it and has just been through the war, and you know things are wrong with it but you can’t put your finger on it. You know we can’t make this a real clean operation. So the manufacturer kind of compounds it because they come back and all they know how to do is go through their procedures list. They cannot offer, that’s the problem, they cannot offer the technical assistance in the field that is necessary to help us fix problems today that are somewhat beyond our scope. If we, we couldn’t possibly educate our mechanics to the degree necessary to keep up to date with everything that changes in a car. You know what I mean; we’re talking about some cars out here having 3/4 205 computer systems and God knows how many micro-processing chips and an incredible amount of electronics. No one back there can possibly be educated enough or be sophisticated enough to understand how everything works and to trouble­ shoot to that degree. So they say, "Well you plug in on a new little nice drain. You know we’ll fix it up. You got all the answers, and it reads it right out." Well, yeah, to a degree if you haven’t identified the problem within that system but maybe there’s something that’s just temporarily interfering with this, there’s the part where it’s trial and error. It’s just a matter of being smart enough to go through that process. So all of these things add to the frustration of the customer, don’t do a hell of a lot for relations between the factory and the dealer, and you know it sounds like I’m playing this like we’re the middle guy, but we are the middle guy, and we’re caught in a pretty tight situation. Ques: What about the support you get from the factory with respect to training of technicians? Resp: They offer to the nth degree, they offer training. But the training is at a level at a school in their technical training center, that can afford basics but in most instances doesn’t get too far into the advanced. One of my complaints with the manufacturer is that we can train all of our young guys coming out, to give them the basics. But you don’t have people that are technically competent to go beyond that. I have mechanics that know more than your trainers know. They would like to learn more, but you don’t have anybody capable of training them. And why is that? Because you don’t have anyone that’s really been out there in the field with a hands-on situation dealing with the problems. You’re doing this theoretically, and theoretically doesn’t work in this real world. Sometimes yes, but not all the time. And so when the mechanics are faced with one of your trainers who is just giving him stuff out of a book that’s supposed to happen, they become very cynical about, you know, well have you ever got any knuckles busted doing this, you know that’s what the book says do it this way. Well, let me tell you, you get yourself a big hammer and a pry bar and that’s how it works. That’s essentially what goes on. And so you develop this, man I’m not going to school; I don’t care about this shit. Because you guys don’t really know what’s going on anyway. So that’s part of the problem. I guess what I’m really saying overall, it’s a matter of understanding, of education and understanding. If there was some way this business could work better, it would be by educating the public a little bit more. I would like to see someone come out and explain to the public what a dealership is all about. How does it 206 run and why do you face some responsibility; I mean, they’ve done tons of analysis on all kinds of businesses. I see "60 Minutes"; they tear people apart. With as much as is done in this country with the automobile, why the public isn’t more informed, I mean really given an objective view of what goes on in a dealership and why. Really giving the true stories would go a long way in helping them to understand, for both sides. I’m not afraid of people knowing what we do. I’m not afraid of them knowing what kind of profits we make. People come in here and say we’ll give you $100 over invoice. Well, big deal! Well, how are you going to pay my lights, my rent, my insurance, and pay this poor guy that’s trying to help you? Come on, $100, $500, I mean that’s not a lot of money today. It doesn’t go very far. But everyone has been conditioned; it’s no one’s fault but the manufacturer’s. They come out here with these rebates and the public starts to believe that there’s tens of thousands of dollars to be tossed around and a $15,000 car. The dealer’s compounded by coming out with invoice and $100 below invoice. I sell a lot of cars. I sell every car at a loss and make it up on ____ . So that hasn’t helped, either. People get the view that we can get by by making a couple hundred bucks a car. Today, with the cost of doing business, it’s extremely difficult. You can have 3 good months and make a lot of money, and in 3 months it can put you right in the hole deeper than you can ever imagine because everything is very volatile. But I think the key to everything, and I’ve said this since the day I got in the business, is communication. Communication with the manufacturer. Communication with your employees. Communication with your customers and the public. If people are not talked to and given the benefit of understanding what goes on and why, how can you possibly expect them to look at you in a favorable fashion? The problem also starts, and this gets into another part of our social and business systems, it also starts with being able to afford to train people. In a small business today, the biggest problem that we face is having enough people when you’re busy and having too many people when you’re not busy. We don’t have the luxury of a big corporation where you can have a separate staff to handle different facets of your business, so you have people playing dual roles, and so when you’re busy you can’t do enough, and when you’re not busy you got more people than you need and the expense is there, so you’re caught in between that fine line and the other side of it is as people change and there is a lot of change in the business, you end up having to train or retrain. Well, who do you get to train? Who trains who? There is no formal training system established to educate a service manager or a parts manager how to run his department. Even the manufacturers are just beginning to have training 207 programs set up for service managers and parts managers and general managers. Volkswagen happens to be one of the pioneers of it. Another interesting part that you could consider is the arrogance that success breeds. It’s a classic in our business. A manufacturer is riding the crest of a hot product, starts to use that big rip on a dealer; two years later he’s at the bottom of a heap and it’s one of these, will you guys please . . . take some cars off our wholesale, will you please. I mean, it’s the whole thing turned around, you would think they would get smart and realize that, hey, you might be up there today, but tomorrow somebody might be kicking your ass. Just remember, we’re the guy that does all this for you. Getting back to that aspect of it, I think it’s a matter of a manufacturer not recognizing that we’re their customer first. If you treat us like shit, then what’s our attitude going to be? How are we going to feel? Give us the opportunity to hit back at you, and we’re going to do it. You don’t create a good relationship, and you never resolve some of those differences by having that kind of attitude. The old timers, the old dealers out there, are like old warriors, troops that have been through wars, and they’ve taken on an attitude that’s really kind of interesting. You talk to the old timers, and they’ve all kind of a developed a, most of them have developed a psychology and a political stance that puts a face on what they really feel about the manufacturer; they’re just playing a game. They know how to play because they know they’re going to go out there and do their business, and they’ll take the hits rather than fight back. Guys like myself who’ve come up in the business, you know I started out in sales and got into buying my own dealership, and I, by hook and by crook, I mean I begged, borrowed, and stole to get in, you know I felt what was right was right, so I did a lot of fighting, got my face bloodied a lot. Just starting to learn at this stage in the game that I’m not going to win, so I have to play politics whether I want to or not. I’m not good at it; I don’t like it, but unfortunately it’s a part of this business. This may be a part of a lot