n 1 c l . I. I | I ‘l‘-.. l A. .. Wu. 1" I u! . J I .. . i 0' 1/ ‘Al? 70.. ill MI ,. I ‘nhvll ROOT! if!» ‘ll..a I -I. ‘ .J .I .-.§ébn...};..wzvu.fi.z. - l. .0 IO"! A»- :n.uus u . 9. . q! ”A t'flluflrlyg.‘ Hull - . 9.! . A0. Ali‘- gfl...ff 1.4%“... . r -P x ‘2 . .-.. . ? I. J z; 1-- I l. I n a y... H n. y.» 1"IRIIA ”an-‘fiu'flrlh‘qwult [ If ; ‘5 . Ann! - A 4 :1 I. v. o , . . 9.1!. m“ t :3 . :2 6.2-9. I . - a..- 1...? -.. {Illhlol )D' . T «t: u. ?.u . . - Z . .3-“ .. H.-. h. .. . R I! 3d:fl’.rtl-1lln¢o¢ftu. . .o. .o n». ‘ I- Z 70.. y'all . - L 1 t to II. n . u..VI . .. l o: P. . :.| . x . ul . Us . ) OI ‘IA - A . \nb - .y . 1 v q 'r . 0 I. t r! I. 4 I . .r 3’9}. o g Y. Q. ‘ . i {.tfiltlfab .. -- L. .. ., . A . I'll I nl.l a v .1 V’DL’V...I..U%IDI‘JIIF .‘ouhll. liflhniltc . “It. In. 1 . ... Ilfl.....:f.l1¢(l‘; .u lfill no“... u. .7. - udbtuflinllwu I ‘..lvn.. ‘v.0.|..t(".? an...” t; I. I luv '1’]. ... u C: . .3- o .3. [thnqlflu‘ffhvlzs‘r‘tthu f ., . ”0‘10””! '1‘. .11.” .‘lco‘a . 9:1 in???” lull I nu . \Ill ‘1 5.1!!" will!!! )IIF , Nihilnltllludo I )nl'l 1H1! . . .vll< 1" ‘4 . . ‘ ‘II"»’J.C¢.IJ’. 'L It”, . 'Ill. ”‘I’Iofltvll . 4..“ ,Y . v‘ . . ‘ . .10. [Unilm “WW; I: ( |V I! ’ .Iila ‘ :lllrj‘uhul.’1xlllf » - D. If J‘vsffhrfirhla (Tun?! gar; Hun: . .. I s I .l. J>Muuu hurlgl '4'. .v? . hm. (Lkrjbi I... 3.1! I . JVrIlllhlu $0? .| 4’1'11- if a. p “03.4 .I’lfll 4...: ’v‘t' Ill-.l. ‘ . . ‘ ,..|Ic....Uu.dVII. 3n . {lid [3420: "o. . i t . “gimme u. llllfllflfljfljfl‘lflfllflflflmfllflmll niversity CHANNEL MEMBER SATISFACTION, PERFORMANCE AND DEPENDENCE: AN EMPIRICAL INVESTIGATION presented by Mary Christine Lewis has been accepted towards fulfillment .‘ of the requirements for Ph.D. Marketing degree in AV! 5” / Major pro essor Date Y MS U i: an Afl'mnan'vc Action/Equal Opportunity Institution 0-12771 9-0. MSU RETURNING MATERIALS: Place in book drop to LIBRAmEs remove this checkout from 1—13—- your record. FINES will be charged if book is returned after the date stamped below. we (w. gm ,9- . DAMAGSPR :9 13:02-8 1998 CHANNEL MEMBER SATISFACTION, PERFORMANCE AND DEPENDENCE: AN EMPIRICAL INVESTIGATION By Mary Christine Lewis 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 Administration 1987 °Copyright by Mary Christine Lewis 1987 All Rights Reserved ABSTRACT CHANNEL MEMBER SATISFACTION, PERFORMANCE AND DEPENDENCE: AN EMPIRICAL INVESTIGATION By Mary Christine Lewis Many companies rely on intermediaries to perform some of the business functions needed to successfully market to their final customers. While the arrangement benefits both parties financially, it does create some unique management challenges for the supplier because the intermediaries are independent firms. To be effective, the supplier needs strategies for managing its intermediaries. Most empirical studies of marketing channels investigated power, conflict and sometimes, satisfaction. Very few studies investigated channel member performance or financial dependence, the variables of utmost concern to practitioners. So, the findings from prior studies had limited managerial usefulness because managers have little interest in channel member satisfaction, except as it relates to these key economic variables and no empirical research had investigated these interrelation- ships. Nevertheless, previous researchers implied that satisfied intermediaries perform at a higher level and also that satisfaction encourages greater financial dependency on the supplier. Therefore, the dissertation's purpose was to conceptualize the interrelationships among channel member satisfaction, dependence and performance, and to empirically test this theory. A mail survey of 204 Mary Christine Lewis qualified franchisees was conducted in the fast food industry. Because several methodological weaknesses have been associated with prior chan- nels research, the author paid strict attention to construct validity, and a two-step causal modeling approach was used to analyze the data. The findings indicated that satisfaction does not lead to perfor- mance or dependence. Rather, financial dependence leads to performance and to the reinvestment of profits in the business. Performance leads to satisfaction but not directly. The relationship between the two vari- ables is mitigated by the extent to which intermediaries assign credit or blame for their performance to the supplier. Thus, increasing intermediaries' satisfaction does not appear to be a prime strategy for improving their performance. But, strategies that encourage financial dependence should be adapted because inter- mediaries are self-interested parties. Given a large financial stake in the business, intermediaries are motivated to perform as a means of protecting and growing their assets. ACKNOWLEDGEMENTS It would be an unusual.person who could successfully complete a work of this magnitude without the support and encouragement of others. Two organizations and several individuals lent support to this project. Taco Bell, a division of PepsiCo, contributed to the dissertation by allowing me to conduct a mail survey of its franchisees. I would like to extend a special thanks to four individuals at Taco Bell who were parti- cularly helpful with the research. John Martin, president of Taco Bell, authorized and personally supported the project. Bruce Klein, formerly vice president of operations at Taco Bell, was the project's initial sponsor, and as such Bruce was instrumental in generating financial and managerial support for the research. Dave Bennet, a regional vice president of operations, and his staff gave me important background information about the industry and the company. Finally, Ron Watson, a personnel manager, arranged for me to use a Taco Bell office for the purpose of conducting telephone follow-up with the franchisees who had not returned questionnaires. Notably, the response rate to the survey was 52.5 percent, an excellent result that was due in part to the follow-up phone calls. The PepsiCo Foundation provided financial support for the research in the form of a doctoral fellowship. I am extremely appreciative of the PepsiCo Foundation's contribution because it enabled me to devote considerable time and effort to the research. I am especially indebted to Professor Douglas M. Lambert of the University of South Florida in Tampa. While at Michigan State University, he held the PepsiCo Professorship and chaired my dissertation, a respon- sibility he voluntarily continued after moving to Florida. Although Professor Lambert made many important contributions to my dissertation, his influence on my personal and professional development far exceeds his involvement in this project. Professor Lambert maintains a high level of commitment to students and practitioners. He was instrumental in my decision to pursue a doctorate and in securing the PepsiCo fellowship that eased the financial burden associated with my decision, and enabled me to concentrate on research during the doctoral program. Also, Professor Lambert is continually involved in scholarly research and is dedicated to the belief that academic research must be designed to yield managerially useful findings. With his uncompromising professional standards and exemplary personal conduct, Professor Lambert provides an excellent example for young scholars to emulate. I would like to express my sincerest appreciation to the other members of my dissertation committee, Professors Forrest S. Carter and R. Dale Wilson for their valuable suggestions. I owe a special thanks to Ron Savitt, Beckley Professor of American Business, University of Vermont, former chairman of the Department of Marketing and Transportation, for his encouragement and counsel. The fact that Professor Savitt headed the department was a key factor in my decision to enroll in Michigan State's doctoral program. Marilyn "Sam“ Miller must also be recognized for her role in the research. Sam typed and retyped the manuscript and her extraordinary vi skill and professionalism were key factors in my ability to complete this dissertation within the required deadlines. Finally, my husband Steven deserves Special thanks for his constant support and encouragement throughout the entire doctoral program. Both Steven and our daughter, Amy, made many sacrifites on my behalf for which I am grateful. vii TABLE OF CONTENTS Page LIST OF TABLES 00.000.00000.000000000.0000..OOOOOOOOOOOOOOOOOOOOOO Xi LIST OF FIGURES .0000.000.000...00.0.0.0...OOOOOOOOOOOOOOOOOOOOOOO Xii CHAPTER 1: INTRODUCTION .................................................. 1 THE BUSINESS PROBLEM .......................................... REVIEW OF RELEVANT THEORY AND PRIOR RESEARCH .................. scum RESEARCH PURPOSE AND OBJECTIVES ............................... HYPOTHESES TESTED ............................................. IO METHODOLOGY ................................................... 14 MANAGERIAL RELEVANCE .......................................... 16 ORGANIZATION OF THE DISSERTATION .............................. 16 CHAPTER 2: REVIEH OF THE LITERATURE O...OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOI... 18 CONSTRUCT DEFINITION AND OPERATIONALIZATION ................... 20 Channel Men‘ber satisfaction 00.0....OOOOOOOOIOOOOOOOOOOOOOO 20 Channel Member Performance ................................ 26 Channel Member Dependence ................................. 28 CURRENT THEORY AND RELATED EMPIRICAL EVIDENCE ................. 31 Power and Conflict ........................................ 31 Satisfaction and Conflict ................................. 37 Conflict and Performance .................................. 39 Satisfaction and Performance .............................. 41 Satisfaction and Dependence ............................... 43 Dependence and Performance ................................ 45 Summary of Current Theory ................................. 46 viii METHODOLOGICAL ISSUES 00.00000000000500000000000000000.0000oooo Single Channel system OOOOOOOOOOOOOOOOOOOOOOOCOOCOOOOOOI... Key Informant MethOd OOOOOOOOOOOOOOIOOOOOOCOOOOOOOOOOOOOOOO SUMMARY OF THE LITERATURE AND MAJOR CONCLUSIONS ............... CHAPTER 3: RESEARCH DESIGN 00......OCOOCOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO OVERVIEW OF THE RESEARCH PROCESS 000......OOOOOOOOOOOOOOOOOOOOO ReseaFCh Objectives OOOOOOOOOOOOOOOOOOOIOOOOOOOOOOOOOOOOOOO The Conceptual Framework .................................. ResearCh MethOdOIOQy OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO THE CONCEPTUAL FRAMEHORK OOOOOOOOOOOOOOOOOOOOOOOOOOOIOOOOOOOOOO Research Hypotheses ....................................... Definition and Operationalization of Variables ............ Performance ............................................ Satisfaction ........................................... Dependence ............................................. Reinvestment ........................................... Attributions of Responsibility ......................... RESEARCH METHODOLOGY 0.00.0000...0.0.0...OOOOOOOOOOOOOOOOOOOOOO ResearCh setting OOOOOOOOOOOOOOOOCOOOOOOOOOOOOOOOOOCOOOOOOO Measures Emp10yed OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO Performance 0.0000000000000000000000...0.0.0.0000...O... Dependence .00...OOOOOIOOOOOOOOOOIOOOOOOOOOOOOOOOOOOOOOO Reinvestment Intentions 00.0.00...OOOOOOOOOOOOOOO0.0.... Attribution Of Respondents O00.0.0.0...OOOOOOOOOOOOOOOOO satisfaction OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO0.... Discussions with Franchisor Management and Industry Experts ......................‘O.................. In-Depth Interviews with Franchisees ...................... Data COVIECtion Methads OOOOOOOOOOOOOOO00.0.00...0.0.0.0... Data Analy515 0.0.0.....0...00......OOOOOOOOOOOOOOOOOOOOOOO SUNMARY 00.0.00...OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO ix 48 48 49 51 73 76 78 79 85 CHAPTER 4: RESEARCH FINDINGS ............................................. MEASUREMENT MDDEL ............................................. STRUCTURAL MODEL .............................................. SUMMARY OF THE FINDINGS ....................;.................. CHAPTER 5: DISCUSSION Am CONCLUSIONS 0.00.00...000......OOOOOOOOOOOOOOOOO REVIEW OF THE RESEARCH OBJECTIVES O...00.000000000000000000000. MAJOR CONCLUSIONS AND IMPLICATIONS OOOOOOOOOOOOOOOOO0.0.0.0.... Understanding th Marketing Channels are Formed ........... The Economic and Behavioral Dimensions are InterPEIated O0.00000000000000000000000000000000000000000 Satisfaction Does Not Influence Performance ............... A Causal Modeling Approach is Appropriate ................. MAJOR CONTRIBUTIONS 0.00.00.00.00..0...OOOOOOOOOOOOOOOOOOOOOOOO SUGGESTIONS FOR FUTURE RESEARCH 0......OOOOOOOOOOOOOOOOOOOOOOO. APPENDIXA OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO BIBLIOGRAPHY OOOOOOOOOOOOOOOOOO0.00.0.0...OOOOOOOOOOOOOOOOOOOOOOOO 86 86 98 107 109 109 112 112 113 113 114 115 117 122 134 Table é—l Table 4-1 Table 4-2 Table 4-3 Table 4-4 LIST OF TABLES Measures of Channel Member Satisfaction .............. Measurement Made] 00.000000...OOOOOOOOOOOOOOOOOOOOOOOO Similarity Matrix for Dimensions of satiSfaCtion 0....0.0.0.000...OOOOOOOOOOOOOOOOOOOOOOOO Correlations Between Constructs ...................... Fit Of the Structural "Odels O...OOOOOOOOOOOOOOOOOOOOO xi Page 24 88 96 97 105 Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure 2-1 3-1 3-2 4-1 4-2 4-4 4-5 4-6 4-7 4-8 LIST OF FIGURES The Interrelationships Among Key Variables in ChanHEI Management 0.0.0...OOOOOOOOOOOOOOOOOOOOOOO The Interrelationships Among Key Variables in Channel Management 0......OOOOOOOOOOOOOOOOOOOOOOOO The Prevailing View OOOOOOOOOIOOOOOOOOOOOOOOOOOOOOOO. An Alternative Model A Measure of Satisfaction with Financial Returns 0.0...O...0......OOOOOOOOOOOOOOOOOOOOO0...... The Climate/Interaction Scale ....................... The Marketing Scale ................................. The Product/Access Scale ............................ The Operations/Personnel Scale ...................... The Prevailing View ................................. The Prevailing View: Added Link from DEP$$ to RstT OOOOOOOOOOOOIOOOOOOOOOOOOOOOOOOOOOOOO. The Alternative Model The Alternative Model Respecified: DEIetEd RstT to PERFM 00......OOOOOOOOOOOOOOOOOOOOOO The Alternative Model Respecified: Added AGAIN to RstT OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO xii Page 31 47 61 62 73 91 92 94 95 99 101 102 103 104 CHAPTER 1 INTRODUCTION Theory and research pertaining to marketing channels has developed into two distinctly different approaches. One approach focuses primarily on the economics of channels while the other emphasizes the channel's behavioral dimensions. Since 1960, the behavioral approach has dominated the literature and has been sharply criticized for its lack of manageri- ally useful findings. Several marketing scholars have urged channel researchers to integrate the behavioral and economic approaches to more realistically represent the manner in which marketing channels operate (Frazier 1983b; Frazier and Summers 1984; Gaski 1984; Stern and Reve 1980). The objective of this research is to integrate the two approaches. Specifically, the research investigates the relationships between channel member satisfaction, a behavioral variable, and channel member performance and dependence which are economic variables. The balance of this chapter outlines the objectives and design of the research which was conducted in the fast food industry. The chapter is comprised of the following sections: 1) the business problem; 2) rele- vant theory and prior research; 3) the research purpose; 4) hypotheses tested; 5) methodology employed; 6) managerial relevance; and, 7) organi- zation of the dissertation. 2 THE BUSINESS PROBLEM Many firms rely on intermediaries to perform some of the business functions required to successfully tap their target markets. However, using intermediaries lessens the supplier's control over its business, especially those functions performed by intermediaries. Because the channel participants take on complementary roles, they become interdepen- dent. This provides the stimulus for cooperation but also establishes the basis for conflict. A supplier needs effective strategies for managing its interme- diaries. Hhile direct control over the intermediary may not be feasible, the supplier may seek to direct, encourage, and influence the inter- mediary‘s performance. Specifically, management wants the intermediary's performance to facilitate the achievement of the firm's objectives. In addition, the supplier wants to encourage greater financial dependency. Management can choose to coerce and threaten intermediaries to comply with the firm's programs and objectives or to attempt to satisfy the intermediary's needs as a means of encouraging a high level of performance and dependence. Intuitively, the latter strategy holds greater appeal. Previous empirical research involving related constructs implies that highly satisfied intermediaries will perform at a high level, and also may maintain a high level of dependence, if not increase their dependence on the supplier (Brown and Frazier 1978; Dwyer 1980; Hunt and Nevin 1974; Lusch 1976a, 1977; Michie and Roering 1978; Rosenberg and Stern 1971; Wilkinson 1981). However, no empirical research has been reported that investigated the interrelationships among channel member satisfaction, performance and dependence. 3 Management needs to know whether or not a high level of channel member satisfaction leads to improved performance and increased dependence. This is because producing and sustaining a high level of intermediary satisfaction is not without cost. .Rather it requires a very deliberate effort to allocate scarce resources in an Optimum manner. In addition to influencing the distribution of the firm's resources, a commitment to intermediary satisfaction also may require an increase in the total resources employed. Subsequently, decisions regarding the level and allocation of resources will impact the firm's costs. Whether the cost is justified depends upon the extent to which a high level of satisfaction motivates the intermediary to perform at a high level, and to become more dependent upon the firm and thereby increases the supplier's profits. REVIEW OF RELEVANT THEORY AND PRIOR RESEARCH After 1960, the focus of channel theory and research shifted to the behavioral dimensions, and the two major constructs of interest were power and conflict. To some degree channel member satisfaction was examined, particularly in relationship to power and conflict. Despite being dominant for nearly 30 years, the behavioral research stream has made few important contributions to channel theory or practice (Frazier 1983b; Gaski 1984; Stern and Reve 1980). Five reasons have been cited for these disappointing results: 1) an inadequate conceptual frame- work; 2) undefined constructs; 3) unsatisfactory operationalizations of the constructs; 4) questionable reliability and validity of the measures; and, 5) inappropriate statistical analysis (Gaski 1984). 4 In short, the theoretical foundatibn of the behavioral approach is weak and the methodological practices lack rigor. Consequently, some marketing scholars have questioned whether it has provided any useful information, or simply served to obscure our thinking about channel management (Frazier and Summers 1984; Gaski 1984; Lambert 1978; Stern and Reve 1980). Interest in channel member satisfaction stemmed from interest in power and conflict. According to the theory, the use of coercive power would decrease satisfaction while noncoercive power would increase satisfaction (Dwyer 1980; Hunt and Nevin 1974; Lusch 1976a, 1977; Michie and Roering 1978; Ross and Lusch 1983). It was postulated that conflict was negatively related to satisfaction (Dwyer 1980; Rosenberg and Stern 1971; Wilkinson 1981). In early studies, channel member satisfaction was not explicitly defined, and it was operationalized with psychometrically poor single- item measures (Brown and Frazier 1978; Hunt and Nevin 1974; Rosenberg and Stern 1971; Walker 1972). Lusch (1977) recognized the multi-faceted nature of the satisfaction construct. According to Lusch a channel member could be more or less satisfied with a variety of issues. “Lusch initiated the practice of using a multi-item role performance measure of satisfaction. Lusch asked intermediaries to use a Likert scale to indicate their beliefs about the supplier's level of performance of each service. Ruekert and Churchill (1984) proposed the following definition: Channel member satisfaction comprises the domain of all characteristics of the relationship between a channel member and another institution in the channel which the member finds rewarding, profitable, instrumental and satisfying...or unsatisfying. 5 The general concensus is that channel member satisfaction is a multi-dimensional construct. The literature suggested that some of its key dimensions may include satisfaction with supplier role performance, the intermediary's financial returns and possibly, the communication between the two parties (Dwyer 1980; Frazier 1983b; Lusch 1977; Michie and Roering 1978; Ruekert and Churchill 1984). Although both Lusch and Michie reported adequate reliability for their measures, only Ruekert and Churchill assessed discriminant validity. Widespread inattention to discriminant validity has led some. marketing scholars to question whether the measures of key constructs are valid (Frazier and Summers 1984; Gaski 1984). Gaski cautioned scholars and practitioners to critically review empirical studies and carefully interpret the findings. Empirical research has consistently supported the hypothesized relationships between power and channel member satisfaction, and conflict and satisfaction. Specifically, empirical evidence supported a negative relationship between satisfaction and coercive power and conflict, respectively (Brown and Frazier 1978; Dwyer 1980; Hunt and Nevin 1974; Lusch 1977; Michie and Roering 1978; Rosenberg and Stern 1971; Wilkinson 1981). Empirical research also provided evidence of a positive relation- ship between the supplier's use of noncoercive power and the channel intermediary's satisfaction (Brown and Frazier 1978; Dwyer 1980; Hunt and Nevin 1974; Lusch 1977; Michie and Roering 1978; Ross and Lusch 1983; Wilkinson 1981). However, in many studies noncoercive power and channel member satisfaction were both represented by role performance measures, so it is not surprising that they would be positively correlated. Hunt and Nevin 6 conducted the first empirical study of noncoercive power and initiated the practice of using a role performance measure. Subsequent studies of noncoercive power employed a similar measure (Lusch 1976a,b, 1977; Michie and Roering 1978; Ross and Lusch 1983; Wilkinson 1981). The noncoercive power sources construct is comprised of four of the five types of power contained in French and Raven's (1959) classification scheme. The four types are: reward, referent, expert, and legitimate. Researchers admitted the difficulty of distinguishing these four types of power empirically, so they created the noncoercive power sources con- struct to capture them (Hunt and Nevin 1974). Hunt and Nevin justified the use of a role performance measure to operationalize the construct based on their belief that good performance: helped suppliers position themselves as experts; legitimize suppliers' efforts to gain power; and lead intermediaries to willingly accept a subordinate role. In other words, if intermediaries are satisfied with the supplier's role perfor- mance, they will be more accepting of the supplier's attempts to influence their performance. Thus, it appears that Hunt and Nevin actually captured the role performance dimension of the channel member satisfaction construct. Not surprisingly, the legitimacy of the noncoercive power sources construct has been questioned (Frazier and Summers 1984). Channel member performance is the variable of utmost concern to practitioners, yet it has been all but ignored in the literature. This is largely due to the dominance of the behavioral approach to channels research in recent decades. Even so, it is difficult to understand, given that a desire to influence the intermediary's thinking and behavior provided the initial rationale for investigating power. It would seem 7 logical for behavioral researchers to ektend their scope to include channel member performance. The disproportionate emphasis given to behavioral research has been criticized because the behavioral variables are difficult to operationa— lize (Gaski 1984; Lambert 1978). Other scholars have argued that the economic and behavioral dimensions of the channel exist in conjunction with one another. Thus, taking either a behavioral or economic approach does not accurately reflect the manner in which channels operate (Frazier I983b; Frazier and Summers 1984; Stern and Reve 1980). No consensus definition of channel member performance exists. Four empirical studies have been reported that investigated channel member performance. Lambert (1978) conducted an exploratory study of channel management practices in eighteen manufacturing firms. He found that performance measurement varied among firms, and with the type of intermediary. Kelly and Peters (1978) and Lusch (1976b) studied the relationship between vertical conflict and the intermediary's financial performance, as perceived by the intermediary. Both studies reported an indirect correlation between conflict and performance. That is, a high level of performance was related to a low level of conflict (Kelly and Peters 1978; Lusch 1976b). Gaski and Nevin (1985) investigated the intermediary's performance from the supplier's perspective. They defined channel member performance as: ...the degree to which a supplier's relationship with a dealer contributes to fulfillment of the supplier's objectives. 8 Gaski and Nevin employed a two-item perceptual measure of performance rather than hard data obtained from the supplier. But, Gaski and Nevin offered no explanation for the use of a perceptual performance measure. Their study investigated the relationship between exercized and unexercized power sources and performance, and found no significant differences. Another variable important to practitioners is channel member dependence. Like performance, dependence has received scant attention in the literature. Channel member dependence has been the subject of two empirical studies by Etgar (1976b) and Phillips (1981). It has been operationalized as an economic variable with three objective measures and one perceptual measure. All four measures were devised by Etgar. Phillips examined the reliability and validity of the perceptual measure and concluded that it was not valid. He used the multiple informant method and found more variance associated with method and random error than with trait. However, two of Etgar's objective measures of dependence evidenced a weak positive correlation with supplier power. The two dependence measures were: percent of income from casualty insurance and the total number of insurers represented. Agents who were heavily dependent upon casualty insurance, and had few insurers yielded more willingly to insurers' controls over their business (Etgar 1976b). Two methodological issues characteristic of the behavioral research stream deserve mention. The first is the custom of using a single channel system as the research setting for empirical work involving behavioral variables (Brown and Frazier 1978; El-Ansary and Stern 1972; Etgar 1976a,b; Frazier 1983a; Frazier and Summers 1984; Hunt and Nevin 1974; Gaski 1986; Gaski and Nevin 1985; Lusch 1976a,b, 1977; Lusch and 9 Brown 1982; Michie and Roering 1978; RoSenberg and Stern 1971; Ross and Lusch 1983; Ruekert and Churchill 1984; Wilkinson 1974, 1981). This practice has been justified on the grounds that most of the constructs under investigation are in the developmental stage, and also because many must be operationalized with empirically derived measures (Ruekert and Churchill 1984). Differences in the operating characteristics of channels makes it imperative to develop measures of constructs like satisfaction using data collected from a single channel. In other words, data collected from multiple channels may be incomparable (Gaski 1985; Lambert 1978). The second methodological issue worthy of note is the use of the single key informant method of data collection. Phillips (1981) argued that multiple key informants should be used to facilitate the decompo- sition of variance into trait, methods and random error. However, many intermediaries are single proprietorships and the owner actively manages the business. Thus, while channel researchers have acknowledged Phillips' point, they continue to use the single key informant method because often it is the only logical choice. To use multiple informants when only one individual is knowledgeable about the relevant issues, and qualified to respond to a survey, does not make sense. In all likeli- hood, it would simply add "noise" to the data. The important point seems to be that researchers should identify all qualified respondents prior to implementing field research. If there are multiple qualified respondents then they should be included in the study. RESEARCH PURPOSE AND OBJECTIVES The purpose of this research was to empirically examine the rela- tionships among channel member satisfaction, performance and dependence. 10 The research was an attempt to integrate the behavioral and economic approaches to channel research. Notably, it was the first empirical test of the relationships among the three constructs of interest. The specific objectives of the research were: 1. To integrate the behavioral and economic approaches to marketing channels theory; 2. To conceptualize the interrelationships among channel member performance, satisfaction and dependence in a theoretically and managerially sound way; and, 3. To empirically test the conceptual scheme. The specific research questions that guided the research were: 1. Does channel member satisfaction lead to performance? 2. Does channel member satisfaction directly influence the member's financial dependence? 3. Does channel member satisfaction lead to the reinvestment of profits? 4. Does financial dependence on a channel relationship directly influence the channel member's performance? HYPOTHESES TESTED Before the researcher formulated formal research hypotheses, it was necessary to develop a conceptual framework to provide scope and direction for the research. This involved conceptualizing each of the constructs of interest. It also meant that each construct had to be explicitly defined and suggestions regarding the operationalization of the constructs also were required. Channel member satisfaction was investigated from the intermediary's perspective and conceptualized as a multi-dimensional multi-item construct. It was defined as the extent to which a channel intermediary is satisfied or dissatisfied with various aspects of the relationship 11 with a supplier (Ruekert and Churchill 1984). The construct should be operationalized by asking respondents to indicate their beliefs about the supplier's role performance, communications between the two parties and the adequacy and equity of their financial returns. Lusch (1977) and Lusch and Brown (1982) empirically evaluated the efficacy of adding an evaluative component to the model and found that it did not improve the predictive power of the model. That is, the authors added an importance scale and asked respondents to evaluate the impor- tance of receiving each service. The findings showed that the importance scale was not needed. It added unnecessary complexity to the model. Each business function performed by the supplier should be identi- fied and an associated list of items developed to measure it. Finally, the measure must be empirically derived. Channel member performance referred to the intermediary's perfor- mance and was investigated from the supplier's perspective. Performance was defined as the degree to which the channel intermediary engages in behavior that contributes to the fulfillment of the supplier's objec- tives (Gaski and Nevin 1985). The performance variable should be operationalized by obtaining financial and task related evaluations of the intermediary's performance. This may be impossible for two reasons. First, the supplier may feel that the data are confidential. Second, the supplier may not conduct formal appraisals of intermediary performance. Faced with these situations, the researcher may need to select another research setting. Use of a perceptual measure of performance would be ill-advised because it would be difficult to construct a valid and reliable measure. 12 Channel member dependence should be depicted as an economic variable. For purposes of this study, the intermediary's dependence was investigated. It was defined as the extent to which a channel member relies on a channel relationship to achieve or maintain a desired level of economic status (Etgar 1976b). Objective measures that reflect the intermediary's financial dependence should be employed. Some examples include percent of sales, percent of income, and percent of net worth. Several researchers have implied that channel member satisfaction leads to performance. A firm conclusion about the directionality of the link can not be supported but the job satisfaction literature has produced findings that suggest that performance leads to satisfaction (Wanous 1974). Given that a channel member depends upon the relationship for income and often makes a sizable investment in the business, it seems most reasonable that performance leads to satisfaction. Also, it is logical to hypothesize that a channel member's satisfaction would be influenced by the extent to which the member assigns responsibility for his performance to his partner. For example, an intermediary that assigns little credit for its performance to its supplier would be less satisfied than an intermediary that credits the supplier highly for the intermediary's performance. A positive relationship is also predicted between channel member dependence and performance. The rationale for this is that most indivi- duals with a high level of financial dependence would be motivated to perform. That is, an intermediary with a large financial stake in the relationship should perform at a high level. Frazier (I983a,b) suggests that satisfaction leads to dependence. But the relationship between channel member satisfaction and dependence 13 is not so straightforward. It is possible for a highly dissatisfied intermediary to be very dependent due to the lack of an attractive alternative or the intermediary's inability to take quick action without violating legal strictures governing the partnership. However, in the long term, given attractive alternatives, the dissatisfied intermediary's dependence would decline. Logically, in the short run, the intermediary would probably not increase its dependence by making additional invest- ments in the business. In contrast, a highly satisfied intermediary may not increase its level of dependence but for other reasons. First, it may have other more attractive options than expanding its investment in this business. Another explanation is that the intermediary wishes to spread its risk by not becoming overly dependent upon one supplier or one source of income. In summary, it is reasonable to postulate that a high level of satis— faction may not directly influence dependence as suggested in the literature. The author constructed and empirically tested two theoretical models. Both were depicted and evaluated as causal models. The first model, termed the Prevailing View, represented the relationships between the variables of interest as suggested by the literature. The second model, called the Alternative Model, was based upon rival explanations of the relationships between the variables that were proposed by the author. The Prevailing View is shown in Figure 3-1 in Chapter 3 and the specific hypotheses associated with the model were: H1: A channel member's satisfaction will directly influence the channel member's performance, dependence and reinvestment. 2: A channel member‘s dependence will directly influence the channel member's performance. 14 H : A channel member's reinvestment in the business will directly influence the channel member's performance. The Alternative Model appears in Figure 3-2 in Chapter 3. The hypotheses associated with the Alternative Model were: H1: A channel member's dependence directly influences the member's reinvestment in the business. H2: A channel member's dependence directly influences the member's performance. H3: A channel member's reinvestment in the business directly influences the member's performance. H4: A channel member's performance directly influences the member's attributions of responsibility. H5: A channel member's attributions of responsibility directly influence the member's satisfaction. METHODOLOGY The literature review revealed problems with the reliability and validity of measures used in previous studies. Of particular note were content, convergent and discriminant validity. The methodology employed in this research corrected the deficiencies associated with prior empirical work in this area. Content validity was ensured by using a two step process to ' empirically derive the measures. First, the researcher held discussions with industry experts and key managers in the sponsoring firm. This enabled the researcher to confirm and/or revise several measures and was particularly helpful in identifying the business functions performed by the franchisor. Discussions with management provided the starting point for the development of the role performance dimensions of the channel member satisfaction measure. 15 The second step in this process was to conduct in-depth interviews with a representative sample of the franchisee population. These inter- views lasted from two to five hours and produced numerous additions, deletions and revisions to the measures. The questionnaire was contin- ually revised as a result of the interview findings. In all, the researcher produced five iterations of the questionnaire. Prior to the mail survey, the researcher telephoned all franchisees to inform them of the study, to identify qualified respondents and to gain their agreement to participate. The telephone calls confirmed what the mailing list had already indicated. Namely, that most of the fran- chises were owned and managed by a single individual. When a franchise had multiple owners, the telephone contact showed that usually one owner actively managed the business, and the owners selected this individual to participate in the survey. The researcher accepted this decision as being logical, and used the key informant data collection method. A comprehensive eight-page typeset questionnaire was sent to the 204 qualified respondents. A limited information estimation procedure was used to evaluate the measurement and structural models. First, the measurement model was evaluated to assess the reliability and validity of the measures and then the structural model was assessed. The use of a two-step analysis proce- dure was particularly important because channels theory is weak and the measures of key constructs are still undergoing develOpment. A limited information estimation procedure enables the researcher to more readily separate measurement problems from theoretical issues. In contrast, a full information estimation procedure like LISREL estimates the measurement and structural models simultaneously which makes it difficult for the researcher to respecify the model (Anderson and Gerbing 1982). 16 MANAGERIAL RELEVANCE Prior research about the behavioral dimensions of marketing channels suffered from many significant weaknesses. Perhaps the most significant shortcoming was the lack of consideration given to key economic variables. This research integrated the behavioral and economic charac- teristics of channels. It provided the first empirical investigation of channel member satisfaction, often cited as an important behavioral variable, and channel member performance, the variable of utmost impor- tance to managers. In addition, the research examined channel member dependence and reinvestment in relationship to satisfaction and performance. Managers need strategies to help them successfully influence their intermediaries' performance. Intuitively, it seems that highly satisfied intermediaries may be high performers. But pursuit of this strategy requires a very deliberate effort to optimize the allocation of the supplier's resources and it may be costly so the benefits must be care- fully weighed. The research provided managerially relevant findings useful in devising strategies to influence intermediaries' performance. ORGANIZATION OF THE DISSERTATION The balance of the dissertation is presented in Chapters 2 through 5. Chapter 2 reviews the relevant literature that served as the theore- tical foundation for this research, and provided guidance regarding methodology.' The chapter also explores the general weaknesses associated with channels research. Next, the three constructs of interest are presented. Each construct is discussed in relation to: its conceptual development; the construct's definition; and operationalizations of the construct used in previous studies. The following section examines 17 current theory and related empirical evidence relevant to the constructs of interest. Then, two noteworthy methodological issues are outlined. The first is the use of a single channel system as the research setting. The second issue involves the use of the key informant data collection method. The chapter ends with a summary of the literature and major conclusions. Chapter 3 presents the research design utilized for this disser- tation. First, the research objectives are presented followed by the development of a conceptual framework to guide the empirical work. Research hypotheses are presented that relate to the Prevailing View and the Alternative Model, respectively. The methodology is reviewed including a description of the research setting; procedures used to develop the measures; data collection methods; and data analysis procedures. Chapter 4 reports the findings of the research. Results from testing the hypotheses are analyzed and reported. Appropriate statistics developed from the study are presented. Finally, Chapter 5 summarizes the research, presents conclusions drawn from the study and explores both the research and managerial impli- cations. The chapter ends with recommendations for future research. CHAPTER 2 REVIEW or THE LITERATURE The purpose of Chapter 2 is to examine the literature relevant to channel member satisfaction, performance and dependence. The review of the literature should elucidate current thinking about these three constructs and suggest issues worthy of further investigation. Channel theory and research have been fragmented into two orientations: an economic focus and a behavioral focus. The economic approach has been concerned with channel systems, especially the structure of channels and the allocation of business functions among channel participants (Stern and Reve 1979). The behavioral approach borrows heavily from social psychology and organization theory and has concentrated on channel member power and vertical conflict (Stern and Reve 1979). Since 1960, the behavioral approach has dominated channel research. Most of this research dealt with channel member power and conflict and less often focused on channel member satisfaction, performance and dependence (Gaski 1984). Despite numerous empirical studies, the behavioral research stream has not yielded many important and managerially useful insights into channel management (Frazier 1983b; Frazier and Summers 1984; Gaski 1984; Stern and Reve 1980). Indeed, this research stream has several major shortcomings including: an incomplete conceptual framework; failure to adequately define the constructs; poor operationalizations; insufficient 18 19 evidence of the reliability and validity of the measures; and inapprOp- riate analysis procedures (Gaski 1984). Given these weaknesses several marketing scholars have expressed grave concerns about what can legitimately be concluded about channel management on the basis of these studies (Frazier 1983b; Gaski 1984; Ruekert and Churchill 1984; Sheth and Gardner 1982; Reve and Stern 1979). An anonymous Journal of Marketing reviewer mused that the shortcomings of this research stream are such that one wonders if we really know anything about the behavioral dimensions of channel management (Gaski 1984). Clearly, there exists a need for a more solid theoretical foundation for, and increased rigor in conducting empirical research regarding channel management. Improvements of this type must be based at least partially on a review of prior conceptual and empirical work, if for no other reason than to avoid past mistakes. Thus, the remainder of Chapter 2 is organized into four sections as follows: (1) construct definitions and operationalizations; (2) current theory and related evidence; (3) methodological issues; and, (4) a summary of the literature and major conclusions. Section one reviews the definitions and Operationalizations of the constructs of interest in this research. Section two presents the current state of theory regarding channel member satisfaction, perfor- mance, dependence and related constructs. Section two also includes a discussion of the empirical findings of major research studies. Section three outlines two important methodological issues: (1) the use of a single channel system as the research setting for empirical work, and (2) the use of the key informant method of data collection. Finally, section four summarizes the literature and presents the major conclusions of interest to future researchers. 20 CONSTRUCT DEFINITION AND-OPERATIONALIZATION This section presents the definitions and Operationalizations of channel member satisfaction, performance and dependence used by researchers previously. It is divided into three parts each of which addresses one of the three constructs of interest. Channel Member Satisfaction Early studies of channel member satisfaction did not explicitly define the construct but implied that it reflected satisfaction with the supplier's role performance or the middleman's willingness to commit to the relationship again, given all that had transpired between the two partners (Brown and Frazier 1978; Hunt and Nevin 1974; Rosenberg and Stern 1971). Lusch (1977) criticized these presentations of channel member satisfaction suggesting instead that satisfaction is a multi- faceted construct. According to Lusch, a channel member could be more or less satisfied with various aspects of the relationship, which he Opera- tionalized as elements of manufacturer role performance.. Subsequently, channel member satisfaction has increasingly been conceptualized as a multi-dimensional construct (Dwyer 1980; Frazier 1983b; Michie and Roering 1978; Ruekert and Churchill 1984). Ruekert and Churchill, citing the similarities between a channel partnership and the relationship between a salesperson and employer, proposed a definition of channel member satisfaction based on the definition of salesperson satisfaction: Channel member satisfaction comprises the domain of all characteristics of the relationship between a channel member and another institution in the channel which the member finds rewarding, profitable, instrumental and satisfying...or unsatisfying. 