of businesses. I guess I have a resentment also to the attitude that is conveyed a lot of times, and that because you’re a little dealer that you don’t have the knowledge, the understanding, the sophistication of the Harvard Business School graduates who work for the manufacturer, who push numbers all day long, can tell you about the classic textbook examples of marketing and dynamics and the marketplace, and all these beautiful examples they give you about how things should work and what you should do--you know, after all, we’re just street people. And we’re just out here doing what we know has to happen to get the job done. It’s like the engineer that comes in here and tells us how to fix a car and never had 208 his hands dirty before. Engineers are the worst people in the world when it comes to fixing something. They only know how it really should work. They don’t really know what happens when it doesn’t work. And it’s true, it’s really true. The worst customers we have are engineers. Engineers will come in here and try to tell us how to fix their car. I swear, it’s a joke. You know, you sit there and you listen and you go, "Yeah." Then when you get done fixing their car, they want to know why you did it that way or why you found out what you did, because that’s not what they analyzed as being wrong. You know it’s that kind of scenario. So, from a dealer’s standpoint, when we start from the basics of selling a car, delivering the car, servicing the customer, and then carrying that through to the process of review with the manufacturer or our sales performance and our service performance. All of the factors that enter into this analysis, a good portion of them are really not understood by the manufacturer. In theory, it really should be a certain way in their mind, in reality trying to explain it to them without them having been in a dealership, really working a dealership, understanding it that you can’t always do what you think you can do because the public is fickle, because the public human beings they are emotional. You know, they do things that are not necessarily fair or rational. You have a salesman who will react in an irrational fashion, and all of these things thrown into the picture just don’t come out to be the simple formula. So my feeling and I have always wanted to be able to convey this message to the public, that if they could understand a little bit more I think the dealer body as a whole ... I mean I’ve been in contact with, I mean I’m involved in dealer groups where there’s a good dealer body out there today, it really is. There’s a younger group of people, it’s a more educated group of people. It’s not your plaid sports coat used-car salesman that happened to make it big. They’re still out there, but it’s changed. You know, you got college-educated people, you got people who have a real understanding of the dynamics of the marketplace, of the psychology necessary to deal with people, that it takes training and communication. I mean, they’re out there, those people are coming of age. But there’s still more to be done. I think the public right now is very confused. I think the frustration of buying a car today, well, if I had to go out and buy a car today, I think it would really piss me off, because I know how frustrating it is to go out there and deal. I mean, we see people coming in and they got a chip on their shoulder right off the bat. It’s like, hey, I know I’m going to get screwed, so let’s make this as painful as possible or I’m going to fight you to the death or my attorney will talk to you before I get out the door. Our competition, we send our new sales people out to shop all of 209 the competition so they know how stupid they are. There are a few smart ones. But the point is that the market place, the business has become so competitive, so massive, there hasn’t been enough done to really help people in the process and which would help everyone in general, and I think that it is tough. It’s really tough. It’s fun, I can tell you that, and as far as I’m concerned this isn’t a biased opinion either. You ask anybody who understands business that the car business is one of the most complicated businesses to run because you have so many facets. You got literally five businesses within a business. It’s not just retailing; it’s not just servicing. And the problems aren’t like a normal retail store where if you have a bad product you bring it in and exchange it; that doesn’t happen. You’re dealing wholesale, you’re dealing retail. You’re dealing with a manufacturer on warranty. You’re dealing with customer faith. You’re dealing with all kinds of conditional circumstances. You have a state that governs you, I mean the state regulations, and it’s just unbelievable sometimes. Granted, they do some good, but I’ll tell you that the irony that the dealer who wants to cheat now has a more formalized method of cheating. There isn’t a lot out there that can’t be gotten around, and the guy can cover his ass and do everything he wants and still get away with it. It’s true; first of all, the state really doesn’t have a very smart group of people. The job itself, I mean when you look at some of the things they have to do, I mean it’s really routine bullshit. It is kind of a farce in a way. I do know they catch certain people. They do understand the public probably abuses a lot of what the law, especially with the department of transportation, where you have to deal with all this crisis. I mean, a customer that knows all that stuff and knows a state representative, he starts off threatening you that if you don’t do this or if you don’t get right to the last five cents, you know, I’m taking you, I’m getting all my money. We’ve had people who have actually set us up, will set us up to get something for nothing. They’re going to abuse the system. They know that they can cheat, and they’re going to do it. They’ll do it one time to me and that’s it. But the state even recognizes that. So it created kind of a monster, and good or bad or otherwise that’s the way it is. Ques: Have you encountered any problem with respect to dealer advertising? In other words, have they got to the point where it creates a discrepancy between demand and supply? Resp: Yes, very much so. The advertising, first of all, is overwhelming, even if it’s as clean in terms of a legal sense. It’s really deceptive in a lot of instances. Naturally, you want to sell what you have on the lot. You 210 want to sell out of stock. But what happens is that you wouldn’t have all that stock there to sell at these give­ away prices if you didn’t have a problem with it. So how do you accomplish getting people in there to convince them to buy what you have and end up selling off of that? You’ve seen the ads--today, today only, or for three days only. What is this creating? Well, in most cases you’re really not reducing what it is that you’re having a problem with in inventory. You’re getting people in that say, "Well, I don’t want that for $9,999." You end up selling them a $12,000 car that they have to order. Is that smart? You sell cars, but you still have to carry that stupid inventory out there, or you take a loss. I haven’t figured out the _____. It’s one of the most difficult aspects of our business, advertising. And tracking the success or lack of success. My personal feeling, and I think I’m beginning to see some results from this, is that you really have to create your dealership image, and it’s what it is that you’re selling, you’re selling your point with services, your people, what you are, not the product itself. Product has to be sold by the manufacturer. Special discounts and occasional events to reduce inventory. It’s legitimate, yeah, I think it accomplishes things. I guess there’s been tremendous success with the cowboy comes out to California who are on the TV 10 times a day, 7 days a week, 24 hours a day. The Ollie Fretters, just advertise, advertise, always advertise. A lot of people just buy because of the fact that they see the thing, time in and time out. Does it help sell more cars? I really don’t believe it, but I guess I don’t have a good argument against it. It’s never been tried yet. There are exceptions; there are dealers in this country who don’t advertise. And they have very successfully sold cars; in fact, they’ve been leaders in their own market area, by nothing more than good phone solicitation and follow-up and make the system work without spending tens of thousands of dollars. We found success that way, too. You can literally tell, you can spend $10,000 in a week and get nothing. What has that done, they may have been legitimate sales give-away deals [but] people don’t respond; why is it? Who turns that faucet on and says go out there and buy that car? I mean, I’ve been faced with it myself. I’ve sent sales ads to stores and I’m not stimulated by it. I don’t care how good the deal is, but ____ why, I don’t know, it’s just a general feeling. You either get turned on or you get turned off, and one of the interesting things that I have begun to gauge my business by a little bit is that. . . . 