21 The manufacturer's or franchisor's role performance has been con- sistently identified as making up several dimensions of the satisfaction construct (Frazier 1983b; Lusch 1977; Michie and Roering 1978; Ruekert and Churchill 1984). In addition, the intermediary's financial returns and communication between the two partners have been cited as dimensions of channel member satisfaction (Dwyer 1980; Frazier 1983b; Ruekert and Churchill 1984). Two aspects of financial returns have been suggested as being relevant to satisfaction, the equity of the returns and the adequacy of achieved returns relative to initial expectations (Dwyer 1980; Frazier 1983b). Thus, the prevailing view of channel member satisfaction is that it reflects the intermediary's satisfaction with multiple dimensions of the relationship, primarily the supplier's role performance, the intermediary's financial returns and communication between the two parties. However, because development of this conceptual scheme paralleled empirical investigations of channel member satisfac- tion, operationalization of the construct often did not reflect the present conceptual view. The measurement of channel member satisfaction has undergone significant change over the years mirroring the continuing conceptual and empirical develOpment of the construct. Early studies employed psycho- metrically poor single-item measures (Brown and Frazier 1978; Hunt and Nevin 1974; Rosenberg and Stern 1971; Walker 1972). Rosenberg and Stern, Brown and Frazier, and Walker measured channel member satisfaction with single-item scales anchored with the phrases "very satisfied" and “very dissatisfied." Hunt and Nevin measured overall satisfaction by asking respondents what they would do "if they had it to do over again." Walker conducted a laboratory experiment with 76 university students while Brown 22 and Frazier conducted an exploratory study with 26 automobile dealers. Hunt and Nevin and Rosenberg and Stern conducted mail surveys of 815 fast food franchisors and 110 dealers and distributors of a household durable good, reSpectively. None of these authors reported any evidence of the reliability or validity of their measures. In a laboratory study with 80 university students, Dwyer (1980) employed a multi-item global measure of satisfaction that captured the intermediary's satisfaction with rewards. He reported an alpha coeffi- cient of .94 for the satisfaction scale but did not report any evidence of the measure's validity. The external validity of the measure is highly questionable because Dwyer's subjects were university students not channel participants. In a mail survey of 567 automobile dealers, Lusch (1977) made an important contribution to the literature when he recognized that an intermediary's satisfaction is based upon a set of issues over which they may be more or less satisfied with the manufacturer or franchisor. He propoSed a 16-item measure of satisfaction comprised of various services automobile.manufacturers provided dealers. The 16 items were identified from exploratory interviews with auto dealers and other industry experts. Included among the 16 items were such services as local advertising assistance, salesperson incentive programs, mechanic training, stock rebates and service manuals. Respondents were asked to use a four-point scale to indicate how dissatisfied (1) or satisfied (4) they were with each of these services. Lusch employed a variety of means to assess the measure's reliabi- lity and validity. The content validity of the measure was attributed to the exploratory interviews and convergent validity was demonstrated 23 (.653) by testing the correlation between the multi-item measure and a single-item measure of satisfaction. Coefficient alpha was employed to assess the internal consistency of the multi-item measure. The measure evidenced sufficient internal consistency as indicated by a coefficient alpha of .87. It is noteworthy that Lusch was the first researcher to employ a multi-item measure and to assess its reliability and validity. Consequently, Lusch should be credited with significantly improving the conceptualization of channel member satisfaction and its operationali- zation. Michie and Roering (1978) and Ruekert and Churchill (1984) also utilized multi-item measures of channel member satisfaction. In a mail survey of 161 automobile dealers, Michie and Roering measured “level of gratitude“ for 15 warranty-related items using a five-point scale. Michie and Roering followed a procedure similar to Lusch's to develop the measure. The authors also reported weak evidence of convergent validity (.515) with a single-item measure of satisfaction and assessed the measure's internal consistency using Cronbach's alpha (.93). It is important to note that the satisfaction measures devised by Lusch and Michie and Roering were measures of supplier role performance. The measures did not represent satisfaction with financial returns or communication between the two parties, two aSpects of the relationship that have since been recognized in the literature as possible dimensions of channel member satisfaction (Dwyer 1980; Frazier 1983b; Ruekert and Churchill 1984). Ruekert and Churchill investigated the satisfaction of retailers and wholesalers of consumer batteries with the manufacturer and, developed two multi-dimensional multi-item measures of channel member satisfaction 24 that included dimensions of role performance, financial returns and communication. The first measure was labeled SATDIR because it required respondents to directly evaluate the relationship using a five-point scale ranging from "very satisfied" to "very dissatisfied." The second measure was named SATIND because it used a five-point scale ranging from “strongly agree" to “strongly disagree“ to indirectly assess the respondent's satisfaction. Ruekert and Churchill identified five dimensions of channel member satisfaction that applied to both measures but the number of items per dimension differed as shown in Table 2-1. Both the items and measures evidenced acceptable internal consistency as indicated by their alpha coefficients which also appear in Table 2-1. Ruekert and Churchill employed the total scores for each measure in a series of pairwise correlations to evaluate convergent and discriminant validity. The two Table 2-1 Measures of Channel Member Satisfaction Number of Items Alpha . Dimension SATDIR SATIND SATDIR* SATIND** Social Interaction 4 5 .70 .87 Product 1 6 -- .76 Financial 3 5 .68 .67 Advertising/ 4 Promotion 5 2 .79 .56 Other Assistances 3 3 .75 .73 .90 .89 *Overall alpha **0verall alpha Source: Ruekert, R. and G. Churchill (1984), "Reliability and Validity of Alternative Measures of Channel Member Satisfaction,“ Journal of Marketing Research, 21 (May), 226-233. 25 multi-dimensional measures had a correlation coefficient of .63. Both SATDIR and SATIND correlated with a single-item measure of satisfaction at .68 and .58, respectively. These findings suggested that the three measures demonstrate convergent validity. Discriminant validity was supported by demonstrating that the correlations between the satisfaction measures were greater than correlations with other constructs, a criterion suggested by Campbell and Fiske (1959). Although Ruekert and Churchill's work marks a major step forward in the conceptual and empirical development of channel member satisfaction, their measures have some significant weaknesses. First, the financial dimension did not address equity of returns or the adequacy of returns both of which have been cited in the literature as key assessments made by intermediaries (Dwyer 1980; Frazier 1983b). Second, a comparison of the items that comprised the Social Interaction dimension of the SATDIR and SATIND measures revealed significant disparity in content. With respect to the social interaction dimension of the SATDIR measure, three of the four items reflected physical distribution activities. In contrast, a review of the social interaction dimension of the SATIND measure revealed that all five items reflected communication with the manufacturer's rep. This leads one to question just what this dimension really represented--communication or physical distribution. This kind of disparity should not exist between two measures of the same construct and suggests the need for further conceptual develOpment of the social interaction dimension. Finally, the items associated with the role performance dimensions vary between the two satisfaction measures, suggesting the need for a more rigorous methodology for developing the dimensions and items that comprise the role performance measure. 26 However, despite the potential for improvement in their measures, Ruekert and Churchill advanced the conceptualization and measurement of channel member satisfaction. One particularly noteworthy contribution is Ruekert and Churchill's attention to the discriminant validity of their measures. Previous research on the behavioral dimensions of channels showed no evidence of the discriminant validity of the measures employed. This omission led some scholars to argue that some measures of key constructs, like power sources, did not really measure what they were supposed to measure (Frazier and Summers 1984; Gaski 1984). Aside from Ruekert and Churchill's study, there is no evidence that the behavioral concepts- commonly investigated by channel researchers are really separate constructs. Some of the measures used to represent these constructs may be measuring different dimensions of the same construct. Given these shortcomings, empirical studies of the behavioral dimensions of channels should be critically reviewed and cautiously interpreted (Gaski 1984). Channel Member Performance Lambert (1978) observed that channel member performance has been all but ignored by channel researchers. Given this, it is not surprising that no consensus definition of channel member performance exists. However, empirical studies have demonstrated that several perSpectives can be taken regarding channel member performance (Gaski and Nevin 1985; Kelly and Peters 1978; Lusch 1976b). For example, the intermediary's performance can be addressed from the customer's viewpoint or the manufacturer's. A researcher also could assess the intermediary's own perception of its performance. Similarly, the supplier's performance 27 could be investigated from its own perspective, the intermediary's viewpoint or the customer's. In a study of automobile dealer performance, Lusch adopted the dealer's perspective of its performance but he did not explicitly define performance. Kelly and Peters conducted a comparative study of channel member performance relative to conflict levels and these authors did not define channel member performance either. Similar to Lusch, Kelly and Peters, took the intermediary's perspective of its own performance. Lusch and Kelly and Peters employed financial measures of perfor- mance. Lusch used two measures of dealer financial returns, return on assets and asset turnover. Kelly and Peters asked intermediaries from a variety of industries to rank their firm's profitability in comparison to similar firms in the same industry. However, the authors did not define profitability. In a study of the distribution channel for an industrial capital good, Gaski and Nevin investigated channel member performance from the supplier's perspective. These authors defined performance as: “...the degree to which a supplier's relationship with a dealer con- tributes to fulfillment of the supplier's objectives.“ This implies that the precise definition of performance depends upon the supplier's objec- tives. It may be defined in financial terms but also could incorporate task related measures related to supplier objectives. Gaski and Nevin employed a two-item perceptual measure of intermediary performance. These authors stated that the scale used to measure performance was “an attempt to combine a more objective rating with a subjective impression of comparative performance." But Gaski and Nevin did not explain why such a measure would be managerially useful. 28 Lambert (1978) performed an exploratory study of channel management practices in 18 manufacturing firms and found that the types of perfor- mance measures used varied across firms and was somewhat related to the type of intermediary. For example, Lambert found that the most frequently used measures of retailer performance were sales, market potential/penetration and performance to plan. He also reported that most of the firms did not measure the performance of the channel, and most significantly, no firm used the contribution approach to profit- ability analysis to measure channel and/or channel member performance. Clearly, the definition of channel member performance is not well developed in the literature. Logically, its specific definition would be determined by the perspective taken by the researcher, the research setting and the business problem under investigation. The operationali- zation would depend upon these factors as well. Channel Member Dependence Marketing channels are characterized by the interdependency of the members. Typically dependence is disproportionate with the supplier being less dependent than the intermediary (Bowersox, Cooper, Lambert and Taylor 1980). For example, a dealer for IBM may derive 70 percent of its income from IBM products. At the same time, this dealer may account for just one percent or less of IBM's total income. Both parties are finan- cially dependent upon the other but not in a balanced way. Channel members are both functionally and financially interdepen- dent. Functional dependence is pre-arranged and the functional roles are uniform in the sense that all intermediaries of a given type perform an identical set of activities. Conversely, the supplier performs the complement of these activities for all intermediaries of this type. 29 However, financial dependence may vary considerably. That is, one intermediary may be wholly dependent upon the supplier for income while another derives only a small percentage of its total income from this supplier. Likewise, the supplier's financial dependence may vary from one intermediary to another. Channel researchers and practitioners have an interest in understanding how a member's financial dependence relates to other constructs such as power, the member's satisfaction with the relationship and its performance as perceived by the supplier. Channel member dependence was originally conceptualized as the inverse of the power of one's partner but empirical evidence does not support this view (El-Ansary and Stern 1972; Etgar 1976b). Frazier (1983a) suggested that channel member power is relative and relates to both parties' dependence. For example, both the supplier's and the intermediary's dependence is a determinant of the supplier's power and vice-versa. To illustrate the point, assume that the intermediary relies on the supplier for 70 percent of its income. Given this, one might conclude that the supplier has a great deal of power over the inter- mediary. However, such a conclusion is premature, because in order to determine the supplier's power it is necessary to know how much the supplier depends upon the intermediary. If, for example, the supplier derives 70 percent of its income from the intermediary then both parties are highly dependent but neither has more power than the other. Thus, in a two-firm relationship power is determined by the relative dependency of the participants. Emerson (1962) proposed that: "... the dependence of Actor 8 upon Actor A is (1) directly proportional to 8'5 motivational investment in goals mediated by A, and (2) inversely proportional to the availability 30 of those goals to B outside of the A-B relationship." Emerson's definition provided the basis for the measurement of channel member dependence in the two empirical studies of dependence reported in the literature (Etgar 1976b; Phillips 1981). Within the context of a marketing channel, dependence refers to the extent to which one party relies on its partner to achieve or sustain its financial goals (Etgar 1976b). It also suggests that the availability of other sources of income influences a member's decision to maintain and possibly increase its dependence. In a mail survey of 113 independent insurance agents, Etgar operationalized the agent's dependence in financial terms. He also measured the difficulty of replacing the income derived from leading insurers. Specifically, Etgar's four measures of dependence were: 1. Total number of insurers the agent represents; 2. The percentage of agent's premiums concentrated with his leading insurer; 3. Difficulty experienced in replacing a leading insurer (l-very difficult; S-very easy); and, 4. Reliance on casualty income (percent of revenues). Phillips (1981) assessed the reliability and validity of measures of organizational characteristics used by previous channel researchers. He operationalized channel member dependence with a measure of the substi- tutability of major suppliers develOped by Etgar. Phillips used the multiple informant method of data collection and reported that Etgar's perceptual measure did not appear to be valid because decomposition of the variance indicated that most of it was attributable to methods and random error variance, not trait. 31 CURRENT THEORY AND RELATED EMPIRICAL EVIDENCE As noted previously, most empirical research involving the behav- ioral dimensions of marketing channels focused on power and conflict. However, some empirical studies have included channel member satisfation, performance and dependence. Current theory regarding the interrelation- ships among these variables will be discussed in this section. Major empirical research studies will be reviewed. Figure 2-1 provides a summary of key relationships. Most of these relationships have been supported by statistically significant research findings. However, Figure 2-1 also includes several correlations suggested in the literature that have not been investigated by channel researchers. Independent Variable Dependent Variable Correlation Noncoercive Power Satisfaction Positive Noncoercive Power Conflict Negative Coercive Power Satisfaction Negative Coercive Power Conflict Positive Conflict Satisfaction Negative Conflict Performance Negative Satisfaction Dependence Positive* Satisfaction Performance Positive* Dependence Performance Positive* *No direct empirical support exists for this relationship. This is a hypothesized relationship based upon the relevant theoretical and empirical literature. Figure 2-1. The Interrelationships Among Key Variables in Channel Management Power and Conflict Empirical studies of channel member satisfaction evolved from interest in channel member power and conflict so it is important to 32 understand why power and conflict were viewed as key variables in channel management. General consensus exists regarding the definition of power as the ability to cause someone to do something he would not have done otherwise (Gaski 1984). In the context of marketing channels, power has consistently been defined as one firm's ability to influence the percep- tion, decisions and/or behavior of its exchange partner (El-Ansary and Stern 1972; El-Ansary 1975; Wilkinson 1974; Hunt and Nevin 1974). Stated another way, power refers to a firm's ability to influence its exchange partner's performance to its own advantage. Power has been conceptualized from the perspective of the manu- facturer or franchisor, presumably the more dominant firm in a two-firm relationship (El-Ansary and Stern 1972). Power has been identified as a key construct in channel management because manufacturers and franchisors generally want to achieve as much control over their intermediaries as possible. When a firm chooses to market through intermediaries the firm Spins ,off some business functions to these partners because management believes that the intermediaries can perform certain functions more efficiently (Mallen 1973). Joining together enables both parties to achieve increased profits but the supplier now has less ability to fully control all aspects of its business, especially those functions now performed by its intermediaries. Thus, manufacturers and franchisorshave a keen interest in strategies for influencing their intermediaries' performance (Frazier and Summers 1984). Conflict has been recognized as an important related construct although there are divergent Opinions about this interrelationship. Some theorists posit that a firm's attempt to control its exchange partners 33 leads to conflict (Pondy 1967). Others-share this view if the type of power used is coercive (Raven and Kruglanski 1970). Most marketing scholars accept the idea that at least the potential for conflict is present in all channels due to the functional interdependency of the relationship (Alderson 1957; Assael 1969; Cadotte and Stern 1979; Lusch 1976a; Mallen 1963; Pondy 1967; Reve and Stern 1979; Stern and El-Ansary (1977); Ross and Lusch 1983; Wilkinson 1981). According to the theory, a channel member willingly yields power to its partner when the partner uses noncoercive power. Conversely, when coercive power is employed a channel member begrudgingly yields power (El-Ansary and Stern 1972; Hunt and Nevin 1974; Lusch 1976a). The rationale for studying power sources was to determine if the use of coercive or noncoercive power produced differential effects on channel member satisfaction. Specifically, it was hypothesized that noncoercive power sources would increase channel member satisfaction and that coer- cive power sources would decrease satisfaction (Brown and Frazier 1978; Dwyer 1980; Hunt and Nevin 1974; Lusch 1977; Michie and Roering 1978; Ross and Lusch 1983; Wilkinson 1981). Empirical tests have consistently supported both hypotheses. Researchers have shown evidence of a positive relationship between noncoercive power sources and satisfaction. Conversely, an inverse relationship was found between coercive power sources, and satisfaction (Brown and Frazier 1978; Dwyer 1980; Hunt and Nevin 1974; Lusch 1977; Michie and Roering 1978; Ross and Lusch 1983; Wilkinson 1981). While these findings seem intuitive, the relationship between noncoercive power sources and satisfaction requires a more comprehensive review because both constructs have been operationalized through role 34 performance measures. Since the operationalization of channel member satisfaction was previously reviewed, the present discussion will focus primarily on the noncoercive power sources construct. Noncoercive power sources was commonly operationalized via a measure of supplier role performance (Hunt and Nevin 1974; Lusch 1976a, 1976b, 1977; Michie and Roering 1978; Ross and Lusch 1983; Wilkinson 1981). Two studies of power also used role performance measures. Frazier (1983a) used a role performance measure of power in study of the distribution channel for automobiles; and Guiltinan, Rejab and Rodgers (1981) did the same in a study of power in a fast food channel. A list of services that suppliers provided their intermediaries was developed on the basis of interviews with channel participants (Frazier 1983a; Guiltinan, Rejab and Rodgers 1981; Hunt and Nevin 1974; Lusch 1977; Michie and Roering 1978; Ross and Lusch 1983; Wilkinson 1981). The multi-item measures of noncoercive power sources ranged from six items (Frazier 1983a) to 24 (Ross and Lusch 1983). Michie and Roering employed 17 items and Lusch and Hunt and Nevin reported 16 and 14 items, respectively. Guiltinan et al. and Wilkinson employed eight items. In each of these studies the respondents were asked to rate the supplier's performance of each service using a Likert scale. However, Lusch (1977) and Lusch and Brown (1982) believed that the importance of each service must also be considered. Citing instrumentality theory, Lusch and Lusch and Brown argued that the strength of a supplier's power sources is a function of both perfor- mance and instrumentality of the services in the intermediary's success. Therefore, the authors employed two scales, a performance scale and an instrumentality (importance) scale. Specifically, in addition to rating 35 the supplier's performance of each service, respondents were asked to indicate how important each service was to the success of their business using a four-point Likert scale. In both cases, the authors tested the hypothesis that a multiplica- tive combination of a respondent's evaluation of the importance of the service, and their belief regarding the supplier's performance of the service will improve the measure's predictive power (Lusch 1977; Lusch and Brown 1982). In both studies, two hypotheses were tested, one in relationship to the measure of noncoercive power and another in rela- tionship to the measure of coercive power. Lusch obtained mixed results. The hypothesis was rejected for noncoercive power sources but weakly supported in the case of coercive power sources. He concluded that because the results were not strong, they should not be strictly interpreted. However, Lusch and Brown reported a different result but without explanation. Based on their reanalysis of the data, both hypotheses were rejected (Lusch and Brown 1982). That is, the importance scale did not improve the predictive power of either measure. Ex post facto, the authors normalized the evaluation component of the measures, and recomputed the correlation coefficients but found no improvement in predictive power (Lusch and Brown 1982). They suggested that one possible explanation for these findings is that a "halo“ effect existed. That is, the belief and evaluation components were not independent (Lusch and Brown 1982). These findings are important, but the major point remains that both noncoercive power sources and channel member satisfaction have been operationalized with role performance measures. According to Bagozzi 36 (1980), measures should be selected based on the definition of the construct the measures are intended to represent. Although it is possible and even desirable to use multiple measures of a construct, the same measure cannot be used to represent two distinct constructs (Churchill 1979). Significantly, the conceptualization of the noncoercive power sources construct went unchanged for ten years following its introduction by Hunt and Nevin (1974). Having accepted the French and Raven classifi- cation of power sources, channel researchers admitted the difficulty of operationalizing all but coercive power sources (Hunt and Nevin 1974). So, referent, legitimate, expert and reward power were combined to form the noncoercive power sources construct. Hunt and Nevin established the practice of using a role performance measure to operationalize noncoer- cive power sources. The authors justified this operationalization based on their belief that good performance, as perceived by intermediaries, helps establish the supplier as an expert; legitimize the supplier's efforts to gain power; and this leads intermediaries to willingly yield power. Stated another way, Hunt and Nevin suggested that satisfaction with the supplier's role performance would increase the supplier's ability to influence intermediaries in a way that benefits the supplier. Thus, it seems that Hunt and Nevin were measuring the role performance dimension of channel member satisfaction. Frazier and Summers (1984) questioned the legitimacy of the noncoercive power sources construct and criticized the continued unchanging application of the French and Raven power framework to channel management research. Frazier and Summers believed that the power and power sources constructs were ill defined and subsequently not very 37 useful to practitioners. They argued that it is not power per se that suppliers desire, but an ability to influence their intermediaries' performance. This suggests that channel researchers should investigate influence strategies including the outcomes (Frazier and Summers 1984). In summary, Ruekert and Churchill (1984) devised two reliable and valid measures that reflected the general consensus that satisfaction is a multi-dimensional construct, including satisfaction with supplier role performance. Given the current conceptual view, their study provideS‘ support for the conclusion that the previously reported correlations between noncoercive power sources and channel member satisfaction were actually correlations between two measures of channel member satis- faction. Satisfaction and Conflict Three of the four studies that investigated the relationship between channel member satisfaction and conflict supported a negative relation- ship between these two constructs (Dwyer 1980; Rosenberg and Stern 1971; Wilkinson 1981). Rosenberg and Stern conducted a mail survey of partici- pants at three levels of the channel for a consumer durable good. The authors interviewed 110 manufacturers, distributors and dealers regarding their relationships with the other two. Thus, Rosenberg and Stern collected data on three dyads, manufacturer-distributor, manufacturer- dealer and distributor-dealer. They hypothesized that the causes of conflict are differences in goals, domains and perceptions. So, respondents were asked to use a five-point scale to “agree" or "disagree" with 32 statements related to goals, domains and perceptions of their partners. Differences in the mean responses between two exchange partners indicated conflict. The authors devised a single-item measure 38 to investigate channel member satisfaction. Respondents were asked to evaluate the performance of their exchange partners as a measure of overall satisfaction. No evidence of the reliability or validity of the measures was reported. Rosenberg and Stern found a significant level of conflict in each of the three dyads. In each case the conflict stemmed from differences in two of the hypothesized causes. Thus, they concluded that goal incon- gruity, domain dissensus and perceptual differences do act as basic causes of intrachannel conflict. Significant levels of dissatisfaction were also reported for distributors with their dealers and for dealers with the manufacturer. This led Rosenberg and Stern to conclude that dissatisfaction with one's exchange partner leads to conflict although the authors admitted that the directionality of the link may be just the opposite. That is, a high level of conflict may cause one firm to devalue their partner's performance as a way of blaming them for the conflict. _ Dwyer (1980) investigated cooperation and satisfaction in a laboratory study with 80 university students. He Operationalized both constructs with multi-item semantic differential scales and reported a high level of reliability for both measures. Dwyer found a strong positive correlation between the cooperativeness of one's exchange partner and a respondent's satisfaction. If cooperation is assumed to be the inverse of conflict, then this finding suggests an inverse relation— ship between satisfaction and conflict. However, it is not at all clear that cooperation can be interpreted this way (Gaski 1984). Furthermore, regardless of how cooperation is viewed, no evidence of the measure's 39 validity was shown which is not surprising given the use of student subjects. Wilkinson (1981) investigated power sources, satisfaction and conflict in the beer industry. He reported a significant positive correlation between noncoercive power sources and satisfaction and a negative correlation between conflict and satisfaction. Regarding the measures employed, Wilkinson reported evidence of reliability but did not address validity. Finally, Assael (1968) conducted an exploratory study of conflict in the auto industry and reported that conflict may be beneficial to the relationship depending upon how the two parties responded to it. Assael also found that satisfaction with one's financial rewards may be an important factor in determining the level of conflict in the channel as evidenced by a correlation between dealer profits and attitude toward the manufacturer. During the mid-1950's, dealer profits were low and most dealers reported a negative attitude toward the manufacturer. However, in the early 1960's, dealer profits were much higher and their attitude toward the manufacturer was positive. Assael's findings support Lambert's (1978) contention that most issues of power and conflict in a channel relationship revolve around the financial well being of the participants. This suggests that satisfaction with financial returns may be of prime importance in determining the overall level of channel member satisfaction. Conflict and Performance Two studies provided limited support for a negative relationship between conflict and performance (Kelly and Peters 1978; Lusch 1976b). Each study will be discussed in turn. In a comparative analysis of 40 vertical conflict in different types of channel systems, Kelly and Peters found a negative correlation between an intermediary's performance, operationalized with a self-reported measure of profitability and conflict. That is, channel members reporting higher levels of conflict often performed at lower levels. The distributors of industrial goods reported the lowest profit- ability and the highest level of conflict. In contrast, franchisees and distributors Of consumer goods ranked themselves significantly above average in profitability and showed low levels of conflict. The authors' concluded that these findings should be viewed as preliminary support for an indirect relationship between channel member profitability and vertical conflict (Kelly and Peters 1978). In a study of the automobile industry, Lusch provided an empirical test of Rosenbloon's (1973) three hypotheses about the relationship between conflict and channel member performance. Specifically, the three hypotheses tested were: . H1: The operating performance of the franchisee will increase (as the franchisee's conflict with the franchisor increases. 2: The operating performance of the franchisee will decrease as the franchisee's conflict with the franchisor increases. H3: The operating performance of the franchisee will increase as the franchisee's conflict with the franchisor increases but only up to a point, after which the franchisee's operating performance will decrease. Lusch used two financial measures of dealer performance which he obtained from the dealers. ‘ The study findings supported Hz, the franchisee's Operating perfor- mance declined as the level of conflict with the franchisor increased. 41 The author concluded that dealers who frequently disagreed with the manufacturer did so at the cost Of reducing their own profitability. However, this explanation assumes that conflict causes decreased pro- fitability. Lusch admitted that he had no evidence to suggest the causal link, so he Offered an alternative explanation. 'He proposed that dealers who are not very profitable may blame the manufacturer for their low performance and that may lead to frequent disagreements (Lusch 1976b). Satisfaction and Performance Channel member satisfaction was first introduced into the literature when McVey (1960) argued the need to recognize middlemen as customers and business partners. He urged manufacturers to sell to their intermediar- ies as well as selling through them. McVey believed that by recognizing and satisfying the middleman's needs manufacturers could improve the entire channel's performance. Mallen (1963) echoed McVey's view that manufacturers should apply the marketing concept to intermediate levels of the channel as a means of improving channel performance. Robicheaux and El-Ansary (1975) developed a model of channel member behavior in which they posited that a member's performance and satisfaction are interdependent. That is, good performance leads to increased satisfac- tion and increased satisfaction motivates a high level of performance. However, due to a derth of empirical evidence, they cited the need for research; but the authors did not define any of the variables in the model or suggest how the variables should be operationalized (Robicheaux and El-Ansary 1975). In a study of food brokers and wholesalers, Lusch (1977) found empirical support for a direct relationship between noncoercive power and satisfaction. Lusch found that the broker's role performance was an 42 important determinant of wholesaler satisfaction. In his conclusions Lusch stressed the importance of this finding to management because he believed that satisfied wholesalers were more likely to cooperate with the broker to achieve the broker's objectives and especially to improve channel performance. ' Other researchers have implied that a low level of satisfaction leads to reduced performance. Dwyer (1980), Rosenberg and Stern (1971) and Wilkinson (1981) all produced empirical support for a negative link between conflict and satisfaction. These researchers implied that reduced conflict leads to increased channel member satisfaction and that satisfaction in turn increases channel member performance. This would suggest that conflict and performance are indirectly related to each other but that both conflict and performance are directly linked to satisfaction. According to this line of thinking conflict may be viewed as a determinant of satisfaction while performance is depicted as a consequence of satisfaction. _ 'But this line of thinking does not account for the fact that channel participants are self interested parties who typically have a large financial stake in the relationship. Marketing channels are business partnerships formed for the mutualeconomic benefit of the participants (Bowersox, Cooper, Taylor and Lambert 1980). Therefore, it seems unlikely that satisfaction determines performance because a dissatisfied _ channel member that performed at a lower level would reduce the member's own rewards. It is more logical to postulate that performance leads to satisfaction. Research in industrial psychology has investigated the relationship between job performance and satisfaction and provided evidence that 43 performance influences satisfaction (Wanous 1974). Sheridan and Slocum (1974) found that performance was a determinant of satisfaction when the individual perceived a direct connection between his effort and subsequent rewards. In a marketing channels context a member could also be expected to evaluate the connection between its partner's effort and the member's performance. A channel member's satisfaction with the relationship would be influenced by the extent to which the member assigned credit or blame for the member's performance to its partner. In summary, the channel literature generally suggests that channel member satisfaction causes channel member performance. But this view does not seem reasonable when one considers that the channel member hurts itself by reducing its performance. Also, the industrial psychology literature has produced evidence that performance causes satisfaction. Therefore, the author suggests that channel member performance is a determinant of channel member satisfaction but that the relationship is influenced by the extent to which the member attributes responsibility for its performance to its partner. Satisfaction and Dependence Although no empirical research has been reported that involved channel member satisfaction and dependence, Frazier (1983b) hypothesized that satisfaction with the supplier's role performanCe and the financial rewards received from the relationship would increase the intermediary's attraction to and trust in the supplier. He suggested that a high level of satisfaction might increase the supplier's expertise as perceived by the intermediary. In addition, Frazier posited that a high level of satisfaction would cause the intermediary to maintain and perhaps increase its level of dependency on the supplier. Frazier based his 44 hypotheses on Blau's (1964) consistency theory. Developed in the context of social psychology, consistency theory predicts that a high level of satisfaction leads an individual to increase his attraction to and dependence on the source. Frazier interpreted this to mean that a high level of satisfaction would result in greater attraction to the supplier, agreement on decision strategy and increased dependency. But Homans (1974) suggested that other factors affect the relationship between satisfaction and dependence. Homans explained that an individual who remained largely dissatisfied with a relationship over an extended period would choose to dissolve the relationship. However, unless a more attractive Option was available the dissatisfied party would maintain his dependence on the source. Risk tendencies also may be important in explaining the relationship between satisfaction and dependence. A risk averse intermediary may not increase its dependence on a supplier even though the intermediary is very satisfied with the supplier. By not increasing its dependency on the supplier the intermediary maintains some measure of control over its economic well being. Indeed, a highly satisfied but risk averse inter- mediary may be expected to intentionally limit its dependence on a supplier. Given these considerations one can conclude that satisfaction does not necessarily lead to dependence but continued dissatisfaction along with an attractive alternative may lead to the dissolution of the relationship. A management issue related to channel member dependence involves the supplier's need to encourage intermediaries to make periodic investments in the business. The reinvestment of profit is needed in order to replace aging or obsolete plant and equipment and to grow the business. 45 As an illustration, McDonalds restaurants have evolved from small hamburger stands with no inside seating to large, air-conditioned restaurants with ample inside seating, drive-thru windows and outdoor playgrounds. Additionally, the menu has expanded from hamburgers, french fries and soft drinks to include an array of breakfast items, chicken and fish entrees and a variety of desserts. The successful evolution of the McDonalds franchise system was at least partially the result of the reinvestment of profits by the fran- chisees. What caused the franchisees to reinvest profits? As with most' owners or managers of businesses, the franchisees most likely reinvested in the business to protect their existing financial stake and to increase future profits. In a channel system like McDonalds where the inter- mediaries are highly (if not wholly) dependent upon the supplier it seems reasonable that the desire to protect their substantial existing invest- ment acts as the primary determinant of the intermediaries' reinvestment of profit. Therefore, if reinvestment is considered a form of increased dependence then one can conclude that an intermediary's existing dependence leads to additional dependence on the relationship. Once again, it is the intermediary's self interest that best explains its behavior. Dependence and Performance A significant number of scholars have advanced the view that a high level of financial dependence leads an intermediary to perform at a high level (Frazier 1983a; Homans 1974; Robicheaux and El-Ansary 1975; Vroom 1964; Walker, Churchill and Ford 1977). The relationship between dependence and performance is easily explained by considering the intermediary's self interest. If an intermediary is highly dependent 46 upon the relationship with the supplier then the intermediary has a vested interest in protecting its economic status. The higher the intermediary's level of dependence the more motivated the intermediary will be to perform. Furthermore, the intermediary evaluates the potential gains to be derived from a high level of performance. If the expected gains justify the effort required to achieve the gains then the intermediary's motiva- tion to perform will be high. Thus, the intermediary assesses both potential losses and expected gains and the combined result determines the level of motivation. Motivation to perform will be highest when the intermediary is highly dependent upon the relationship and expected gains are large. Motivation will be lowest when the intermediary's level of dependence is low and expected gains are small or negative (Frazier 1983a; Homans 1974; Robicheaux and El-Ansary 1975; Vroom 1964; Walker, Churchill and Ford 1977). Summary of Current Theory Empirical research supports the hypothesis that reliance on noncoer- cive power increases intermediary satisfaction and reduces conflict but it is important to recall that noncoercive power was Operationalized with a role performance measure. Therefore, noncoercive power should be viewed here as a dimension of channel member satisfaction for reasons that were discussed earlier. Given this perspective, the research provides evidence of an inverse relationship between satisfaction and conflict. Conversely, it appears that coercion only serves to reduce satisfaction and escalate conflict. Finally, there is limited evidence of a negative relationship between conflict and performance. All of these relationships are depicted in Figure 2-2 but it is important to 47 Noncoercive 47> Conflict Power Sources (2*.3*.6*. lO,I3*,16*. (2*.8*,l9*) + (ana*. 19g 19*) - (3*,15*,19*) NV Satisfaction - - - 1 ----- 9Performance . (9.10.11.12.14) a | / I I I + ' (2*.6*.10~) ' ’ +I (1.4.5) I , ’ (4,5,14,17,18) I l ' I v ,r’ Coercive Dependence Power Sources 1. Blau 1964 KEV; ——————-—-direct empirical 2. Brown and Frazier 1978 evidence exists 3. Dwyer 1980 to support this 4. Frazier 1983b relationship 5. Homans 1974 6. Hunt and Nevin 1974 -------- the relationship 7. Kelly and Peters 1978 has been suggested 8. Lusch 1976a but not empirically 9. Lusch 1976b tested 10. Lusch 1977 11. Mallen 1963 * provides empirical 12. McVey 1960 support 13. Michie and Roering 1978 14. Robicheaux and El-Ansary 1975 15. Rosenberg and Stern 1971 16. Ross and Lusch 1983 17. Vroom 1964 18. Walker, Churchill and Ford 1977 19. Wilkinson 1981 Figure 2-2. The Interrelationships Among Key Variables in Channel Management 48 note that the direction of the causal links has not been firmly established. Three additional interrelationships have been suggested in the literature but not yet empirically validated. These include links between channel member satisfaction, performance and dependence respectively. These variables are also depicted in Figure 2-2. They are connected by dotted lines to signify the lack of supporting evidence. Finally, the author has presented some alternative explanations regarding the relationships between satisfaction, dependence and performance which formed the basis for the research hypotheses tested in this dissertation. Chapter three includes a discussion of the specific research hypotheses. METHODOLOGICAL ISSUES Two methodological conventions of channel research deserve special attention and will be discussed in this section. The two practices, targeting a single channel system for empirical studies and, use of a single key informant will be discussed in turn. Single Channel System It is customary in channel research to investigate the characteris- tics of a single channel system (Brown and Frazier 1978; Frazier and Summers 1984; Hunt and Nevin 1974; Gaski 1986; Gaski and Nevin 1985; Lusch 1976a,b, 1977; Lusch and Brown 1982; Michie and Roering 1978; Rosenberg and Stern 1971; Ross and Lusch 1983; Ruekert and Churchill 1984; Wilkinson 1974, 1981). Most studies target a single channel system due to the cost and effort required to collect data from several channel systems. But more significantly, the data from multiple systems may be noncomparable (Gaski 1986). Indeed, the findings of Lambert's (1978) 49 exploratory study of eighteen firms across three industries indicated that data collected from firms in the same industry may not be comparable. The issue of noncomparable data is especially relevant because the focus of most empirical channels research currently is development of measures of key constructs. Often the specific Operationalization of a construct depends upon the channel situation. Channel member satisfaction is a prime example of such a construct. Using multiple channel systems would force the researcher to emphasize the commonalities among those systems and in so doing one would probably overlook those things that really mattered in producing satisfaction (Ruekert and Churchill 1984). For these reasons it seems appropriate to use a single channel system as the research setting when conducting channel research. However, strictly speaking, the research findings cannot be generalized but must be replicated to evaluate applicability to other channel systems. Key Informant Method The key informant method has been the predominant method of data collection in channel research. Phillips (1981) criticized the use of a single key informant because he argued that it is difficult to assess the reliability and validity of measurements obtained from a single source. That is, one cannot model variance attributable to methods. He concluded that: "... the degree to which informant reports are valid indicators of the organizational characteristics they are intended to measure is an unresolved issue." Essentially what Phillips wanted was proof that informant reports are valid and he suggested that channel researchers gather data from multiple key informants. However, this recommendation does not reflect a thorough understanding of the ownership 50 and management structures of most channel intermediaries. Empirical research shows that many distributorships, dealerships and franchises are owned and managed by a single individual. When a single individual is the only qualified respondent to a survey, use of the multiple informant method is not applicable and may introduce “noise" into the data. Frazier and Summers (1984) stated that: “While Phillips (1981) emphasizes the desirability of utilizing multiple respondents within an organization when using the key informant method, he also notes the necessity of verifying that all such respondents are qualified to provide the requested information." Thus, while channel researchers acknowledge the need to gather data from all relevant parties, in most instances there was a single principle from whom it made sense to gather information (Brown and Frazier 1978; El-Ansary and Stern 1972; Etgar 1976a,b; Frazier and Summers 1984; Hunt and Nevin 1974; Gaski 1985; Gaski and Nevin 1986; Lusch 1976a,b, 1977; Lusch and Brown 1982; Michie and Roering 1978; Rosenberg and Stern 1971; Ross and Lusch 1983; Ruekert and Churchill 1984; Wilkinson 1974, 1981). These authors asked qualifying questions to determine whether one or more individuals should be surveyed in each firm and to specifically identify the individual(s). Usually the firms were contacted prior to the survey to ascertain this information. In addition, the authors reported that the questionnaires typically contained one or more questions to confirm that the respondents were qualified to participate in the survey. For example, respondents may be asked to indicate whether they are the principal owner/manager of the firm. Thus, it seems that the major consideration in selecting partici- pants for a field study is to evaluate whether each individual is 51 qualified to represent the organization. Given the availability of qualified multiple key informants, the researcher should solicit responses from all qualified individuals (Phillips 1981). However, the ownership and operating structure of many channels precludes the use of multiple informants, and makes it necessary for the researcher to rely on a single key informant. SUMMARY OF THE LITERATURE AND MAJOR CONCLUSIONS Channel theory and research has been dichotomized into two approaches, behavioral and economic. Several marketing scholars have urged channel researchers to adapt an integrated approach that encom- passes both the behavioral and economic dimensions of marketing channels (Frazier 1983b; Frazier and Summers 1984; Stern and Reve 1980). Of the three constructs of interest in this dissertation, channel member satisfaction has received the most attention. Channel member performance and dependence have been all but ignored in the literature. Perhaps this lack of attention is because both performance and dependence have been depicted as economic variables in empirical studies (Etgar 1976b; Kelly and Peters 1978; Lusch 1976b), and for almost three decades most channel research has centered on behavioral constructs. Several major conclusions can be drawn about this research stream that may benefit future researchers: 1. An Integrated Approach. Marketing scholars have criticized the emphasis given to the behavioral dimensions of channels for two reasons: the variables are difficult to measure and, the approach does not accurately represent the manner in which channels operate (Frazier 1983b; Frazier and Summers 1984; Lambert 1978; Stern and Reve 1980). Stern and Reve and Frazier argued that the economic and behavioral dimensions of 52 channels are inextricably linked. They urged researcher to adOpt an integrated approach that reflects this interrelationship (Frazier 1983b; Stern and Reve 1980). The key point is that marketing channels may be behavioral systems, especially vertically administered channels but the participants are economically motivated (Bowersox, Cooper, Lambert and Taylor 1980). Firms join together to form marketing channels because they can enjoy mutual economic gains. Thus, it is not realistic or managerially useful to study the behavioral aspects of channels in isolation. Future researchers should adopt an approach that integrates the behavioral and economic characteristics of marketing channels. 