211 Import Dealership Ques: What I’m really interested in, basically, is there are so many issues or problems, if you will, that come about that have happened in the last year or so that have led to a disagreement between yourself and your primary supplier. I realize that you’ve got more than one supplier that you deal with. But, for your primary supplier, problems or issues that have arisen that have led to some disagreement between you and your supplier. Resp: OK, as far as disagreements, I think I could phrase them more in the line of differences of opinion, how to handle various problems that we’re faced with. The biggest frustration I think the auto industry faces right now, and I’m talking the entire industry, not just my particular business--we deal with 6 manufacturers, Volkswagen, BMW, Porsche, Audi, Mercedes Benz, and Subaru. And the biggest problem that the industry faces is how to stimulate the automotive buying public. And everybody has kind of followed the leader in this respect as far as the, you know, Chrysler, General Motors, Ford, they’ve come out with all these buyer incentive programs. And all that has done is to add further confusion to the automotive buying public. So what happens is we kind of follow the market; even though we’re in the import business, we pretty much parallel the market. If the domestic business is good, our business is usually good. If their business is down, our business is down. They’re really very closely related. And even though we are trying, we do a small percentage of the national business; imports in this country are selling about 27% of the total industry; on a dealer level we don’t sell that much. We’ll sell probably around 15% to 20% of the total market. So our frustration has been, you know, when are they gonna come with the program? And when are they gonna get back, when are they gonna eliminate all these damn incentive programs and get back to running a business like a business should be run? In other words, you’d sell a car, you’d sell a product, you’d sell a service based on the value, what the real value is, not the hocus-pocus value that’s plagued the industry because of incentives or low interest rates, discount financing, rebates. You know, that’s been the frustrating part. And, unfortunately, the manufacturers themselves don’t know the answer. And so, the biggest disagreements that we have basically evolve around those market issues. You know, how and when to stimulate the market. Ques: OK. Have you experienced a problem with respect to the factory making sales projections in terms of the number of 212 cars that you should or can sell versus the number that you want to buy from the factory? Has that ever caused any differences of opinion? Resp: Yeah. Those are basically distribution items. And, yeah, there’s always times where we’d like more of a given model. Just like right now, there are some various hot, you know, hot models. And that’s true with almost any franchise. The hot models in Volkswagen right now are the Fox. The hot models with BMW are the Cabriollet 325 series. The hot model with Mercedes Benz is the 300E. The hot model with Porsche is the 911 series. So, yes, we do, we’re always constantly trying to get more of those hot cars. But that’s pretty much where the industry is, everybody has a hot car and they, and you never get what you want of the hot car. Ques: What do you do in the way of taking action when, let’s say, they recommend that you buy a certain number of cars. Is there anything you do in particular to handle that situation or to sort of resolve it or reduce it? Resp: Well, what you try to do is try to figure out exactly what it is that you can sell. Now whenever any manufacturer gets a hot car, they, and they will never come out and tell you this, but they do packaging. And they package their cars and say, "OK, now if you want to get this hot car, you’re gonna have to buy 2 of the not-so-hot cars." And they’ll come out and say, "Well, we don’t do that." Well, they do do it. OK. It’s, you know, we know it’s illegal for them to do it, but they’ll do it anyway and say, "Well, gee, your minimum stock requirements are X number of units," and based on those minimum stocking requirements, you’re supposed to have so many models of each particular style. And the dealer, in this respect, wants to eliminate stocking a hell of a lot of cars because there’s a tremendous expense that goes along with it. Ques: Have you experienced any sort of differences of opinion with respect to the warranty program in terms of the factory more or less living up to the warranty program that is laid out? Resp: We’ve been pretty fortunate in that respect. I think that all of the lines that we sell, the factories are very responsible. And they do have good warranty policies. Where we do run into frustrations from time to time are parts situations, where we’ll get a particular problem that may develop on a car and it goes on a back-order parts status where the parts aren’t available anywhere in the country. Well, the customer right away doesn’t, you know, they don’t give a damn. All they want is their car fixed. And it is frustrating. And the one that we’ve had the most 213 problem with right now in the parts situation is Porsche. We have had an extreme amount of problem with car-down situations with Porsche. Ques: What about the training of your service technicians, getting them trained? Resp: There again, our, these manufacturers that I deal with are very, very responsible. And I think the training responsibility is a two-fold responsibility. It’s a responsibility of the dealer and the manufacturer. A lot of dealers don’t want to spend the money. They’re basically cheap and they’d just as soon hire somebody off the street and let them go through trial and error and, you know, learn by the seat of their pants. We have a very, we spend a lot of money on training. We have a very well-trained staff of people. A lot of dealers will wait toward the tail-end of the year; they figure, "Well, I’ll put them off as long as I can so I don’t have to spend the money." Well, my philosophy has always been service is very important. Service is the most important part of my business. And we get of our, all of our training is done now through almost the entire year. But in the last 30 days, I’ve had 6 or 7 people in school. Well, that’s a tremendous expense. Because all of our training centers are ... at least the closest one is 225 miles away. So they have at least 2 days just in transportation. You have to either give them a vehicle to drive there or a car or an airplane ticket, and then you also have to pay them while they’re gone, plus pick up their expenses while they’re there. So it’s a very expensive proposition for me to send a technician to school for a week; it’s about $1,000 to $1,500. So it’s a major expense. Ques: So I guess there are dealers who sort of slough-off on that and really look at the expenses involved. Resp: Exactly. They’re looking at the expense. You know, in the last 60 days, I’ve spent close to $10,000 in training. Well, you know, if you aren’t committed to quality and to service, you aren’t gonna spend that $10,000. It takes a major commitment. Like I say, it’s a two-fold responsi­ bility. Ques: Interesting. Do you see these Consumer Satisfaction Index ratings as threats that the factory uses? Do they cause a difference of opinion between yourself and your primary supplier? Resp: If you’re doing your job, all it’s gonna do is further enhance, you’re reinforcing the quality image in the 214 community. We’ve always ranked in the top 10 as far as customer satisfaction with all of our franchises. We’re number, in the last, well in the last 5 years, we’ve been in either number 1, 2 or 3 slot in customer satisfaction, so we pride ourselves in that. And our people know that we’re conscientious and concerned about satisfied customers. Because if the customer isn’t satisfied, he’s not coming back. Ques: Then this is a. Resp: Exactly. But now see, again the dealer that has a poor attitude toward service--he’s telling you that "Hey, these customer satisfaction surveys are a bunch of malarkey and they’re unfair to the dealers, etc." What he’s not telling you is that he doesn’t give a damn about the service department. He hates service. He wishes that it would burn that part of the business down because he doesn’t want to deal with it. Because it is a hard part of the business. It’s a tough part of the business. Ques: That’s interesting. Now, in the area of using, let’s say, new technology or products that come out that are used in the service aspect of cars, in other words, having the tools necessary to use certain new products that come out, or a new computer-type product where you check automobiles’ electronics, etc. Have you incurred any problem with that aspect with the factory? Resp: Yeah, we, the factory will come out from time to time and say, "You have to have this piece of equipment. If you don’t get this piece of equipment, we’re not gonna send you the cars." I’ll give you one example. BMW has an engine tester that is $30,000. They said, "Well, if you don’t buy that piece of equipment, you’re not gonna be able to get the new 735 series because all systems have to be tested on this car, and this is the only way you can do it." Well, we’ve had that machine now for almost 3 years and we’ve used it less than 5 times. Well, you know, when I look at how much it’s cost me, you know $6,000 every time we use that machine? You know, it upsets me. That makes me angry. If they’re gonna have tools and require me to have tools, then those tools, you know, there’d better be a payback. You know, the cost-benefit relationship has to be there and with a lot of these tools, there is no cost-benefit relationship. Well, and the worst part is, what added insult to injury is, they said now 3 years ago, "Now, you buy this machine and that’s all you need." When they came out with the new model this year, they said now I have to have another $11,000 update on the machine. So now I’ve got a machine that’s $41,000. One machine. 215 Ques: And it’s not really being used. Resp: That’s right. And if you get a, now a dealership like ours where we’ll sell, let’s say, this year we’re forecasting to sell about 80 BMWs. That’s $500 a car. And the consumer isn’t gonna pay that $500. He doesn’t give a damn if that machine’s here or not. He will if he needs it. But when they come in to buy the car, they won’t know if I’ve got that equipment or not. Ques: Now, has there been a difference of opinion with respect to the factory more or less requesting the dealer to increase their cost or share in the cost of factory-sponsored programs, like advertising, for example? Have you incurred any problems in that area? Resp: Increasing our costs? Ques: Yes. Taking a larger share, in other words--a larger dollar contribution in factory-sponsored programs? Resp: All of the manufacturers are promoting advertising groups. And one of the concerns that I have about that is, are they still gonna spend the same amount of dollars as they did before, or is it the dealers now who are spending the money to promote the product? Really, the factory’s responsibility is to create the overall awareness of the product and sell people on the benefits of owning that product and create traffic and interest for that particular product. The dealer’s responsibility is to tell them where it is. Where we are and who we are and why we’re here. And that is a concern. Because all manufacturers are promoting more and more of these advertising groups. And you have to really, and one of the other really frustrating things is the factories know how much we spend but we don’t know how much they spend. They’ll never tell you. And I’ve sat on national councils and we’d say, "Well, how much are you spending?" They won’t, you know, they ____ that one. They don’t want to in dollars and cents. Ques: It’s sort of like a leverage thing, if you will, I guess, that they’re using in some sense. Resp: That’s correct. Ques: Now, let’s get back to these incentive programs. This discount-financing thing. When that comes about, how do you normally, what action do you take or things that you do to deal with them when the factory sponsors this discount financing? 216 Resp: What you have to do is take a look at it and see if they’re strong enough to use as a marketing tool. And if they are, then you go out and spend the money promoting that. To give you an example, Volkswagen just came out with a 3.9% interest program that’s good ’til the end of this month. Well, it’s, you know, people haven’t had low-interest financing on the imported products in the past, so we thought, well, maybe this is the time to spend some money. So our advertising group, we spent close to $15,000 promoting that fact on television and newspaper. Ques: So is it a shared thing between yourself and the factory in the nature of this discount financing? Resp: Right. Yes. Ques: So, it is really a difference of opinion there with respect to the contribution that you’re making relative to what they are contributing in order to make the public aware of this? Resp: No. In this particular situation, I think they’re doing their fair share. I think they’re handling it properly. You know, the dealers can always argue, "Well, they could do more." I mean, but that’s typical dealer reaction. They never do enough. Ques: Now, let’s take a situation like the extended warranties. Has there been a difference of opinion with respect to the operationalization of the extended warranties--what you have to do versus what the factory says that you should be doing in servicing these warranties? Resp: Well, as far as the warranties themselves go, now the extended warranty programs that are available out there are either number 1, offered by the manufacturer. Now to give you an example--Volkswagen products is a 2-year unlimited mileage warranty. If you drive that 100,000 miles in 2 years, and you’re still under factory warranty--there is no mileage limitation for 2 years. We have other manufacturers that are 5-year, 6-year, 3-year warranties. If you sell the extended warranty, it just, you just follow the guidelines of the program, so I think they’re fair. I think they’re very fair and very equitable. The advantage of a factory extended warranty program in the minds of the consumer is it has more credibility. In other words, if you come in here and you’re looking at a Volkswagen product, and I say that this car has a factory extended warranty program that’s available for a cost of $395, and it’s through Volkswagen, right away you’ve got confidence. But if I say we’ve got an extended warranty program that’s through XYZ Warranty 217 Company, you say, "Well, who the hell is XYZ Warranty Company?" Whereas the factory extended warranty programs have credibility. And those are the ones that we choose to sell. Ques: Is there a situation that occurs when, if there’s an over­ production of certain models, and the factory sort of puts the pressure on--they’re trying to get these slow-moving models to get them sold; is there a difference of opinion with respect to dealing with over-production of certain models--they’re not selling well, let’s say. Resp: Yeah, there is. But, again, that’s something that’s really beyond our control. We, as dealers, don’t have control over that. When they’re built, there’s nothing you can do. You can’t put them in _____ . Somebody’s got to sell them, and it’s the dealers that have to sell them. So when we get in a situation like that, where there is an over-supply, or they come out with a bad model, somebody has to bite the bullet. And it’s usually the factory suffers the biggest part of the bullet, and the dealers have a demented margin. So we suffer, as well. Ques: So, in essence, when you’re dealing with a slow-moving item like that and the part that you play more or less in terms of what you would normally do to get that item sold, is you may not be able to sell it as much as if you ordered another popular model, where your margin would be way higher. Resp: That’s right. To give you an example, Audi. When Audi first came out with their new body style in 1984, it was a very, very appealing ... it was a hot model. Our gross margins on that car, gross profit margins, were running about 13-1/2 to 14% of the selling price. With all the negative publicity that Audi’s had as a result of the 60 minute unattended acceleration issue and all of that, our margins went from 13-1/2, 14% . . . well, we were down to about 4-1/2%. In order for us to break even, we’ve got to retain about a 7% margin. That tells you what happens. And everybody suffers. I mean, the factory is losing money right now. The dealers are losing money right now. So basically what we have to do is say, "OK. Is it a strong enough franchise? Is it a strong enough product line? Do we want to keep doing that? And do we see light at the end of the tunnel?" In this situation, yes, we do. Ques: Let’s say that you have a problem more or less and it comes down to a situation in which there’s a problem with getting, the availability of a model when you order it from the factory. You can’t get it on the spot, let’s say. What would you normally do in terms of action you would take and 218 the decision you would make in order to get a popular model locally, or from some other dealers, etc.? What would normally happen in those situations? Resp: Well, when a model is hot, chances are it’s hot all over. So, for someone else to deal er-trade or to get that car up to me wouldn’t make economic sense. But there are situations where you--the key here is timing. Timing is the essence of anything. But if you can see early on that this model is going to be a very desirable model, then you can kind of position yourself and try to build up your inventory or availability. But the real bottom line is, there’s only a limited amount of them. This is supply and demand. There’s only a limited amount that you can work with. Now, right now, the best seller is Mercedes Benz roadsters, the two-seater 560 SL. That’s always been a very hot car. Well, traditionally, we sell those cars before we even get them. People come in here and they look and say, "We’d like to see an SL." We don’t have one. Well, the biggest fear that we have as a retail merchant is that their interest is out there for that automobile. They come in here, they don’t see one, chances are they’re gonna go someplace where they can see one. And if they do, we’re gonna lose the sale, we’re gonna lose a customer. Now I just bought an SL Roadster out of Texas through a broker, that I paid $8,000 over my normal cost, just to have that car here. Well, I’ll end up selling that car with probably less than a 4 or 5% gross profit margin. But the reason why we do that is that we know that this time of the year, people are looking for that type of a car. And if they come in here and don’t have the opportunity to see one, they won’t order one. And if they don’t see one, they will go somewhere else to buy one. So if we can get them in here and slow them down enough, so they can see the car that we have, and if they’d want to order one, we can take an order, and hopefully keep their interest. There again, that’s a management decision that has to be made. A lot of people say, "Well, you’re crazy to pay that kind of, why pay $8,000 more for a car than you know that you can order one and get one 3 months later for $8,000 less?" Well, that’s a marketing decision that you have to make. Ques: Does it present a problem then when you order these, you want to get that car from the factory, let’s say, and they won’t, let’s say there’s something wrong with respect to allocation, who gets what models depending on things like the size of the dealer, for example? Resp: Well, see, Mercedes Benz probably has the fairest system of all. And you start with basically 100 cars. For every 100 cars that Mercedes Benz builds, 5 of them will be SL 219 Roadsters, maybe 6. So, when I look at my allocation, I know that basically 5 to 6% is all I can ever expect to get. Because that’s all they build. And they distribute those cars equitably through the whole system. And so, a guy that sells 300 cars, sure, he’s gonna get 3 times as many SL’s as I’m gonna get, but that’s just the way the system works. But it’s hard, again, it goes back to supply and demand. There’s only so many out there. You’re only gonna get so many. And a lot of people, consumers think, "Well, gee, you can replace that car." That’s where the role of price comes in. If you came in to me today and said, "Wayne, I’d like to buy an SL Roadster," I’m gonna say to you, "Well, we can order one for you. It’s gonna probably take 2 to 3 months to get you one." And you say, "Well, what’s the best price you’re gonna give me?" Well, we’re gonna be pretty inflexible, knowing that I’m only gonna get 5 or 6 of those a year. So, because, again you’ve got a small number. But it does one other thing. It also says to the customer, "There is a value there," and what’s why the SL Roadsters have the highest traditional value of any car that was ever built. Because they hold their value, because dealers know they’re scarce, there are a limited small number, and, therefore, the value is there when they’re used. Now we just sold an ’83 for $33,000. Well, that car brand new sold for about $42,000, so that car lost $9,000 in 4 years. Well, I mean that’s phenomenal. But the new one today now is $58,000, so the person that bought that new one is betting on the same concept. So that, again, supply and demand really determine what the market is. Ques: Well, it sounds like a lot of interesting things that go on here with respect to the relationship between yourself and dealing with your suppliers for all the various lines that you do carry. I guess the nature of what I’m doing with respect to this research is trying to sort of focus in on some of the areas that lead to differences of opinion between dealers and the factory, and I’m trying to look at that from the standpoint of how do dealers really react to those types of things that the factory, actually the factory takes, that may or may not be perceived as things in your best interest, and when it happened, how did you react to that? That’s the nature of my research. Resp: Sure. Well, if there’s any one message that I want to carry back to you, it is that there is in the minds of some automotive executives, these are executives at the manufacturer level, that the dealers are their adversary. That the dealers are out there making, you know, just wild profits, and they don’t do anything at all. It’s the manufacturers that really create the market. So why do they need the dealers? This is what a lot of manufacturers-- 220 their opinions of dealers are, whereas the real role of the dealer and the manufacturer should be one of partnership. And they should be working closer together to develop a better understanding of what is really going on in the marketplace. Because there is no one that can tell you better about what’s happening in the marketplace than the person that sits here at this chair, because I watch that floor. I listen. I know what the customers want, what they’re asking for. And, sure, the manufacturer will say, "Well, we do surveys, we do all this, we do that." But if you aren’t there on the firing line on a day-in and day-out basis, you really don’t know. And I keep telling them. I said, "Take the time, spend a week, 10 days, on a sales floor, or at the service desk. Talk to the consumers. Find out what they really want. You know, what they’re looking for." That’s what has to happen in this whole industry if we’re gonna get back to business as normal. The confusion of the buying public as a result of all these incentives that started, you know, back in the early 80s, it’s nothing but confusion and distrust. Because they don’t believe you. They don’t believe the dealer. And the dealers have created that problem for themselves. There’s been a lot of prostitutes in this business. A lot of unethical dealers that are flim-flammers that are, they’re carnival, they use a lot of carnival tactics, and it’s flat bullshit. That’s how they sell cars. What the consumer wants to do, all they want to do is to buy a good car at a fair price, and don’t want to be lied to. And they want to be taken care of after they buy the product, with a good warranty. And that’s how I’ve survived in 27 years in this business, is by doing exactly that. When they’ve got a problem, we talk about it. We discuss it. We try to work it out, try to solve the problem for the customer. We don’t leave them high and dry. But if the consumer doesn’t trust the dealer, wants to get the best deal he can, well then the dealer can’t afford that $41,000 piece of equipment to really diagnose his problem when it comes in, or to spend that $10,000 or $12,000 for training. The consumers, all they, the consumer, says, "Well, how much over cost can I buy the car for?" And what they don’t understand is cost. What is cost? And see, my cost is basically 7% of the gross profit of a new car. So you buy a $10,000 car, my cost of that car is about $7000. So before I make, if I want to make any profit on that car, I’ve got to sell that car for at least $11,000. APPENDIX B QUESTIONNAIRE TYPE A 221 PLEASE RESPOND TO THIS QUESTIONNAIRE ON THE BASIS OF YOUR PRIMARY SUPPLIER AND YOUR PRIMARY MAKE OF CAR (THE ONE WHICH YOUR DEALERSHIP SOLD THE MOST UNITS OFF IN THE PAST 3 YEARS) SECTION A: FACTORY-DEALER RELATIONSHIPS In this section, we are seeking your perceptions of factory-dealer relationships. (PLEASE CIRCLE THE APPROPRIATE NUMBER) Very Very To what extent are the goals of the factory COMPATIBLE with Incompatible Compatible the goals of your dealership? 