2. The Conceptual Framework. The behavioral approach to channel research cannot be credited with major, managerially useful advances in channel management. This is not too surprising given that channel member performance, the variable of utmost importance to management, has received very little attention in the literature. Other important economic variables such as financial dependence have also been neglected. Thus, key variables have been omitted from consideration. In addition, the conceptual model that has developed primarily focuses on three variables: power, conflict and satisfaction. Any model that adopts such a narrow perspective is unlikely to produce managerially meaningful findings. A secondary problem involves the legitimacy of the three constructs addressed by behavioral researchers. Both the power and conflict constructs are not well defined. Hence, some marketing scholars have questioned whether these two constructs measure what they purport to measure (Frazier and Summers 1984; Gaski 1984). 53 In summary, the conceptual model must be broadened to include channel member performance and related economic variables that are of critical concern to practitioners. In addition, each variable must be well defined to facilitate consistent and meaningful Operationalization. 3. Construct Validity. Previous channel researchers placed little or no emphasis on the validity of the behavioral constructs they studied (Gaski 1984). Few empirical studies reported on the reliability of the measures employed (Dwyer 1980; Lusch 1976a,b, 1977; Michie and Roering 1978; Phillips 1981; Ruekert and Churchill 1984; Wilkinson 1981). Only two studies reported evidence of convergent or discriminant validity (Phillips 1981; Ruekert and Churchill 1984). However, as Gaski (1984) pointed out, much of this work was completed before marketing scholars attributed increased priority to construct validity. 4. Objective Measures. Whenever possible researchers should use objective measure rather than perceptual measures. Etgar (1976b) employed both types of measures to Operationalize channel member depen- dence but Phillips (1981) showed that the perceptual measure lacked reliability and validity. In constrast, two of the objective measures Etgar (1976b) used correlated significantly with other variables Of interest. Similarly, two authors who investigated channel member performance used Objective measures and reported statistically signifi- cant findings (Kelly and Peters 1978; Lusch 1976b). Conversely, Gaski and Nevin (1985) employed a perceptual measure of performance comprised of two items that did not demonstrate adequate reliability or validity. In reference to the increased emphasis given to behavioral constructs that are usually Operationalized with perceptual measures, Lambert (1978) concluded that: "At the very least there are severe 54 implementation problems with this behavioral orientation, and the potential for application by managers is doubtful." 5. Measures. Satisfaction with the supplier's role performance is an important component of channel member satisfaction. Researchers who employed a role performance measure of channel member satisfaction reported that the measure had sufficient content validity (Lusch 1977; Michie and Roering 1978; Ruekert and Churchill 1984). Other researchers used role performance measures to operationalize power or power sources and they also indicated that the measures possessed content validity (Frazier 1983a; Guiltinan, Rejab and Rodgers 1980; Hunt and Nevin 1974; Lusch 1977; Michie and Roering 1978; Ross and Lusch 1983; Wilkinson 1981). Even so, the content validity of these measures is questionable because most of these studies used multiple items but did not identify multiple dimensions Of role performance. Ross and Lusch (1983) and Ruekert and Churchill (1984) provided evidence of the multidimensionality of role performance. The dimensions reflected the busineSs functions the supplier performed, so those authors used different sets of items to measure each dimension. Future researchers should identify all functions of the supplier's role and devise separate sets of items to represent each function. In addition, the items should be complete enough to enable the study findings to provide useful managerial guidance. 6. Scales Employed. When measuring the role performance dimensions of channel member satisfaction, it does not improve the predictive power of the model to have respondents evaluate the importance of services provided by the supplier (Lusch 1977; Lusch and Brown 1982). Asking 55 respondents to indicate their beliefs about the supplier's performance of the services is sufficient. 7. Research Setting. Almost all field studies in recent decades have been confined to a single channel system (Brown and Frazier 1978; El—Ansary and Stern 1972; Etgar 1976a,b; Frazier and Summers 1984; Hunt and Nevin 1974; Gaski 1986; Gaski and Nevin 1985; Lusch 1976a,b, 1977; Lusch and Brown 1982; Michie and Roering 1978; Rosenberg and Stern 1971; Ross and Lusch 1983; Ruekert and Churchill 1984; Wilkinson 1974; 1981). Two reasons have been cited for this convention. The first reason for studying a single channel system is the intractable cost and effort required to collect data from multiple channel systems. Second, the data from multiple systems may be uncomparable (Gaski 1986). The findings Of Lambert's (1978) exploratory study of channel management in eighteen manufacturing firms suggested that the data most likely would not be comparable from one firm to another, even if the firms were in the same industry. Consequently, using multiple channel systems would require researchers to emphasize commonalities in developing and measuring many constructs, like channel member satisfaction, and that would obfuscate the important dimensions of the satisfaction construct. That is, the researcher may overlook factors that really mattered in generating satisfaction or dissatisfaction (Ruekert and Churchill 1984). Thus, channel researchers should continue to focus on a single channel system when conducting empirical work. 8. Key Informant Method. Most mail surveys of channel systems targeted the intermediary, that is, a distributor, dealer or franchisee (Brown and Frazier 1978; El-Ansary and Stern 1972; Etgar 1976a,b; Gaski and Nevin 1985; Lusch 1976a,b, 1977; Lusch and Brown 1982; Michie and 56 Roering 1978; Rosenberg and Stern 1971; Ross and Lusch 1983; Ruekert and Churchill 1984; Wilkinson 1974; 1981). In all of these studies the authors obtained data from a single key informant. Although Phillips (1981) criticized the use of a single key informant, it is sometimes the only logical approach. If the decision making unit for the issues relevant to the study is comprised of one individual, the single key informant method should be used. Many dealerships, distributorships and franchises are Owned and managed by a single individual. In these situations, use of the multiple informant method of data collection would most likely introduce "noise“ into the data. Future researchers should attempt to identify the members Of the decision making unit prior to mailing the survey. If the decision making unit is comprised of multiple participants, the multiple informant method should be used. CHAPTER 3 RESEARCH DESIGN Chapter 3 describes the research design and methodology employed in this study. The chapter begins with a brief overview of the research process followed by more in-depth discussion of the conceptual framework and the research methodology. Each section will be further subdivided into the relevant component parts. The chapter will end with a summary of the entire research design. OVERVIEW OF THE RESEARCH PROCESS This section will present the research objectives of the study and then briefly outline each of the two main elements of the research design. Research Objectives As noted previously, channels researchers have recognized channel member satisfaction as an important variable in channel management but they have given scant attention to channel member performance and depen- dence. NO empirical research has been reported that investigated the manner in which performance, satisfaction and dependence are related. The primary purpose of this study was to suggest how the three variables are related and to empirically test this conceptual scheme. HOpefully, the findings will stimulate further conceptual and empirical develOpment 57 58 of these relationships so that managerially useful guidelines can be developed. Specifically, the objectives of this study were: 1. To integrate the behavioral and economic approaches to marketing channels theory; 2. To conceptualize the interrelationship among channel member performance, satisfaction and dependence in a theoretically and managerially sound way; and, 3. To empirically test the conceptual scheme. The Conceptual Framework The constructs of interest in the study were: channel member perfor- mance; satisfaction; dependence; reinvestment; and, attributions Of responsibility. In addition to presenting the research hypotheses each variable is defined and suggestions are made for how the variables should be Operationalized. Research Methodology The purpose of this section is to outline the process through which measures of the variables were developed and empirical data collected and analyzed. Specifically, the section will discuss: the research setting; measurement of the variables; data collection methods; and, analysis procedures used to purify the measures, assess reliability and validity, and test the research hypotheses. THE CONCEPTUAL FRAMEWORK An incomplete conceptual framework has been repeatedly cited as a major shortcoming of past empirical channel research (Frazier 1983b; Frazier and Summers 1984; Gaski 1984; Ruekert and Churchill 1984; and, Sheth and Gardner 1982). The purpose of this section is to present the conceptual framework that served as the basis for the research. Two 59 additional steps important in develOping a conceptual framework are the explicit definition of each variable to be studied and and an explanation Of how each variable should be operationalized (Churchill 1979). Given explicit construct definitions and guidance as to the Operationalization of the constructs, measures can be developed that facilitate an empirical assessment of the research hypotheses. The variables of interest in this study will be discussed in turn and the discussion will be comprised of two parts: variable definition and operationalization of the variable. But first the research hypo- theses will be presented. Research Hypotheses Referring to the focus of channel research, Lambert (1978) commented: It is interesting to note that the majority of the channels literature is concerned with the 'soft' behavioral aspects of channel management even though these kinds of data are the most difficult to measure and evaluate. The more easily quanti- fiable areas of cost trade-offs, and performance measurement and evaluation have received very little attention in the literature although the payoffs in these areas would seem to be the greatest. Although the rationale commonly given for investigating the behavioral aspects of channel relationships is management's desire to influence channel member perceptions, decisions, and performance, most empirical studies did not include performance. Instead, empirical research has addressed behavioral variables like power and conflict without evaluating the relationship between these variables and performance. Nevertheless, given that management's interest in influencing the performance of channel members provided the rationale for studying power, it seems quite remarkable that channel member performance has received so little 6O attention in the literature. In addition to a concern with performance, suppliers have a desire to establish long lasting relationships with intermediaries. Since the intermediary enters the relationship to achieve economic gain, suppliers need to know how to encourage inter- mediaries to maintain, and even increase their financial dependency. So, the two variables of primary interest to managers are performance and dependence, and the interrelationship Of the two. Also, managers have a keen interest in their role partners' reinvestment intentions. Previous channel research focused on satisfaction and, power and conflict as determinants of satisfaction. However, channel relationships are business partnerships based on interdependency and mutual profit. Therefore, it seems unlikely that practitioners harbor much concern with channel member satisfaction, except as it relates to performance, depen- dence and reinvestment in the business. If channel member satisfaction does influence performance, dependence and/or reinvestment it becomes a relevant managerial concern. Otherwise, managers would have little interest in strategies useful for improving satisfaction.' Therefore, this research addressed the interrelationships among channel member performance, dependence, reinvestment and satisfaction that were suggested in the literature. Figure 3-1 is a causal model that depicts the prevailing view of the interrelationships among channel member performance, satisfaction and dependence. Chapter 2 presented the relevant literature and suggested that a dominant or prevailing view exists regarding the relationships among these constructs. 61 Figure 3-1. The Prevailing View The specific research hypotheses implied by the Prevailing View were: H1: A channel member's satisfaction will directly influence the channel member's performance, dependence and reinvest- ment. H2: A channel member's dependence will directly influence the channel member's performance. H3: A channel member's reinvestment in the business will directly influence the channel member's performance. The data collected in this study were used to empirically test the Prevailing View and an alternative view of the relationships among channel member satisfaction, dependence and performance. The alternative view was suggested in Chapter 2 by the author and is shown in Figure 3-2. Hereafter it will be referred to as the "Alternative Model.“ Note that this causal model includes an additional variable called CREDT. CREDT represents the attributions of responsibility construct which was discussed in Chapter 2. Briefly, the CREDT variable reflects the fact 62 that intermediaries assign credit or blame for their performance to themselves, their role partners and/or the situation. The research hypotheses associated with the Alternative Model were: H1: A channel member's dependence directly influences the member's reinvestment in the business. H2: A channel member's dependence directly influences the member's performance. H3: A channel member's reinvestment in the business directly influences the member's performance. H4: A channel member's performance directly influences the member's attributions of responsibility. H5: A channel member's attributions of responsibility directly influence the member's satisfaction. RNV$T DEP$$. Figure 3-2. An Alternative Model Definition and Operationalization of Variables A major shortcoming of previous channel research was the failure to explicitly define the constructs and to suggest how the constructs should be operationalized. Thus, the definition and Operationalization of each variable is described in the following section. 63 Performance. Channel member performance herein refers to the intermediary's performance and can be evaluated by a number of role partners that can be categorized into two groups, customers and suppliers. The first group of role partners are those individuals or institutions served by the channel member. The second group is comprised of those firms that supply goods or services to the channel member and may include manufacturers or other channel intermediaries. Aside from its role partners, the intermediary may perform a self-evaluation. Here the performance construct was limited to the supplier's view of the intermediary's performance. The author does not mean to suggest that the customer's view is unimportant but that consideration of it is beyond the scope of this study. The purpose of the research was to investigate the relationship between the channel intermediary's: satisfaction with its major supplier; level of dependence; and, performance as perceived by the supplier. In other words, the research sought to answer questions such as: Is there a significant relationship between the intermediary's dependence and performance as a business partner? Does a high level of dependence lead to a high level of performance? Given the focus of the research, channel member performance Was defined as the degree to which the channel intermediary engages in behavior that contributes to the fulfillment of the supplier's objec- tives. The performance variable should be Operationalized by obtaining the supplier's evaluation of the channel intermediary's performance. The performance evaluation system should include both task and profitability measures. Ideally, the supplier has a formal performance evaluation system in place. Using this system, the supplier evaluates each of its intermediaries on all relevant dimensions and computes an overall 64 performance score for each one. In this situation the research should use the supplier's performance data. It is possible that the supplier may not conduct a formal perfor- mance evaluation of its intermediaries or may conduct an incomplete or otherwise inadequate evaluation. In these situations, the researcher should consider choosing a new target organization as the focal point for the study, or accept responsibility for guiding the target firm in developing performance evaluations of its intermediaries. With the researcher's assistance, the supplier should devise performance measures and standards, and uniformly apply them to its intermediaries. Satisfaction. Channel member satisfaction has been recognized as a key construct among channel theorists and researchers, but in spite of the attention satisfaction has received its conceptual and operational development are in need of further refinement (Ruekert and Churchill 1984). When conceptualizing the concept of satisfaction it is important to ask, "satisfaction with what?" A review of past channels research indicated general agreement that the appropriate answer is satisfaction with the relationship between the channel member and another channel participant. All past studies of satisfaction have focused on the relationship between a channel intermediary and a major or sole supplier. Previous studies investigated the intermediary's satisfaction with the relationship. In this study, channel member satisfaction was conceptualized similarly and the Ruekert and Churchill definition was adopted. Specifically, they suggested that channel member satisfaction encompasses all facets of the relationship between the channel member and an exchange partner that the member finds satisfying or unsatisfying. 65 Channel member satisfaction should be operationalized as a multi- dimensional construct. The dimensions Of satisfaction may differ somewhat among different channels but generally the relevant dimensions will reflect the channel member's satisfaction with the supplier's role performance and the intermediary's financial returns. Previous research has shown the role performance dimension to be a multi-faceted variable. The primary elements of role performance reflect the various business functions associated with the supplier's role and will depend upon the Specific channel system under study. The starting point for developing a satisfaction measure would be to identify all business functions performed by the manufacturer. This task can best be accomplished by interviewing key managers for the manufacturer. Next, a set of items should be generated to represent each business function. The items should be worded such that they provide managerially useful information. To simply state, “mechanic training," typical of items used in past studies, is not sufficient to guide management action. This point can be illustrated by trying to interpret the meaning of a low performance rating for "mechanic training.“ Would this imply a problem with the content of the training, the delivery method, the availability of training, or something else? It should be clear that items used to measure supplier role performance must be quite specific. A comprehen- sive and meaningful set of items should be developed through in-depth interviews with members of the target population. It is important to note that even if a business function is germane to different research settings, the set of items used to represent that function must be empirically derived. That is, one set of items may not accurately represent the same function in different channels due to 66 differences in the scope of the function. For example, product related support might be important to both an office equipment distributor and a fast food franchisee but the specific types of support desired may vary considerably between the two. Indeed, differences in the type Of support provided by the supplier in a given area may exist even among suppliers in the same industry. As an example, it is common practice for fast food franchisors to provide franchisees some support in the area of real estate and construction. However, one franchisor may actually perform location analysis on new store sites and make the site selection decision while another franchisor may simply perform an approval function. In the latter case, the franchisee conducts the location analysis. Given these differences, the real estate and construction dimension of role perfor- mance could not be validly operationalized in both channel settings using the same items. In conclusion, the dimensions and associated items of role performance may vary among channels and should be empirically derived and validated. The financial dimension of satisfaction should capture two key assessments made by channel members regarding financial rewards. First, channel members compare the adequacy of their actual results to their prior expectations. If the financial rewards achieved through the relationship exceed the channel member's prior expectations then they are satisfied. The level of satisfaction is positively related to the magnitude of the differences between prior expectations for financial rewards and the actual financial rewards received. The second key assessment made by channel members involves the equity of the distribution of the financial rewards between the supplier 67 and the intermediary. The channel member determines the level of equity by making two comparisons (Frazier I983b). First, the intermediary compares the ratio of its rewards to its contributions. Second, the ratio of the supplier's rewards to the supplier's contributions is esti- mated. Finally, these two ratios are compared. 