1 2 3 4 5 6 7 Not Very Very In general, how WELL would you say your dealership and the Well Well factory work together toward goals which you and the factory 1 2 3 4 5 6 7 both wish to attain? Not At All Very In general, how COMMITTED are you to finding goals shared Committed Committed with the factory? 1 2 3 4 5 6 7 Very Constant In general, how would you describe the FREQUENCY OF CONFLICT Infrequent Conflict between your dealership and the factory? 1 2 3 4 5 6 7 No Tension Very High In general, to what extent is there TENSION between your At All Tension dealership and the factory? 1 2 3 4 5 6 7 Not At All Very In general, how COMMITTED are you to reducing conflict with Committed Committed the factory? 12 3 4 5 6 7 Not Satisfied Very How SATISFIED do you think the factory is with your At All Satisfied dealership's sales performance in your trade area? 1 2 3 4 5 6 7 Not Satisfied Very How SATISFIED are you with your overall relationship with At All Satisfied the factory? 1 2 3 4 5 6 7 Not At All Very Suppose the factory decided to drop your dealership as its Difficult Difficult distributor due to reorganization. How DIFFICULT would it 1 2 3 4 5 6 7 be for you to get another dealership? Not Satisfied Very In general, how SATISFIED are you with the overall quality At All Satisfied of cars made by the factory? 1 2 3 4 5 6 7 SECTION B: CONFLICTS WITH PRIMARY SUPPLIER Below are several conflict situations. For each situation, you are asked to respond to several questions regardless of whether the conflict is a current problem or not. Please view each situation as an independent event. (PLEASE CIRCLE THE APPROPRIATE NUMBER) 222 CONFLICT SITUATION 1: NEW VEHICLE ALLOCATION AMD DISTRIBUTION THE FACTORY (PRIMARY SUPPLIER) HAS AM UNWRITTEN POLICY WITH RESPECT TO NEU VEHICLE ALLOCATION AND DISTRIBUTION. THAT IS, TO GET THE DESIRED QUANTITY OF A PARTICULAR MODEL, YOU MUST ACCEPT MORE THAN YOU WOULD LIKE OF OTHER MODELS. How FREQUENTLY have you faced this particular conflict situation in the oast 3 years? (PLEASE WRITE IN THE TOTAL NUMBER OF TIMES THIS CONFLICT HAS OCCURRED) ____________ times. Not Very Very How INTENSE would you characterize typical disagreements Intense Intense over this conflict, regardless of whether the disagreements 1 2 3 4 5 6 7 were settled or not? Unclear or Very No Guidelines Clear To what extent does your current contract with the factory At All Guidelines (primary supplier) provide clear guidelines about how you 1 2 3 4 5 6 7 might resolve this conflict? PLEASE CIRCLE THE NUMBER NEXT TO EACH QUESTION THAT BEST CORRESPONDS TO THE DEGREE TO WHICH YOU DISAGREE OR AGREE WITH EACH STATEMENT, KEEPING IN MIND THE ABOVE CONFLICT SITUATION. Strongly Strongly Disagree Neither Agree I am very concerned about the factory's vehicle allocation 1234567 and distribution policy to my dealership. I am very motivated to resolve this conflict. 1234567 The factory's policy of new vehicle allocation and 1234567 distribution is very important to my dealership* success. LISTED BELOW ARE SOME WAYS OF RESOLVING THIS CONFLICT SITUATION IN TWO CASES. WE WOULD LIKE YOU TO EVALUATE HOW EACH RESPONSE OPTION WOULD RESOLVE THIS CONFLICT IN EACH CASE. BELOW TO THE RIGHT OF EACH RESPONSE, PLEASE ALLOCATE 100 POINTS AMONG THE RESPONSES SO THAT YOUR ALLOCATION REFLECTS THE RELATIVE EFFECTIVENESS OF EACH ALTERNATIVE RESPONSE TO RESOLVE THIS CONFLICT SITUATION. MAKE SURE THE POINTS ADD UP TO A TOTAL OF 100. Example: If there are 3 responses, the allocation can be 33-33-34 points or 34-33-33 points if they are EQUALLY EFFECTIVE or 20-30-50 points or 20-30-50 points or 10-40-50 points, etc. if they are NOT EQUALLY EFFECTIVE. Allocate similarly for 4 or more responses. Make sure the points add up to a total of 100. Case A. Assume you oo along with the factory's new vehicle allocation and distribution policy. 1. Order more than the desired quantity of new cars to ensure that you receive the desired quantity. 2. Engage in demand forecasting for different models and build inventory for those models which may become popular in the future. 3. Accept the factory's typical dealer allocation of new models. 100 TOTAL Case B. Assise you do not oo along with the factory's new vehicle allocation and distribution policy. (PLEASE ALLOCATE 100 POINTS IN THE SAME MANNER AS ABOVE). 4. Use a computerized locating system or a conventional locating system to find special vehicles and engage in a dealer trade. 5. After locating a vehicle you purchase it outright. 6. Concentrate your efforts on buying and selling used cars when new models are not available. 7. Appeal to your district manager to increase your allocation. 8. Protest the factory's unwritten policy in regards to new vehicle allocation and distribution. 100 TOTAL If you go along with the factory's new vehicle allocation and distribution policy, what is the typical (most likely) way you resolve this conflict given the 3 options listed in Case A above? (PLEASE CIRCLE THE APPROPRIATE RESPONSE NUMBER) Response 123 If you do not go along with the factory's new vehicle allocation and distribution policy, what is the typical (most likely) way you resolve this conflict given the 5 options listed in Case B above? (PLEASE CIRCLE THE APPROPRIATE RESPONSE NUMBER) Response 45678 223 CONFLICT SITUATION 2: ADOPTIOH OF NEW PRODUCT TECHNOLOGY THE FACTORY (PRIMARY SUPPLIER) HAS INFORMED YOU THAT YOU WILL NOT RECEIVE YOUR TYPICAL ALLOCATION OF NEW MODELS UNLESS YOU PURCHASE THE NEW PRODUCT TECHNOLOGY DESIGNED TO SERVICE THEN. How FREQUENTLY have you faced this particular conflict situation in the past 3 years? (PLEASE WRITE IN THE TOTAL NUMBER OF TIMES THIS CONFLICT HAS OCCURRED) times. Not Very Very How INTENSE would you characterize typical disagreements Intense Intense over this conflict, regardless of whether the disagreements 4 5 6 7 were settled or not? Vague or Very To what extent does your current contract with the factory No Guidelines Clear (primary supplier) provide clear guidelines about how you At All Guidelines might resolve this type of conflict? 1 2 3 4 5 6 7 PLEASE CIRCLE THE NUMBER NEXT TO EACH QUESTION THAT BEST CORRESPONDS TO THE DEGREE TO WHICH YOU DISAGREE OR AGREE WITH EACH STATEMENT, KEEPING IN MIND THE ABOVE CONFLICT SITUATION. Strongly Strongly Disagree Neither Agree I am very concerned about the factory tying allocation to 1 2 3 4 5 6 7 the purchase of new product technology vehicle service. I am very motivated to resolve this conflict. 1 2 3 4 5 6 7 The factory's policy of tying allocation to the purchase of 1 2 3 4 5 6 7 new product technology for vehicle service is very important to the success of my dealership. LISTED BELOW ARE SOME WAYS OF RESOLVING THIS CONFLICT SITUATION IN TWO CASES. WE WOULD LIKE YOU TO EVALUATE HOU EACH RESPONSE OPTION WOULD RESOLVE THIS CONFLICT IN EACH CASE. BELOW TO THE RIGHT OF EACH RESPONSE, PLEASE ALLOCATE 100 POINTS AMONG THE RESPONSES SO THAT YOUR ALLOCATION REFLECTS THE RELATIVE EFFECTIVENESS OF EACH ALTERNATIVE RESPONSE TO RESOLVE THIS CONFLICT SITUATION. MAKE SURE THE POINTS ADD UP TO A TOTAL OF 100. Example: If there are 3 responses, the allocation could be 33-33-34 or 33-34-33 points if they are EQUALLY EFFECTIVE or 50, 30, 20 points or 20-30-50 points or 10-40-50 points, etc. if they are NOT EQUALLY EFFECTIVE. Allocate similarly for 4 or more responses. Make sure the points add up to a total of 100. Case A. Assuae you go along with the factory's policy of new product technology adoption. 1. Purchase the special tools and equipment to service a new model from the factory even though the new product technology is rarely used. 2. Purchase the special tools and equipment to service a new model from alternative supply sources that are approved and recommended by the factory. 