'If each party's rewards relative to their contributions is equivalent, then the distribution of financial rewards is equitable and this contributes to satisfaction with the relationship. Dependence. The conceptualization of channel member dependence should reflect the following characteristics of marketing channels: the voluntary participation of the parties; profit as the prime goal; and the long term nature of most channel relationships. In the channels literature dependence has been conceptualized as an economic variable. For purposes of this study, channel member dependence was defined as the extent to which a channel member relies on this relationship to achieve or maintain a desired level of economic status. This definition was suggested by Etgar (1976). I In the short term, a channel participant may have limited alter- natives to the current relationship but long term most can exercise other options. This holds true for both partners to the exchange and is especially characteristic of vertically administered channels where the intermediary has only one supplier. A franchise system is an example of this type of channel. Most franchisees are bound by long term contracts that are not easily or quickly broken. Channels are characterized by task Specialization which leads parti- cipants to adopt complementary roles. Subsequently, each party usually makes significant recurring investments in plant and equipment related to 68 their role. Thus, the concept of dependence applies to both the inter- mediary and the supplier. In this study, the intermediary's dependence was investigated. Dependence should be operationalized by employing Objective indicators Of the intermediary's financial dependence. Specifically, measurement of dependence Should capture the following: . the percentage of total income derived from the relation- ship; and, . the percentage of total net worth invested in the rela- tionship. Reinvestment. Reinvestment of some percentage of profits is neces- sary in all businesses. In vertically administered channels like fran- chise systems it is important that franchisees reinvest profits to replace and/or upgrade plant and equipment, to expand existing outlets and to add outlets. Thus, reinvestment is defined as the extent to which a channel participant intends to reinvest profits in the business. This studyis concerned with the franchisee's reinvestment not the franchi- sor's. The reinvestment variable can be Operationalized as a percentage of the profits the franchisee intends to reinvest in the business. Per- centage of profits is an appropriate measure because it enables the researcher to make meaningful comparisons among franchisees with substan- tially different levels Of profitability. Attributions of Responsibility. Attributions of responsibility refers to the extent to which channel participants assign credit or blame for their performance to themselves, their partners or the situation (Frazier 1983). The study investigated attributions of responsibility from the franchisee's standpoint. That is, the study focused on the 69 extent to which franchisees' assign credit or blame for their performance to themselves, the franchisor or the situation. The attributions of responsibility variable can be operationalized by asking respondents to distribute 100 points among self, suppliers (role partners) and the situation. This type of measure most accurately reflects the fact that channel participants attribute responsibility for their performance in a relative manner. The next section will outline the research methodology used in the study including the research setting; measures employed; data collection methods; and, the data analysis procedures used to assess the reliability and validity of the measures and to test the research hypotheses. RESEARCH METHODOLOGY Research Setting It is customary in channels research to investigate the charac- teristics of a single channel system (El-Ansary and Stern 1972; Etgar 1976a,b; Frazier and Summers 1984; Hunt and Nevin 1974; Gaski 1986; Gaski and Nevin 1985; Lusch 1976, 1977; Lusch and Brown 1982; Rosenberg and Stern 1971; Ruekert and Churchill 1984; Wilkinson 1974, 1981). AS noted in Chapter 2, it is essential to study one system at a time because the participants in multiple channels would likely interpret the measures differently. In keeping with the practice, data for the study were collected from a single channel system, a fast food franchise system. The franchisor was a wholly owned subsidiary of another firm. In addition to its network of franchisees, the franchisor operated company- Owned stores. A major tenet of the franchisor's strategic plan was to achieve much of the firm's expansion goal through the growth of the franchise system. This would occur by encouraging existing franchisees 70 to reinvest in the business so they could expand their existing stores and/or open new stores. Additionally the firm sought out new qualified franchisees. The firm had 229 franchisees at the time the data for the study were collected and its heaviest concentration of stores was in California, Arizona, Texas and Florida. Most franchises were owned and operated by a single individual and franchises varied considerably regarding such characteristics as: number of years with the franchise system; number of stores operated; geographical location; and, education. Finally, although this channel System could be characterized as a contractual system, the contract did not prevent the franchisees from having other businesses or jobs, and a significant number did own other businesses or had other jobs. Measures Employed The content validity of most of the measures used in previous studies of channel member satisfaction, performance and dependence was questionable. For that reason it is important to explain the process used to assess the content validity of the measures used here. However, before discussing this process in detail the measures used to represent the constructs are described. Performance. The franchisor investigated in the study had a formal performance evaluation system in place and each franchisee was evaluated annually. Performance was evaluated in four key areas: operations; organization; financial; and, upgrades/development. On the basis of franchisee performance in these four areas, an overall composite rating was assigned to each franchisee. The franchisor provided the researcher 71 with these composite scores for use as a measure of franchisee perfor- mance. The performance scores used in this research study reflected the franchisor's evaluation of franchisee performance during the year in which the study was conducted. Dependence. Three measures of financial dependence were employed in the study. Franchisees were asked to indicate the percentage of their total annual income and net worth derived from the franchise respec- tively. In addition, franchisees reported whether they had a job other than the franchise or another business. Reinvestment. The respondents indicated the percentage of profits they intended to reinvest for the current year and two succeeding years. Attributions of Resppnsibility. The attributions of responsibility variable was measured by asking respondents to distribute 100 points among three factors that may have contributed to their performance: 1) their own performance; 2) the franchisor's performance; and, 3) other situational factors such as economic conditions. The factor believed to have been most instrumental in the franchisee's performance was assigned the highest number of points. Satisfaction. Three measures of channel member satisfaction were employed in the study, two Single item and one multi-item, multi- dimensional measure. The single item measures dealt with satisfaction with role performance and whether franchisees would choose to become franchisees of this system if they had it to do over again. Likert scales were used in both cases. Regarding role performance, respondents were asked to mark a point on a line that best expressed their level of satisfaction with the franchisor's overall performance. The line was 72 anchored by the words “poor“ and "excellent.“ A midpoint was placed on the line to correspond to satisfactory performance. Similarly, franchisees indicated whether they “would do it again "by circling a number on a scale from 1 (Strongly Disagree) through 5 (Strongly Agree). The multi-item, multi-dimensional measure of channel member satisfaction reflected satisfaction with the interorganizational climate, the franchisees' financial returns and the franchisor's role performance. Respondents used a five-point Likert scale to indicate their agreement or disagreement with statements regarding the interorganizational climate and their financial returns. The franchisor's role performance in a number of areas was evaluated by using a seven point scale ranging from 1 (Poor Performance) through 7 (Excellent Performance). The items used to measure satisfaction with role performance were empirically derived as recommended by the literature. The items used to measure satisfaction with interorganizational climate were derived from the literature and discussions with practitioners. Finally, the items used to measure satisfaction with the franchisee's financial returns were developed by the researcher because the literature contained no Specific suggestions. Therefore, the researcher generated six items that appear in Figure 3-3. Three of the Six items used to measure satisfaction with financial returns reflected the comparison Of actual returns to initial expectations. The other three related to the equity of the distribution of the returns. Unlike the role performance dimension of satisfaction, the inter- organizational climate and financial returns dimensions may be widely applicable to a variety of different channels with little content change. Insufficient empirical data exists to support any firm conclusion, but 73 logic suggests that the intermediary's concerns with the general climate and the adequacy and equity of their financial returns would be fairly uniform across channels. Nevertheless, the climate and financial returns measures must be subjected to empirical confirmation. Equity of Returns In relation to the franchisor's investment in time and money, the monetary rewards received by the franchisor are quite reason- able. Given our respective investments, the rewards are equitably distributed between myself and the franchisor. My ratio of rewards to effort is comparable to the franchisor's ratio of rewards to effort. Rewards Compared to My actual monetary rewards compare favor- Initial Expectations 1 ably to my initial expectations. My monetary rewards exceed my initial expectations. My rewards compare favorably to my initial expectations. Figure 3-3. A Measure of Satisfaction with Financial Returns. Discussions With Franchisor Management and Industry Experts The researcher engaged in extensive discussion with key managers in the franchisor's organization for the following reasons: 1) to identify the business functions performed by the franchisor; 2) to generate items useful in measuring each function; 3) to confirm and/or revise the items used to measure the financial returns dimension of channel member satisfaction; 4) to identify specific measures of financial dependence and including reinvestment intentions; 5) to ensure that questions were 74 worded such that respondents to the mail survey would interpret them in the same way; and, 6) to identify key demographics that may be useful in explaining differences in satisfaction, performance and dependence among respondents. Corporate and regiOnal executives with both line and staff respon- sibilities were targeted for these discussions. AS a group, these managers represented all of the functions performed by the franchisor. Many of these individuals participated in several different discussions because development of the measures was an interactive process. Upon completion of the final version of the research instrument, it was submitted to the senior executives of the franchisor organization for approval. In addition to discussions with franchisor management, several experts in the industry were interviewed to garner additional insights into industry practices. In particular, information was gathered about the division of the business functions between the franchisor and franchisee in the fast food industry. The researcher attended the Multi-Unit Food Service Organization's (MUFSO) Annual Meeting where She met and interviewed many senior executives of fast food companies. MUFSO is the major trade organization for the industry. In addition, she attended several meetings with industry leaders that were set up by the faculty of the Hotel, Restaurant and Institution Management (HRI) Department at Michigan State University. These discussions with franchisor management and industry experts enabled the researcher to identify eight dimensions of franchisor role performance: product, promotion; price; physical distribution/customer service; Operations; real estate and construction; personnel; and, training. follows: 1. 75 Specifically, the eight dimensions were characterized as A product dimension that included new product development and introduction, food preparation, quality, and pack- aging; A promotion dimension that included national promotions, local promotions, merchandising programs, new product promotions, franchisees' ability to provide input to promotions and/or to preview promotions, the extent to which promotions stimulate sales and the return on promotion dollars; A pricing dimension that encompassed recommendations from the franchisor regarding new prices, margins, and supply costs; A physical distribution/customer service dimension that assessed how well the franchisor supports the franchisee in ordering, Shipping and billing for supplies; An operations dimension that reflected a variety of assistances the franchisor provides the franchisee such as assistance with equipment repair and maintenance, labor scheduling, record keeping, business plan development and sanitation; A real estate and construction dimension that captured the support services related to site location, remodeling, and building and leasing stores; A personnel dimension that included franchisor assistance with hiring, staffing, salary and termination guidelines; and, A training dimension that reflected support of crew training, provision of manuals, films and a training center. Also, the discussions provided confirmation of the items used to measure satisfaction with the interorganizational climate, financial returns and the two Single item measures Of satisfaction. In addition, the measures of channel member dependence, the franchisee's willingness to reinvest in the business and key franchisee demographics were 76 identified. All of these measures were further refined by interviewing a sample of potential respondents. In-Depth Interviews with Franchisees In-depth interviews lasting two to five hours were conducted with fifteen of the 204 franchisees qualified to participate in this study. As a group, these franchisees represented a cross section of those comprising the total franchise population. That is, the fifteen fran- chisees varied considerably regarding key demographics such as years with the franchise system; number of stores operated; age; education; and, geographical location. For example, the geographical locations repre- sented by the fifteen franchisees were as follows: . California ............ 4 . Florida ............... 4 . Texas ................. 3 . Georgia ............... 2 . Illinois .............. 1 . Wisconsin ............. 1 There were four objectives for conducting the in-depth interviews. The first objective was to gain general knowledge about the industry and more specific information about this franchise system. The second objective was to refine the measures of the variables, especially channel member satisfaction. This included adding, deleting and/or revising (the wording of) items to ensure that the list of items was comprehensive and meaningful to the franchisees. That is, this process provided assurance of the measures' content validity. The third objective of the in-depth interviews was to revise the list of key franchisee demographics to reflect the franchisees' input. The fourth objective was to facilitate 77 modification of the instructions contained in the questionnaire to ensure that respondents to the survey would interpret them in the same way. The questionnaire format and layout were also reviewed to confirm that the franchisees understood the interrelationships of certain tasks and to reduce the time and effort needed to complete the questionnaire. In conducting the in-depth interviews an iterative process was employed which resulted in five revisions of the questionnaire. The revisions reflected the findings of successive in-depth interviews. The in-depth interviews resulted in many changes to the role performance dimensions of channel member satisfaction, one addition to the financial returns dimension of channel member satisfaction, and two new items useful in measuring overall satisfaction with the relationship. Specifically, the role performance dimensions were revised as follows: . Product - 6 questions added/revised 7 questions deleted/combined . Promotion - 2 questions added 2 questions deleted . Pricing - 2 questions added 2 questions deleted . Physical Distribution/ 15 questions added Customer Service - 3 questions revised 2 questions deleted . Operations - 7 questions added/revised 2 questions deleted . Real Estate and Construction - questions added/revised . Personnel - questions added/revised . Training - questions added/revised questions deleted «bNNU'l In all, the role performance measure contained 117 items distributed among the eight functions as follows: 78 Product ................................... 26 Physical Distribution/Customer Service .... 25 Operations ................................ 19 Promotion ................................. 14 Real Estate and Construction .............. 13 Pricing ................................... 9 Personnel ................................. 7 Training .................................. 4 In summary, the franchisor's data were used to measure franchisee performance in this study. Three perceptual measures of the franchisee's satisfaction with the relationship were employed. A multi-dimensional measure comprised of nine dimensions and 161 items was used. In addition to the eight dimensions of role performance, the measure included a communications dimension (38 items) and a financial returns dimension (6 items). Also, two single item measures were employed. Three objective measures of channel member dependence and reinvest- ment intentions respectively were used. Data regarding 27 demographic characteristics of the franchisees were also collected. Data Collection Methods AS noted previously, the data regarding the franchisees' performance were obtained from the franchisor. The data regarding the other con- structs of interest and demographic characteristics were collected via a mail survey of the franchisees. All franchisees in this franchise system were included in the study. The franchise system was comprised of fran- chisees with widely varying characteristics. For example, one franchisee operated over 200 outlets while others Operated just one or two stores. A mailing list was provided by the franchisor which contained 237 listings. However, eight names were listed twice because the list was arranged by zone and region, and these eight franchisees operated in more than one zone or region. The deletion of duplicate listings reduced the 79 number of franchisees to 229. Prior to the mail survey, the researcher telephoned all franchisees to apprise them of the research, to identify qualified informants, and to gain their agreement to answer the questionnaire. Franchise owners were listed on the mailing list, and it showed that the largest share of the franchises were owned by single individuals. The telephone calls confirmed that most franchises were single propri- etorships. In the case of multiple owners, telephone contact revealed that generally just one owner actively managed the franchise and the owners selected this individual to respond to the survey. The researcher accepted this posture as being logical and thus, collected data from the franchisees using the key informant method. Thus, the final number of qualified respondents was reduced to 204 from the original sample of 237 franchisees due to the following: . Duplicate listings of the same name .......... 8 . Multiple owners of the same franchise ........ l9 . Deletions due to retirement, sale or death ... 6 The response rate for the mail survey was 52 percent (107/204). In comparison to similar studies this was an excellent response rate and may be attributed to the pre and post survey telephone calls. Finally, early and late responses were compared and no statistically significant differences (p<.05) were found. Therefore, nonresponse bias did not appear to be a problem (Armstrong and Overton 1977). Data AnaLysis A limited information estimation procedure was employed to analyze the data collected in this study. Specifically, PACKAGE developed by Hunter, Gerbing, Cohen and Nicol (1980) was used. This is a two-Step 80 analysis procedure. First the measurement model is evaluated in order to assess measurement error. Secondly, the structural model is evaluated using a least squares estimation procedure which is preferred to maximum likelihood estimation when the sample size is small. The two-step procedure was most appropriate because both theory and measurement are not well developed in the marketing channels area. Thus, it is essential to employ a procedure that limits interpretational confounding (Anderson and Gerbing 1982; Burt 1976; Hunter and Gerbing- 1982). According to Burt, interpretational confounding occurs when an unobserved variable is assigned empirical meaning in terms of both epistemic and structural criteria. In other words, the meaning assigned to the unobserved variable is based on correlations between indicators of that construct (epistemic criteria) and also on correlations between constructs (structural criteria). In this case, the two-step procedure is superior to the more commonly used LISERL program (JBreSkog and SBrbom 1981), that evaluates the measurement and structural models simulta- neously. Burt reported that with a full information estimation procedure like LISERL the construct would be defined “in terms of both epistemic and structural criteria but in some unspecified, flexible ratio of the two criteria....“ Thus, the empirical meaning attributed to the construct may be very different from what the researcher intended. Also, if a full information estimation procedure is used and the theoretical model does not fit the data it is nearly impossible to know whether the problem is related to the measurement or structural model because the full information approach provides little guidance to the researcher as to how the model should be respecified (Anderson and Gerbing 1982). 81 According to Bagozzi (1980), the two-step procedure “keeps the interpre- tation of the theoretical variables constant in the analysis and makes for a more accurate estimation of the relationships between the theo- retical variables.“ Hunter and Gerbing (1982) offer another reason for using a two-step procedure. They presented empirical evidence that LISERL spreads misspecification error in the theoretical model over both the structural and measurement models. In Hunter and Gerbing's example, LISERL produced incorrect factor correlations even though a correctly specified measure- ment model was employed. This finding further supports the efficacy of using a limited information estimation procedure. Next the steps taken to evaluate the measurement model are described followed by a similar discussion of the structural model. The measurement model was evaluated by a five-step process as follows: 1. Established content validity; 2. Respecified the scales and assessed external consistency; 3. Evaluated the internal consistency of the scales; 4. Identified multi-dimensional constructs; and, 5. Created a correlation matrix of the constructs. Step One: The content validity of the items used in the scales was established through discussions with industry experts and franchisor management and through in-depth interviews with a sample of fifteen franchisees judged to be representative of the franchisee population. Content validity was established prior to data collection as it is essential to the develOpment of an appropriate research instrument. The 82 remaining four steps undertaken to evaluate the measurement model were completed following collection of the data. Step Two: The oblique centroid Multiple Groups Analysis (MPRG) program in PACKAGE (Hunter and Cohen 1969; Hunter et al. 1980) was used to develop the scales. Anderson and Gerbing (1982) suggested that MPRG be employed for this purpose because it produces similarity coefficients which offer some advantages over correlation coefficients: The problem with correlation coefficients is that they are too . general. Any two variables related by a linear transformation correlate perfectly; but proportionality is described only by those linear transformations with an intercept of zero. Thus an index is needed 'that does not reduce the data to deviation scores' (Hunter 1973). The result 2 r... r. ,. is the similarity coefficient. The square Of this coefficient is called the 'index of proportionality' by Tryon and Bailey (1970). The value of this index ranges from -1 to l with these extreme values representing perfect internal and external consistency. The usefulness of this coefficient for exploratory analysis of multiple-indicator measurement models is outlined by Hunter (1973): A matrix of indicator correlation coefficients can be trans- formed into a matrix of similarity coefficients and then ordered according to the following criterion. The first variable has the highest sum of squared coefficients with the remaining variables. The second variable has the highest coefficient with the first, the third has the highest coeffi- cient with the second, etc. The result is an ordering of the variables with relatively large drops in adjacent Similarity coefficients indicating the cluster boundaries. 83 All items retained for further analysis had a Similarity coefficient of at least .90 with alternative items of their respective constructs. This criterion was more stringent than the cut-off point of .80 suggested by Anderson and Gerbing (1982) as a useful guide when purifying measure- ments. The matrix of similarity coefficients also was examined to assess the parallelism or external consistency of the scale items. When the similarity coefficient for two items equals -1 or +1 the two items are perfectly parallel. Additionally, all items comprising a Specific scale should exhibit similar patterns of correlations with items making up other scales (Hunter and Gerbing 1982). Step Three: The next step in the process was to conduct a confir- matory factor analysis on the scales using MPRG in PACKAGE (Hunter and Cohen 1969; Hunter, et al. 1980). The MPRG program provides standard score coefficient alphas for each group Of multiple indicators. Coefficient alpha is a measure of a scale's internal consistency or reliability. Step Four: The development of the scales was an iterative process. Thus, the similarity coefficient matrix and confirmatory factor analysis were produced, evaluated and the scales respecified several times. Given satisfactory scales it was then necessary to assess the dimensionality of the constructs, particularly channel member satisfaction because previous studies depicted satisfaction as a multi-dimensional construct. The dimensionality of the scales was determined by creating a similarity coefficient matrix of the scales. If two scales had a similarity coefficient Of .90 or greater they were judged to represent different dimensions of a multi-dimensional construct (Hunter and Gerbing 1982). A confirmatory factor analysis was performed on the revised scales as 84 further evidence of dimensionality. For example, scales measuring satis- faction with various aspects of the franchisor's role performance were confirmed as multiple dimensions Of the channel member satisfaction construct. Each of the scales is discussed in Chapter 4 in detail. Step Five: After valid and reliable scales were obtained, the final step in the develOpment of the measurement model was the creation of the correlation matrix of the scales for use in evaluating the structural model. The correlation matrix is the input data used by the path program. The structural model was evaluated with the ordinary least squares routine in PACKAGE. Chi-squared is used to test the overall fit of the model. The customary alpha level of .10 was employed. Thus, p>.90 was interpreted as an indication of a good fit (Bagozzi 1980, Fornell and Larker 1981). Nested or hierarchical models can be compared by examining the difference in x2 associated with the two models. The difference between the X2 values is itself a X2 with degrees of freedom equal to the difference in the degrees of freedom related to the two models (Kenny 1979)~ If the difference in x? is not significant, the most parsimonious model should be retained (Fornell and Larker 1981; JOreSkog and SOrbom 1982; Kenny 1979). The researcher tested two competing models previously referred to as the Prevailing View and an Alternative Model. The two models were not hierarchical because the Alternative Model contained an additional variable, attributions of responsibility. However, each of the two models were respecified and the new versions were "nested" in the original models. Thus, the nested or hierarchical models were assessed using the procedure described above. 85 Finally, it is important to note that an established test for the significance of the structural parameters has not been developed. Although the test is theoretically possible the extensive mathematics involved has deterred researchers from pursuing its development. Thus the use of the causal modelling approach enables the researcher to evaluate the interrelationship of several variables simultaneously so the directionality of the links as specified by theory can be supported or rejected. But the strength of the relationship between any pair of variables cannot be assessed (Bagozzi 1980). SUMMARY In summary, Chapter 3 outlined the research objectives that guided the research, the conceptual framework that provided direction and scope, the research hypotheses, and the methodology which facilitated an empiri- cal test of the research hypotheses. The conceptual framework identified the variables of interest, defined them and suggested how the variables should be operationalized. Several research hypotheses were presented to test the proposed interrelationships among channel member satisfaction, performance, dependence, and related variables. The research setting, data collection methods and data analysis procedures were also described. Chapter 4 presents the research findings. CHAPTER 4 RESEARCH FINDINGS Chapter 4 reports the results of the data analysis. Previous channel research has been criticized for using unapprOpriate statistical analysis procedures and for not reporting evidence of the reliability and validity of the measures employed (Gaski 1984). In order to avoid those shortcomings a causal modeling approach to data analysis was used. A limited information estimation technique was used because it enables the researcher to evaluate the measurement model prior to testing the struc- tural model. Thus, measurement issues can be separated from theoretical matters. The two-step approach to data analysis avoids the problem of interpretational confounding and that is especially relevant here due to the limitations of previous research cited above. The chapter is divided into three parts. First, the findings relevant to the measurement model are presented. Second, the results related to the structural model are described. Finally, the chapter ends with a summary of the findings related to both the measurement and structural models. MEASUREMENT MODEL Chapter 3 described the five-step process employed to evaluate and refine the measurement model. Specifically, the five steps were: 1. Establishment of content validity; 2. Respecification of the scales and assessment of external consistency; 86 87 3. Evaluation of the internal consistency of the scales; 4. Identification of multi-dimensional constructs; and, 5. Creation of a correlation matrix of the constructs. The purpose of step one was to establish the content validity of the items that comprised the scales used in this research. The results of step one were presented in Chapter 3 so this section pertains to the remaining four steps in the process. Step two was undertaken to respecify the scales and to assess the external consistency of the items. Both tasks were accomplished by creating a matrix of similarity coefficients. The items exhibiting similarity coefficients of .90 or greater were grouped together to form scales. Respecification of the scales was made easier by reordering the similarity coefficient matrix so that items with high similarity coeffi- cients appeared adjacent to one another. Another benefit of ordering the matrix in this way is that it facilitated an assessment of the external consistency of the items. External consistency was evidenced by higher correlations between the items in the same scale than between items contained in one scale and items associated with other scales. Also, the pattern of correlations was similar for all items that comprised a particular scale. On the basis of these two criteria the researcher concluded that the scales evidenced sufficient external consistency. The purpose of step three was to confirm the reliability of the scales. Thus, the scales were subjected to confirmatory factor analysis using the MPRG routine in PACKAGE. Steps two and three were undertaken jointly in an iterative manner until the scales evidenced a high degree of external and internal consistency. Internal consistency was evaluated using the standard score coefficient alpha scores produced by the MPRG routine. 88 Steps two and three resulted in the identification of ten scales, one each related to channel member performance, dependence, reinvestment intentions, and attributions of responsibility and six scales associated with channel member satisfaction. Table 4-1 summarizes the findings relative to the scales. ‘Four Of the six scales associated with channel member satisfaction were believed to be different dimensions of one measure because these scales were based on items associated with the multi-item, multi—dimensional satisfaction measure used in the research. However, the dimensionality of the scales was the focus of step four in the evaluation of the measurement model so a brief discussion of each of the ten scales is presented next. Then the findings that pertain to the dimensionality of the measures are discussed. Table 4-1 Measurement Model No. of Similarity Coefficient Scale Items Coefficients 4 Alpha PERFM 1 --- --- DEP$$ 2 .90 .84 RNVST 2 .89 .84 CREDT 1 --- --- AGAIN I --- --- SATRL 1 --- --- Climate/Interaction 12 .92-.98 .92 Marketing - 20 .92-.98 .95 Product/Access 15 .93-.98 .93 Operations/Personnel 9 .93-.97 .90 89 As noted in Chapter 3, the performance measure used in the research was provided by the franchisor. The measure was comprised of a single item that represented a composite performance score. The measure was constructed by the franchisor on the basis of the franchisee's perfor- mance in four key areas: operations; organizatiOn; financial; and, upgrades/development. Only two of the three items included in the research as measures of dependence loaded on this scale. These two items were: percentage of. total annual income derived from the franchise and percentage of net worth attributed to the franchise. Whether the franchisee had another job or business was not a useful measure of financial dependence. The similarity coefficient for the two items that were retained was .90 and the scale had an alpha coefficient of .84. One of the three items believed to measure reinvestment intentions did not produce a similarity coefficient of .90 or greater with the other two items so it was drOpped from the scale. The two remaining items had a similarity coefficient of .89. Although the coefficient is Slightly below the cut-off point both items were included because the scale had a coefficient alpha of .84. By design, the attributions of responsibility (CREDT) variable was measured with a single item and a review of the matrix of Similarity coefficients revealed that this item was not highly related to any other item. Likewise, the confirmatory factor analysis provided evidence that the CREDT measure was a separate and unique construct. Six scales were related to channel member satisfaction and reflected the franchisees' satisfaction with: 9O . The franchisor's overall role performance (SATRL); . Their initial decision to join this franchise system (AGAIN); . Interorganizational climate/interaction; . Marketing support; . Product support/access to the franchisor; and, . Operations/personnel support. The two single-item satisfaction measures, SATRL and AGAIN appeared to measure related but separate constructs. That is, the two measures were related to each other and to the remaining four satisfaction scales but not at the level that would suggest that these scales measured the same construct. SATRL and AGAIN had a similarity coefficient of .78 and neither item had a similarity coefficient of .90 or greater with any of the items in the other four satisfaction scales. On this basis, SATRL and AGAIN were retained in the analysis; but they were recognized as separate albeit highly related constructs. The climate/interaction scale was comprised of twelve items with Similarity coefficients ranging from .92 through .98. The alpha coeffi- cient for this scale was .92. AS shown in Figure 4-1 the items measured the extent to which the franchisor provided support and coaching to franchisees and their level of agreement regarding business philOSOphy, goals and decision strategy. The marketing scale included twenty items representing the fran- chisees' satisfaction with the franchisor's performance of a variety of marketing activities. The twenty items reflected all of the marketing mix variables and are summarized in Figure 4-2. The items showed the franchisees' concern with quality products, new products, promotional support, customer service and competitive pricing. The similarity coefficients for the twenty items ranged from .92 through .98 and the 91 1. 10. 11. 12. The goals developed by the franchisor work to the mutual benefit of both the individual franchisee and the parent organization. My business philosophy and the franchisor's are very much alike. This franchise organization is highly interested in the welfare Of its franchisees. I am generally satisfied with my franchise. Through it, I am able to meet all of my professional goals. The corporate management of this franchise system is so far removed from my situation that their ideas Often do more harm than good. My franchisor and I have very compatible business goals. Once they've sold you the franchise, they just forget all about you, that is, until your fees are due again. My franchisor and I are very compatible on decision strategy. There is a definite lack of support, coaching and feedback from the franchisor. My franchisor's role performance has not met my initial expectations. My franchisor's support and interface with our franchisee organization or its representatives is excellent. My franchisor explains how and why changes are occurring in the franchise system, and their effects on the local environment. Figure 4-1. The Climate/Interaction Scale. 92 I. The extent to which a quality assurance program establishes and enforces standards for key product characteristics for appearance. 2. The use of merchandising programs for new products. 3. Your franchisor's assistance in developing quality standards. 4. The shelf life of raw ingredients. 5. The extent to which the products represent a good value to consumers. 6. Co-operative advertising funds provided by your franchisor. 7. Length of promised order cycle (lead) times on promotional materials. 8. The Shelf life of finished products. 9. Equipment provided for use in preparing raw products. 10. Equipment provided for use in preparing menu items. 11. The ability of the franchisor to meet promised delivery dates on new product introductions. 12. The quality or grade of raw ingredients. 13. Your franchisor's assistance in developing customer service standards. 14. Promotions support in general. 15. A new product testing program consistently applied to new products. 16. The number Of approved suppliers for purchase of food. 17. The number of approved suppliers for purchase of paper supplies. 18. The number of approved suppliers for purchase of equipment. 19. Advertising support for new products. 20. Competitiveness of retail prices. Figure 4-2. The Marketing Scale. 93 marketing scale was highly reliable as evidenced by an alpha coefficient of .95. Although the marketing scale contained a number of items related to product, the product/access scale included additional product related items. Ten of the fifteen items in the product/access scale related to product and many of the ten items dealt with new product develOpment and introduction. Three items reflecting the accessibility of middle and upper level executives in the franchisor's organization also loaded on this factor. A complete list of the fifteen items that made up the product/access scale appears in Figure 4-3. Table 4—1 reveals that the similarity coefficients for items in the product/access scale ranged from .93 through .98 and that the alpha coefficient was .93. The tenth scale identified from the matrix of similarity coeffi- cients reflected the franchiSor's role performance in the areas of Operations and personnel respectively. The operations/personnel scale had nine items with similarity coefficients of .93 through .97. The scale's reliability was evidenced by an alpha coefficient of .90. Figure 4-4 lists the nine items and shows that several reflect the franchisor's role performance relative to financial support. For example, assistance with recordkeeping, controlling variable expenses, inventory and costs of goods sold were included. Other items represented support with unemployment claims and personnel support in general. Finally, two items were related to promotions and physical distribution support: promotions for specific local events and a system for handling complaints and claims. In summary, ten scales were identified through steps two and three of the evaluation process but four of the scales were believed to 94 9. IO. 11. 12. 13. 14. 15. The ability of the packaging to help market the product. Competitiveness of the product line. The ability to minimize the time and effort required to prepare menu items for customers. The timing of new product rollouts. Periodic use of retail price discounts and coupons. Advance notice on new products. Product related support in general. The ability to minimize the time and effort needed to prepare raw products. The development and introduction of new products. The introduction of fully tested and proven new products. A new product testing program that encompasses all phases of operations and marketing. The accessibility of overall franchisor management. The accessibility of corporate franchisor management. The accessibility of regional franchisor management. Your ability to preview new advertising and promotions. Figure 4-3. The Product/Access Scale. 95 1. 2. Your franchisor's assistance in recordkeeping. Your franchisor's assistance in controlling variable expenses. Your franchisor's ability to educate you regarding unemployment claims. In general, franchisor support in the area of personnel. Your franchisor's assistance in inventory control. General pricing recommendations from the franchisor. From the franchisor, convenient and clearly communicated handling procedures for your complaints and claims. Your franchisor's assistance in controlling cost of sales. Assistance in developing advertising and promotional programs for special local events. Figure 4-4. The Operations/Personnel Scale. 96 represent separate dimensions of a multi-dimensional measure of channel member satisfaction. SO, the fourth step undertaken to develop the measurement model was an assessment of the dimensionality of the scales. Dimensionality was evaluated by creating a matrix of the similarity coefficients of the nine scales. If two scales exhibited a similarity coefficient of .90 or greater they were judged to be different dimensions of the same construct. As anticipated, the climate/interaction, marketing, product/access and operations/personnel scales were dimensions of the satisfaction construct. The matrix of similarity coefficients for the four scales is shown in Table 4-2. With one exception, these four scales had similarity coefficients of .90 or greater, the cut-off point adhered to by the researcher. Table 4-2 Similarity Matrix for Dimensions of Satisfaction* Operations/ Product/ Climate/ Marketing Personnel Access Interaction Marketing 100 Operations/Personnel 94 100 Product/Access 94 90 100 Climate/Interaction 92 85 92 100 *Decimals are omitted. The similarity coefficient for the operations/personnel and climate/ interaction scales was only .85. But because the other coefficients exceeded .90 the researcher judged these two scales to be dimensions of satisfaction. Subsequently, the four scales were summed to form one 97 scale and the revised set of scales were subjected to confirmatory factor analysis using the MPRG routine in PACKAGE. The MPRG routine provided an alpha coefficient for the newly formed scale that indicated a high degree of internal consistency (alpha=.97). Unfortunately the SATMD measure did not contain a financial returns dimension because none of the six items associated with financial returns had similarity coefficients of .90 or greater with each other or with any other items. Therefore, the SATMD measure must be interpreted in terms of its empirical content which did not reflect the franchisees' satisfaction with their financial returns. Finally, the fifth step in developing the measurement model was to produce a correlation matrix of the constructs. Note that the correla- tion matrix shown in Table 4-3 supported the conclusion that SATMD, SATRL and AGAIN are highly related but separate constructs. Table 4-3 Correlations Between Constructs* SATMD AGAIN SATRL CREDT PERFM DEP$$ RNVST SATMD 100 56 49 27 6 -6 -12 AGAIN 56 100 43 21 12 5 8 SATRL 49 43 100 34 3 -5 -1 CREDT 27 21 34 100 18 18 9 PERFM 6 12 3 18 100 28 DEP$$ -6 5 -5 18 28 100 15 RNVST -12 8 -1 9 7 15 100 *Decimals are omitted. It seems that each measure captured some unique aspect of the channel member's satisfaction. Thus, it was appropriate to include all three constructs in the structural model. 98 STRUCTURAL,MODEL Two models representing competing explanations of the relationships among the variables of interest were tested: the Prevailing View and the Alternative Model suggested by the researcher. First, the results of the path analysis for the Prevailing View are presented and then the findings related to the Alternative Model are discussed. However, before discuss- ing the results of the path analysis it is necessary to explain how the three separate satisfaction constructs were depicted in the models. It was hypothesized that SATMD was a determinant of SATRL. It would seem logical that satisfaction with the channel partner's performance of a variety of activities (SATMO) would determine one's overall level of satisfaction with role performance (SATRL) and would also affect the likelihood that one would become a member of the channel system if they had it to do over (AGAIN). Other than this adjustment, the two models were the same as depicted in Chapter 2. The ordinary least squares routine in PACKAGE was used to test both mOdels. As shown in Figure 4-5 the Prevailing View path model did not fit the data. The model had eight degrees of freedom, a x2 of 4.78 and p=78. The general rule is that a probability of more than ninety percent supports the conclusion that a theoretical model fits the data (Hunter and Cohen 1969). Thus, it was concluded that the Prevailing View could not be supported with these data. Subsequently, the researcher respecified the model in an attempt to improve the fit. Respecification is desirable if the researcher can base the model's revision on theory (Kenny 1979). A link was added from DEPSS to RNVST because most franchisees have no source of income aside from their franchise and thus might be motivated 99 X2 -4.78 df-8 p'78 Figure 4-5. The Prevailing View 100 to make reinvestments in order to protect their existing investment. The relationship between DEPSS and RNVST was suggested by the author in Chapter 2 and forms the basis for one of the research hypotheses contained in the Alternative Model. Therefore, it is appropriate to add the causal path between DEPSS and RNVST because the link can be justified on theoretical grounds. However, the revised version of the Prevailing View did not improve the model's fit. As shown in Figure 4-6 the fit (p=.75) was slightly worse. In contrast, the Alternative Model Shown in Figure 4-7 fit the data quite well (p>99.5). All of the path coefficients had a positive Sign as predicted and with the exception of one, the path coefficients were substantially different than zero. The path coefficient from RNVST to PERFM was only .03. Although there is no test to evaluate the signifi- cance of path coefficients it seemed reasonable to respecify the model, deleting the link from RNVST to PERFM. This version of the Alternative Model is shown in Figure 4-8. The respecified Alternative Model is more parsimonious than the initial version and fit the data equally well (p>99.5). That is, deleting the link from RNVST to PERFM did not have a negative effect on the model's fit. A third and final version of the Alternative Model was subjected to path analysis. The third version of the model included a link from AGAIN to RNVST. The purpose of this alteration of the model was to empirically test the hypothesis that satisfaction causes additional investments as suggested by Blau (1964) and Frazier (1983b). Shown in Figure 4-9, this path model did not fit the date as well as the preceeding model (p=99.5). Additionally, the path coefficient from AGAIN to RNVST was quite small (.07). 101 .49 43 .05 .11 x2 - 4.26 df - 7 p875 Figure 4-6. The Prevailing View: Added Link from DEP$$ to RNV$T RNVST .15 DEPSS X ' 2.93 Figure 4-7. 102 .03 .28 df 3 13 The Alternative Model SATMD .49 .46 103 RNVST \ . . 8 . _1, 1 27 SATMD .23 .45 DEP$$ .49 / .zo SATRL x2 . 