3. Purchase the special tools and equipment to service a new model from alternative supply sources of your own choice. ^0® TOTAL Case B. Assume you do not go along with the factory's policy of new product technology adoption. (PLEASE ALLOCATE 100 POINTS IN THE SAME MANNER AS ABOVE). 4. Request that the factory use "turn and earn" solely to determine whether or not you receive your allocation. 5. Request that the factory permit you to use existing product technology to service new models if it is appropriate. 6. Do not purchase the new product technology and thus do not receive typical new models. 100 TOTAL If you go along with the factory's policy of new product technology adoption, what is the typical (most likely) way you resolve this conflict given the 3 responses listed in Case A above? (PLEASE CIRCLE THE APPROPRIATE RESPONSE NUMBER) Response 123 If you do not go along with the factory's policy of new product technology adoption, what is the typical (most likely) way you resolve this conflict given the 3 responses listed in Case B above? (PLEASE CIRCLE THE APPROPRIATE RESPONSE NUMBER) Response 4 5 6 224 I CONFLICT SITUATION 3: PARTS AVAILABILITY THE FACTORY (PRIMARY SUPPLIER) HAS REQUESTED THAT YOU PURCHASE ALL PARTS FOR YOUR PRIMARY MAKE FROM IT EXCLUSIVELY. How FREQUENTLY have you faced this particular conflict situation in the past 3 years? (PLEASE WRITE IN THE TOTAL NUMBER OF TIMES THIS CONFLICT HAS OCCURRED) ____________ times. Not Very Very How INTENSE would you characterize typical disagreements Intense Intense over this conflict, regardless of whether the disagreements 1 2 3 4 5 6 7 were settled or not? Unclear or Very To what extent does your current contract with the factory No Guidelines Clear (primary supplier) provide clear guidelines about how you At All Guidelines might resolve this type of conflict? 1 2 3 4 5 6 7 PLEASE CIRCLE THE NUMBER NEXT TO EACH QUESTION THAT BEST CORRESPONDS TO THE DEGREE TO WHICH YOU DISAGREE OR AGREE WITH EACH STATEMENT, KEEPING IN MIND THE ABOVE CONFLICT SITUATION. Strongly Strongly Disagree Neither Agree I am very concerned about parts availability for my primary 1 2 3 4 5 6 7 make of automobile. I am very motivated to resolve this conflict situation. 1234567 The availability of parts for my primary make from the 1 2 3 4 5 6 7 factory is very important to my dealership's success. LISTED BELOW ARE SOME WAYS OF RESOLVING THIS CONFLICT SITUATION IN TWO CASES. WE WOULD LIKE YOU TO EVALUATE HOW EACH RESPONSE OPTION WOULD RESOLVE THIS CONFLICT IN EACH CASE. BELOW TO THE RIGHT OF EACH RESPONSE, PLEASE ALLOCATE 100 POINTS AMONG THE RESPONSES SO THAT YOUR ALLOCATION REFLECTS THE RELATIVE EFFECTIVENESS OF EACH ALTERNATIVE RESPONSE TO RESOLVE THIS CONFLICT SITUATION. MAKE SURE THE POINTS ADO UP TO A TOTAL OF 100. Example: If there are 3 responses, the allocation could be 33-33-34 or 33-34-33 points if they are EQUALLY EFFECTIVE or 30- 50-20 points or 20-30-50 points or 10-40-50 points if they are NOT EQUALLY EFFECTIVE. Allocate similarly for 4 or more responses. Make sure the points add up to a total of 100. Case A. Assise you oo along with the factory's request (purchase all parts for your primary make from it exclusively). 1. Purchase all parts from the factory. 2. Purchase all parts from the factory except for those which require back ordering. 3. Purchase all parts from the factory except those which could be purchased less expensively from an alternative source. 100 TOTAL Case B. Assise you do not go along with the factory's request (purchase all parts for primary make from it exclusively). (PLEASE ALLOCATE 100 POINTS IN THE SAME MANNER AS ABOVE) 4. Purchase "slow-moving" parts from the factory and "fast-moving" parts from an alternative source of supply. ___________ 5. Purchase all parts for your primary make from an alternative source of supply if the parts are less expensive. 6. Do not purchase all parts for your primary make from the factory exclusively. 100 TOTAL If you go along with the factory's request (purchase all parts for your primary make from it exclusively), what is the typical (most likely) way you resolve this conflict given the 3 responses listed in Case A above? (PLEASE CIRCLE THE APPROPRIATE RESPONSE NUMBER) RESPONSE 1 2 3 If you do not go along with the factory's request (purchase all parts for your primary make from it exclusively), what is the typical (most likely) way you resolve this conflict given the 3 responses listed in Case B above? (PLEASE CIRCLE THE APPROPRIATE RESPONSE NUMBER) RESPONSE 4 5 6 225 CONFLICT SITUATION 4: CUSTOMER SATISFACTION INDEX (CSI) SURVEY THE FACTORY HAS INFORMS) YOU THAT THE CSI SURVEY IS DESIGNED TO ASSESS THE QUALITY OF YOUR SALES AND SERVICE RENDERED TO CUSTOMERS RATHER THAN THE QUALITY OF THE PRODUCT. How FREQUENTLY have you faced this particular conflict situation in the past 3 years? (PLEASE WRITE IN THE TOTAL NUMBER OF TIMES THIS CONFLICT HAS OCCURRED) _____________ times. Not Very Very How INTENSE would you characterize typical disagreements Intense Intense over this conflict, regardless of whether the disagreements 1 2 3 4 5 6 7 were settled or not? Unclear or Very To what extent does your current contract with the factory No Guidelines Clear (primary supplier) provide clear guidelines about how you At All Guidelines might resolve this type of conflict? 1 2 3 4 5 6► 7 PLEASE CIRCLE THE NUMBER NEXT TO EACH QUESTION THAT BEST CORRESPONDS TO THE DEGREE TO WHICH YOU DISAGREE OR AGREE WITH EACH STATEMENT, KEEPING IN MIND THE ABOVE CONFLICT SITUATION. Strongly Strongly Disagree Neither Agree I am very concerned about the factory's use of the CSI 1 2 3 4 5 6 7 survey. I am very motivated to resolve this conflict situation. 1 2 3 4 5 6 7 The use of the CSI survey is very important to the success 1 2 3 4 5 6 7 of my dealership. LISTED BELOW ARE SOME WAYS OF RESOLVING THIS CONFLICT SITUATION IN TWO CASES. WE WOULD LIKE YOU TO EVALUATE HOW EACH RESPONSE OPTION WOULD RESOLVE THIS CONFLICT IN EACH CASE. BELOW TO THE RIGHT OF EACH RESPONSE, PLEASE ALLOCATE 100 POINTS AMONG THE RESPONSES SO THAT YOUR ALLOCATION REFLECTS THE RELATIVE EFFECTIVENESS OF EACH ALTERNATIVE RESPONSE TO RESOLVE THIS CONFLICT SITUATION. MAKE SURE THE POINTS ADD UP TO A TOTAL OF 100. Example: If there are 3 responses, then the allocation could be 34-33*33 points or 33-33*34 points if they are EQUALLY EFFECTIVE or 30-50-20 points or 20-30-50 points or 10-40-50 points, etc., if they are NOT EQUALLY EFFECTIVE. Allocate similarly for 4 or more responses. Make sure the points add up to a total of 100. Case A. Assuae you go along with the factory's use of the CSI survey. 1. Work with the factory to isolate areas for improvement. 2. Try to understand the factory's perceptions of the CSI survey's use. 3. Use the NVIP (New Vehicle Inspection Program) to correct any errors before the vehicle is released to a customer. 4. Use customer relations personnel to call on customers who recently purchased a new vehicle and those whose vehicles were serviced to discover areas for improvement. 5. Offer some monetary incentives to customers to fill out the survey to yield positive ratings. 100 TOTAL Case B. Assume you do not oo along with the factory's use of the CSI survey. (PLEASE ALLOCATE 100 POINTS IN THE SAME MANNER AS ABOVE) 6. Develop your own interpretations of the CSI survey results. 7. Request that the factory redesign the CSI survey to allow customer to evaluate both the quality of the product and the quality of your sales and services. 8. Request that the factory solicit from dealers specific areas to improve the quality of the product given that you provide the service for the product when there are problems. 9. Protest the current use of the CSI survey. 100 TOTAL If you go along with the factory's use of the CSI survey, what is the typical (most likely) way you resolve this conflict given the 5 responses listed in Case A above? (PLEASE CIRCLE THE APPROPRIATE RESPONSE NUMBER) RESPONSE 12345 If you do not go along with the factory's use of the CSI survey, what is the typical (most likely) way you resolve this conflict given the 4 responses listed in Case B above? (PLEASE CIRCLE THE APPROPRIATE RESPONSE NUMBER) RESPONSE 6789 226 SECTION C: ABOUT TOUR DEALERSHIP In this last section, we would like to collect routine statistical information about your dealership for classification purposes. This information will be held strictly confidential and not identified with your dealership in any way. (PLEASE CHECK THE APPROPRIATE CATEGORY OR WRITE IN THE APPROPRIATE NUMBERS) 1. How would you describe your trade area? Urban; Suburban; Rural or Small Town 2. In your trade area, how many other franchised dealers are there for your primary make of vehicle? dealers 3. How long has your dealership been in business under its present name? years months 4. How long has your dealership been in business at its present location? years months 5. In how many dealerships are you the sole owner or a partner? dealerships 6. Compared to other franchised dealers in your trade area, how would you describe your dealership with respect to new car sales volume? ______ high volume franchise ______ moderate volume franchise ______ low volume franchise 7. Of your total part purchases over the past 3 years, approximately what percent is accounted for by your primary make? _____ percent 8. Approximately, what is your dealership's AVERAGE NUMBER OF SALESPEOPLE over the oast 3 years? ______ years 9. Approximately, what is your dealership's AVERAGE AIMJAL TURNOVER (new car sales volume in units) over the past 3 years? ______ units 10. Approximately, what is your dealership's AVERAGE AIMJAL SALES* REVENUE for new cars over the past 3 years? $________________________ 11. Approximately, what is your average net profit as a PERCENTAGE OF SALES over the past 3 years? (e.g., 1%, 2%, etc.) percent 12. Please indicate the approximate PERCENTAGE of your total new car business (oast 3 year average) represented by each of your suppliers: 1 ofSales Vol one GM ___ X Ford ___ X Chrysler ___ X Jeep/Eagle __ X Honda ___ X Toyota ___ X Nissan ___ X Mazda ___ X VW & Other German Cars ___ X Subaru ___ X Others ________________ ___ X ___ X ___ X ___ X 100 X 13. Please indicate which of the following groups you are a member of and the number of years a member: National Automobile Dealers Association years Your regional automobile dealer association years Michigan Automobile Dealers Association years ________ A special interest association, such as years the 20 Group, Ford Dealer Alliance, etc. 227 14. If you would like to make any comments about factory* deal er relations or any other aspects of the automobile business, please write in the space provided. THANK YOU FOR YOUR HELP. PLEASE INSERT THE QUESTIONNAIRE IN THE POSTAGE-PAID ENVELOPE PROVIDED FOR YOUR CONVENIENCE AND DROP IT IN THE MAIL IF YOU WOULD LIKE A COPY OF THE RESULTS OF THIS SURVEY, PLEASE PROVIDE YOUR MAILING ADDRESS IN THE SPACE PROVIDED BELOW. MAILING ADDRESS: _ _ ___________ APPENDIX C QUESTIONNAIRE TYPE B 228 INSTRUCTIONS Below are several conflict situations involving factory-dealer relationships. For each situation, you are asked to respond to several questions. Please view each situation as an independent event. Please respond to this questionnaire on the basis of of a dealer's PRIMARY SUPPLIER. SECTION A: DEALER'S CONFLICTS WITH ITS PRIMARY SUPPLIER In this section, we would like you to evaluate the EXPECTS) IMPACT of dealer responses to each conflict situation on its performance. (PLEASE CIRCLE THE APPROPRIATE NUMBER) CONFLICT SITUATION 1: NEW VEHICLE ALLXATION AND DISTRIBUTION THE FACTORY (PRIMARY SUPPLIER) HAS AN UNWRITTEN POLICY WITH RESPECT TO NEW VEHICLE ALLOCATION AND DISTRIBUTION. THAT IS, TO GET THE DESIRS) QUANTITY OF A PARTICULAR POPULAR MODEL, A DEALER MUST ACCEPT MORE THAN IT WOULD LIKE , TO OF OTHER MODELS. _______________________________________________________ Very Very How FRE QUENTLY do you think this conflict is encountered in Infrequent Frequent factory'-dealer relationships? 1 2 3 4 5 6 7 Not Very Very How INTENSE would you characterize typical disagreement over Intense Intense this conflict situation, regardless of whether the 1 2 3 4 5 6 7 disagreements are settled or not? PLEASE CIRCLE THE NUMBER NEXT TO EACH QUESTION THAT BEST CORRESPONDS TO THE DEGREE TO WHICH YOU DISAGREE OR AGREE WITH EACH STATEMENT, KEEPING IN MIND THE ABOVE CONFLICT SITUATION. Strongly Strongly Disagree Agree A dealer is very concerned about the factory's new vehicle 1 2 3 4 5 6 7 allocation and distribution policy. The factory's new vehicle allocation and distribution policy 1 2 3 4 5 6 7 can affect a dealership's success. A dealer is usually very motivated to resolve this conflict 1 2 3 4 5 6 7 situation. LISTED BELOW ARE SONE WAYS OF RESOLVING THIS CONFLICT SITUATION 1fl TH THE FACTORY. PLEASE INDICATE YOUR EXPECTATIONS OF THE IMPACT OF EACH RESPONSE ON DEALER PERFORMANCE. 1. Dealer accepts the factory's typical allocation of new models. Extremely Extremely Negative No Positive Impact Impact Impact * Impact on new car sales volume (total): 1 2 3 4 5 6 7 * Impact on trade area market share: 1 2 3 4 5 6 7 * Impact on profits: 1 2 3 4 5 6 7 2. Dealer uses a computerized locating system or the conventional location system to find special models and engages in dealer trading. Extremely Extremely Negative No Pos i t i ve Impact Impact Impact * Impact on new car sales volume (total): 1 2 3 4 5 6 7 * Impact on trade area market share: 1 2 3 4 5 6 7 * Impact on profits: 1 2 3 4 5 6 7 3. After locating a vehicle, the dealer purchases it outright. Extremely Extremely Negative No Positive Impact Impact Impact * Impact on new car sales volume (total): 1 2 3 4 5 6 7 * Impact on trade area market share: 1 2 3 4 5 6 7 * Impact on profits: 1 2 3 4 5 6 7 4. Dealer concentrates its efforts on buying and selling used cars when new models are not avaiIable. Extremely Extremely Negative No Positive Impact Impact Impact * Impact on new car sales volune (total): 1 2 3 4 5 6 7 * Impact on trade area market share: 1 2 3 4 5 6 7 * Impact on profits: 1 2 3 4 5 6 7 229 5. Dealer orders more than the desired quantity of new cars to ensure that iit receives the desired quantity. Extremely Extremely Negative No Positive Impact Impact Impact * Impact on new car sales volume (total): 1 2 3 4 5 6 7 * Impact on trade area market share: 1 2 3 4 5 6 7 * Impact on profits: 1 2 3 4 5 6 7 6. Dealer appeals to the district manager to increase its allocation. Extremely Extremely Negative No Positive Impact Impact Impact * Impact on new car sales volume (total): 1 2 3 4 5 6 7 * Impact on trade area market share: 1 2 3 4 5 6 7 * Impact on profits: 1 2 3 4 5 6 7 7. Dealer engages in demand forecasting for different models and builds inventory for those models which will become popular in the future. Extremely Extremely Negative No Positive Impact Impact Impact * Impact on new car sales volume (total): 1 2 3 4 5 6 7 * Impact on trade area market share: 1 2 3 4 5 6 7 * Impact on profits: 1 2 3 4 5 6 7 8. Dealer protests the factory's unwritten policy in regards to new vehicle al location and distribution. Extremely Extremely Negative No Pos i t i ve Impact Impact Impact * Inpact on new car sales volume (total): 1 2 3 4 5 6 7 * Impact on trade area market share: 1 2 3 4 5 6 7 * Impact on profits: 1 2 3 4 5 6 7 CONFLICT SITUATION 2: ADOPTION OF HEM PRODUCT TECHNOLOGY THE FACTORY (PRIMARY SUPPLIER) HAS INFORMED THE DEALER THAT IT WILL NOT RECEIVE ITS TYPICAL ALLOCATION OF NEU MODELS UNLESS IT PURCHASES THE NEW PRODUCT TECHNOLOGY DESIGNED TO SERVICE THEM. Very Very How FREQUENTLY do you think this conflict is encountered in Infrequent Frequent factory-dealer relationships? 1 2 3 4 5 6 7 Not Very Very How INTENSE would you characterize typical disagreement over Intense Intense this conflict situation, regardless of whether the 1 2 3 4 5 6 7 disagreements are settled or not? PLEASE CIRCLE THE NUMBER NEXT TO EACH QUESTION THAT BEST CORRESPONDS TO THE DEGREE TO WHICH YOU DISAGREE OR AGREE WITH EACH STATEMENT. KEEPING IN MIND THE ABOVE CONFLICT SITUATION. Strongly Strongly Disagree Neither Agree A dealer is very concerned about the factory's practice of 1 2 3 4 5 6 7 tying allocation to the purchase of new product technology for vehicle service. The use of new vehicle allocation to enforce dealer purchase 1 2 3 4 5 6 7 of new product technology can affect a dealership's success. A dealer is very motivated to resolve this conflict 1 2 3 4 5 6 7 situation. 230 LISTED BELOW ARE SOE WAYS OF RESOLVING THIS CONFLICT SITUATION WITH THE FACTORY. PLEASE INDICATE YOUR EXPECTATIONS OF THE IMPACT OF EACH RESPONSE ON DEALER PERFORMANCE. Dealer purchases the special tools and equipment to service a new model from the factory even though the new product technology is rarely used. Extremely Extremely Negative No Pos i t i ve Impact Impact Impact * Impact on new car sales volute (total): 1 2 3 4 5 6 7 * Impact on trade area market share: 1 2 3 4 5 6 7 * Impact on profits: 1 2 3 4 5 6 7 Dealer purchases the special tools and equipment t<> service new models from alternative supply sources that