3.19 df - 14 p>99.5 Figure 4-8. The Alternative Model Respecified: Deleted RNV$T to PERFM 104 .28 .18 .27 SATRL x2 - 3.19 df - 13 p - 99.5 Figure 4-9. The Alternative Model Respecified: Added AGAIN to RNV$T 105 A comparison of the fit of the structural models tested in this research is depicted in Table 4-4. Recall that when hierarchical or “nested" models are examined it is useful to assess the significance of Table 4-4 Fit of the Structural Models Prevailing View Alternative View x2 df p )12 01’ p 4.78 8 78 2.93 13 >99.5 Added DEP$$-—> RNVST 4.26 7 75 Deleted RNVST’--PERFM 3.19 14 >99.5 Added AGAIN-—’-RNV$T 3.19 13 99.5 2 the difference between the x values for the two models. If the diffe- rence in the X? values is statistically significant then the model with the better fit is accepted (Bagozzi 1980). Because neither version of the Prevailing View fit the data no further analysis of these two models was undertaken. In contrast, all three versions of the Alternative Model fit the data. Because there were no statistically significant differences between the three hierarchical models, the models should be compared on the basis of Simplicity or parsimony and, the most parsimonious model should be accepted. Therefore, the model depicted in Figure 4-8 was accepted. The findings supported the hypotheses that the franchisee's high level of financial dependence directly influenced the franchisee's performance and reinvestment intentions. However, it does not appear that reinvestment intentions lead to performance. The path coefficient 106 from RNVST to PERFM was only .03 and deleting this link from the model did not significantly affect the model's fit. But it would be premature to conclude that reinvestment is not a determinant of performance. That is, this result may reflect the need for a better measure of reinvest- ment. Here the franchisee's reinvestment intentions were measured rather than the actual reinvestment made in the business during some previous period(s). Future researchers may obtain a different result if the level of previous investment is measured because actual dollars are a more meaningful indicator of the amount at stake. As outlined in Chapter 2, it is the amount of money that the franchisee has at risk that determines the franchisee's performance. So, the relationship between previous investment and performance may be quite different from the relationship between reinvestment intentions and performance. One final note about reinvestment is required and that is to make it clear that a relative measure should be used rather than an absolute measure. A relative measure, like the percentage of profits reinvested enables the researcher to make meaningful comparisons whereas an absolute meaSure does not. As an illustration, it is not particularly useful to know that two fran- chisees invested one hundred thousand dollars in their franchises because one franchisee may have earned two or three times the profit earned by the other franchisee. Clearly, it is much more useful to know what amount both individuals reinvested as a percentage of profits. The hypothesis that performance is a determinant of attributions of responsibility was supported. Thus, Homans' (1974) hypothesis that attributions of responsibility determine performance was rejected. This finding provides further support for the theory that the franchisee acts out of self interest. Given a high level of financial dependence 107 franchisees have a vested interest in performing. Therefore, dependence is a determinant of performance and performance is a determinant of attributions of responsibility. That is, the amount of credit or blame assigned to the franchisor is determined by the franchisee's performance. The findings also supported the hypothesis that the franchisees' attri- butions of responsibility influenced satisfaction with the relationship. Specifically, attributions of responsibility directly influence satisfac- tion with the franchisor's role performance and the interorganizational climate as represented here by SATMD. Satisfaction with these dimensions of the relationship influence satisfaction with the franchisor's overall role performance (SATRL) and whether the franchisee would join this channel system again if they had it to do over (AGAIN). The findings also provided weak support for the hypothesis that the franchisees' satisfaction with their initial decision to join the channel system (AGAIN) directly influences reinvestment. In summary, the rationale for studying channel member satisfaction had been the untested belief that increased satisfactiOn led to improved performance and greater financial dependence. However, the findings reported here lead to the rejection of this theory. Indeed, the direc- tion of the causal links is just the opposite. Channel participants are economically motivated to form marketing channels and it is the high level of financial dependence on the relationship that drives performance and reinvestment. However, satisfaction may influence reinvestment. SUMMARY OF THE FINDINGS Chapter 4 presented the results of the data analyses. A limited information estimation procedure was used because previous channels research has been criticized for not reporting evidence of valid and 108 reliable measures and for using inapprOpriate statistical analysis techniques. Both problems were avoided through the use of this approach. A five-step process was used to evaluate the measurement model and it produced ten scales. The scales included two single-item measures of satisfaction (SATRL and AGAIN) and one single-item measure of attri- butions of responsibility (CREDT) and performance (PERFM), respectively. Three multi-item measures were identified. Both channel member depen- dence (DEP$$) and reinvestment intentions (RNVST) were represented with two-item measures. A four dimensional measure of channel member satis- faction (SATMD) also was identified. The evaluation of the two theoretical models showed that the Prevailing View model did not fit the data even when it was respecified. For this reason, the hypotheses associated with the Prevailing View were not supported by the data. Therefore, we can conclude that channel member satisfaction, a behavioral variable is not a determinant of three key economic variables: channel member performance, financial dependence and reinvestment. In contrast, the findings provided Support for the hypotheses associated with the Alternative Models. Three versions of the Alternative Model were tested and all three models fit the data. Therefore, the most parsimonious model was accepted and it appears in Figure 4-8. Chapter 5 presents a more detailed discussion of the findings and several important conclusions regarding the causes and consequences of channel member performance. CHAPTER 5 DISCUSSION AND CONCLUSIONS Chapter 5 begins with a review of the research objectives and an evaluation of the findings as they relate to the research questions presented in Chapter 1. Next, the major implications and conclusions associated with the research are discussed followed by a presentation of the dissertation's major contributions to the literature. The chapter ends with suggestions for future research. REVIEW OF THE RESEARCH OBJECTIVES As outlined in Chapter 1, the specific objectives of the research were: 1. To integrate the behavioral and economic approaches to marketing channels theory; 2. To conceptualize the interrelationship among channel member performance, satisfaction and dependence in a theoretically and managerially sound way; and, 3. To empirically test the conceptual scheme. The need to adopt an integrated approach to research involving the relationship between channel participants must be underscored. The members of a marketing channel are economically motivated and choose to work together because such an arrangement benefits both parties finanCially. For this reason, the study of sighs: the behavioral or economic dimension of marketing channels is inappropriate because such an 109 110 approach does not accurately reflect the operating characteristics of a marketing channel. This dissertation adopted an integrated approach. Notably, the research empirically tested the relationship between channel member satisfaction and three important economic variables: channel member performance; dependence; and, reinvestment. However, to be useful empirical research must be grounded in theory. Consequently, the second objective of the research was to develop a meaningful conceptual frame- work to guide the empirical work. The review of the literature revealed that channel member satisfac- tion had received considerable attention from channels researchers. The importance of satisfaction to channel managers was consistently cited based on the untested belief that satisfaction influenced performance (Mallen 1963, McVey 1960). But the directionality of this link was questionable particularly since the job satisfaction literature had produced evidence to the contrary (Wanous 1974). To satisfy the second research objective the authOr first con- structed a causal model to represent the prevailing view regarding the relationships between channel member satisfaction, performance, depen- dence, and reinvestment. The Prevailing View is shown in Figure 3-1 in Chapter 3. Next, the author constructed an alternative model shown in Figure 3-2 in Chapter 3. The primary difference between the two theories is that the Alternative Model depicts the economic variables as causal antecedents or determinants of the channel member's performance. The model is built on the assumption that most channel members have a substantial financial Stake in the relationship and would therefore be motivated to perform as a means of protecting their investment. In 111 contrast, the Prevailing View suggests that satisfaction leads to performance, dependence and reinvestment. Thus the third objective of the research was to empirically examine the two competing theories about the relationships between these variables. The specific research questions that guided the research were: 1. Does channel member satisfaction lead to performance? 2. Does channel member satisfaction directly influence the member's financial dependence? 3. Does channel member satisfaction lead to the reinvestment of profits? 4. Does financial dependence on a channel relationship directly influence the channel member's performance? In the following sections the findings of the study will be reviewed with respect to each of the research questions. 1. Does channel member satisfaction to to performance? The findings did not support the view that satisfaction leads to performance. Instead, the findings supported the Alternative Model which depicts satisfaction as a consequence of performance.' Furthermore, the relationship is not direct. An intervening variable, attributions of responsibility mediates the extent to which performance leads to satis- faction. 2. Does channel member satisfaction directly influence a Ehannel memberTs financialdependence? As reported in Chapter 4, the findings did not Suggest that satis- faction determined financial dependence. Satisfaction was depicted as a determinant of financial dependence in the Prevailing View path models but neither of these models fit the data. 112 3. Does channel member satisfaction lead to the reinvestment of profits? The findings of the study appeared to provide some support for this relationship albeit it weak. AS shown in Figure 4-9 in Chapter 4, the path coefficient from satisfaction (AGAIN) to reinvestment (RNVST) has a positive Sign as predicted and the overall fit of the model was excellent (p=99.5). However, two other versions of the Alternative Model which did not contain a link from AGAIN to RNV$T also fit the data (p>99.5). Because the difference in the fit of the three nested models was not statistically Significant the most parsimonious model was accepted (Bagozzi 1980). This model appears in Figure 4-8 and does not include a link from AGAIN to RNVST. 4. Does financial dependence on a channel relationship directly infTUence a ChanneT member's perfOrmance The findings supported the view that financial dependence directly influences performance. The relationship between dependence and perfor- mance is shown in Figures 4-7, 4-8 and 4-9 in Chapter 4. As noted above, each of the three versions of the Alternative Model fit the data. Each model depicted dependence (DEP$$) as a determinant of performance (PERFM). MAJOR CONCLUSIONS AND IMPLICATIONS Four major conclusions can be drawn on the basis of the research presented in this section. Each conclusion is discussed in turn. Understanding Whprarketing Channels Are Formed Managerially useful findings result from empirical studies that are based on sound theory. To be useful, theory must accurately represent the phenomenon being investigated. In this instance, the phenomenon of 113 interest was a marketing channel so the researcher must be knowledgeable about the reasons firms join together to form marketing channels in order to develop theory useful in prescribing guidelines for channel manage- ment. Notably, most previous channel research involving channel member satisfaction failed to accurately account for the channel participants' self interest. This study avoided this problem by recognizing that the following fundamental characteristics apply to marketing channels and. must be considered when developing channels theory: 1. Increased profit is the primary motive for joining a marketing channel; 2. The members adopt complementary roles so they are usually highly dependent upon one another for success; and, 3. Most channel participants make sizable investments in plant and equipment related to their role and thus channel participants are motivated to protect their investment. The Economic and Behavioral Dimensions Are InterrETated The findings provided empirical evidence of the interrelationship between the economic and behavioral dimensions of marketing channels. As noted above, the formation and structure of marketing channels generally depend upon economic criteria. The participants are economically motivated and typically highly dependent upon the relationship. So, individual channel members should be expected to behave in a manner that promotes their particular economic interests. Satisfaction Does Not Inflience Performance Satisfaction is not a determinant of channel member performance as suggested by the literature. But financial dependence does directly 114 influence performance and reinvestment. Thus, increasing intermediaries' satisfaction with the relationship does not appear to be a prime strategy for improving their performance. Instead, suppliers should pursue strategies that encourage the intermediaries' financial dependence. The key may be in the selection of intermediaries that have most of their assets invested in the business. An intermediary that has most of its net worth invested in the business has ample incentive to perform at a high level because the downside risk associated with not performing is so great. One possible strategy for encouraging financial dependence is to offer a more extensive array of support services to intermediaries that agree to invest a larger percentage of their net worth in the business. This strategy would discourage owner-investors who are not interested in participating in the day-to-day operations of the business such as doctors or lawyers. Instead, investors would be encouraged to become owner-operators due to their large financial stake in the business. A Casual Modeling Approach is Appropriate Bagozzi (1980) observed that: Perhaps no single concept is more pervasive and important in marketing than the notion of cause and effect. Marketing practitioners depend on it in their planning and implemen- tation of programs designed to obtain responses from consumers . . . Our propositions, theories, and methodologies are all fundamentally based on the concept of causality . . . Whether one wishes to understand the world or to change it, causality will invariably play a central role. Gaski (1984) criticized previous research because causal relation- ships were commonly suggested but almost never examined with a causal modeling analysis procedure. The causal relationships proposed in this study were presented in the form of a causal model and the structural 115 model was tested using the PATH routine in the PACKAGE program develOped by Hunter and Cohen (1969). A limited information estimation procedure should be employed because it solves the measurement model before investigating the theoretical model. Thus, the researcher can more easily separate measurement and theoretical issues. For example, evaluation of the measurement model used in this study revealed that none of the items used to measure satisfaction with financial returns were included in the multi-dimensional satisfaction measure. These items were not retained in the analysis because they did not correlate sufficiently with other items or with each other. Subsequently, the multi-dimensional satisfaction measure used to test the structural model did not represent the financial returns dimension. Because the researcher used a two-Step analysis procedure, the empirical content of the satisfaction variable was noted and considered when the findings were interpreted. MAJOR CONTRIBUTIONS It makes sense to judge the dissertation's contributions to the literature in relation to the major shortcomings associated with the channel research stream. Taking this perspective, one must conclude that the dissertation makes several major contributions. First, the researcher developed an alternative view of the relationships among channel member satisfaction, dependence, performance, reinvestment and attributions Of responsibility. Thus, the first major contribution of the dissertation was a more complete conceptual framework upon which the empirical test was based. The most significant criticism of previous work in this area was the lack of an integrated approach. The Alter- native View presented by the researcher addressed both the behavioral and 116 economic dimensions of the marketing channel. Thus, the Alternative View provides a basis for expanding channel theory in such a way that the theory more accurately reflects the interlocking nature of the economic and behavioral aspects of the channel. Second, the constructs of interest in the dissertation were explicitly defined and the reliability and validity of the measures were assessed using a rigorous five-step process. Additionally, the dimen- sionality of the constructs was evaluated. Two of the major shortcomings of previous channels research identified in Chapter 2 were inadequate construct definitions and insufficient evidence of the reliability and validity of the measures employed. Thus, the dissertation makes another important contribution by employing a more rigorous methodology that addressed these key issues. Third, the researcher employed a causal modeling approach to analyze the data because the research hypotheses logically took the form of a causal model. More importantly, several of the constructs were unobser- vable and only by using causal modeling can one account for measurement error. Therefore, a third major contribution of the dissertation is the use of an appropriate statistical analysis technique. Fourth, the dissertation presents managerially useful insights about marketing channel relationships and the lack of managerially useful insights has been consistently cited as a major shortcoming of channel research. Specifically, the research supports the view that the actions of channel participants are consistent with the participants' self interest and, it appears that the participants' primary objective is the preservation of the monies they have already invested in the business. 117 For this reason, strategies designed to increase channel member perfor- mance should not be founded on improving satisfaction. But managers should not neglect channel member satisfaction altogether because the findings provided some evidence that satisfaction does lead to reinvest- ment in the business and the periodic reinvestment of profits is essential to ensure the continued viability of every business. Another reason that channel member satisfaction should not be ignored is that continued dissatisfaction may lead to dissolution of the relationship given the availability of an attractive alternative (Homans 1974). The study did not produce empirical evidence to support the relationship between dissatisfaction and dissolution but nonetheless Homans' hypothesis makes sense. Therefore, prudent managers Should treat channel member satisfaction as an important variable in managing channel relationships. 7 However, the study provided evidence that financial dependence is a primary determinant of performance and reinvestment. Consequently, strategies designed to influence performance and reinvestment should encourage financial dependence. Clearly, the role of financial depen- dence should be addressed when management devises criteria for selecting intermediaries. In addition, the threshold cost of joining the channel should be reviewed. In summary the dissertation made several important contributions but many issues are not yet resolved. The final section of Chapter 5 presents some guidelines for future research. SUGGESTIONS FOR FUTURE RESEARCH The research represents a substantial contribution to marketing channel theory and practice because the study successfully integrated the F.— L a 118 economic and behavioral approaches to channel research. Nevertheless, many questions relevant to the management of channels remain unanswered. Therefore, the following section presents several suggestions for future research. 1. Certainly, the hypotheses tested in this research should be replicated in other channel systems. The data used in the research reported here was collected from a fast food franchise system. The research should be replicated in other firms within the fast food industry and also in other industries. It would be important to determine the extent to which the Alternative Model presented in Chapter 3 applies to various channels. 2. Past studies focused almost exclusively on behavioral variables such as power, conflict and satisfaction. But the findings reported here showed that the economic variables are the "drivers" that determine channel member behavior. Future research should heed Lambert's (1978) advice and pay greater attention to the economics of channel relationships as a means for improving channel management. The fact that channel participants are economically motivated and behave in accordance with their best interests must be accounted for in the development of theories about channel relationships. 3. Whenever possible, researchers should use objective measures rather than perceptual measures to avoid the problems of reliability and validity that plagued previous studies. If economic variables receive more attention in 119 1 future studies the use of objective measures is parti- cularly appropriate and may be developed from cost data obtained from study participants. Financial dependence has been represented by objective and perceptual measures but only the objective measures evidenced sufficient reliability and validity to be useful in testing theory. Furthermore, using objective measures of dependence the author Obtained a significant finding as did Etgar (1976) in his study of the insurance industry. Channel member satisfaction deserves further study because there are many dimensions of satisfaction that have not been adequately addressed. As an example, the multi- dimensional measure of satisfaction used in this study did not include the financial returns dimension. Future research Should attempt to devise a valid and reliable measure of satisfaction with financial returns because it is logical that this dimension may be instrumental in channel management. Indeed, the study provided weak support for the hypothesis that satisfaction leads to reinvestment in the business. Perhaps if the satisfaction measure had represented the franchisees' satisfaction with their financial returns a larger path coefficient would have been produced. For these reasons a valid and reliable measure of satisfaction with financial returns must be developed. Longitudinal studies must be undertaken to adequately investigate the role that many Situational variables play 120 in the management of channel relationships. For example, the franchisor in this study was purchased by a larger firm less than two years prior to the study. Subsequently the franchise system underwent many changes which may have influenced the research results. But the effect of these managerial changes cannot be explained on the basis of a cross sectional study. Likewise, many other variables relevant to the management of channel relationships may change over time. The important point is that the relationships between the participants in marketing channels are dynamic and long lasting. Consequently, longitudinal studies must be undertaken. Finally, the integration of the behavioral and economic approaches to marketing channel research holds great promise because an integrated focus more accurately reflects the way that channels operate. However, unless strict attention is given to methodological issues, more empirical research will be of limited value. Evidence of the reliability and validity of the measures used must be reported. Furthermore, a causal modeling approach to data analysis is recommended because it enables the researcher to simultaneously test causal inferences about several variables. Also, a causal modeling approach enables the researcher to account for measurement error. But as a first Step, Stronger theories must be advanced that accurately reflect the basic characteristics of a 121 marketing channel. Then, the constructs Should be expli- citly defined and guidelines for operationalizing the constructs should be suggested. APPENDIX A A STUDY OF FAST FOOD FRANCHISE SYSTEMS QUESTIONNAIRE MICHIGAN STATE UNIVERSITY A STUDY OF FAST FOOD FRANCHISE SYSTEMS PART IA: SATISFACTION WITH RETURNS. EQUITY AND MANAGEMENT STYLE INSTRUCTIONS Listed on the following pages are several statements about your franchise system. the franchisor and you. Please circle, on a scale of 1 to 7. the number that best expresses the extent to which you agree with each statement. ll you strongly meet a statement. circle number 1 (Straw Olaegreel. Reserve a rating of 7 (SW A'eel for those statements you wholeheartedly embrace. Example: am . 3w 5w Dlaevee Neutnl Agee 1 3 4 5 7 1 . My monetary rewards exceed my initial expectations ..................... 1 2 3 4 I a 5 7 2. Good ideas from franchisees often get passed along to franchise management ..... 1 2 m 4 5 5 7 5w Dlae'ee 1 2 1 The goals developed by the franchisor work to the mutual benefit of both the individual franchisee and the parent organization . . ..... . . . . 1 2 3 4 5 6 7 2. Gwen Our respective investments. the rewards are equrtably distributed between myself and the franchisor . ..... . ...... . . . . ...... . ....... 1 2 3 4 5 5 7 3. The market my storelsl are located in is larel intensely competitive . . ..... , . 1 2 3 4 5 5 7 4 My business philOSOphy and the franchisor‘s are very much alike . . ..... . 1 2 3 4 5 6 7 5. In relation to the franchisor's investment in time and money. the monetary rewards received by the franchisor are quite reasonable ................ 1 2 3 4 5 5 7 6. The overall quality of my franchisor's performance has been excellent during: Thelastthreevears . . . . .. ......................... 1 2 3 4 5 5 7 The laSt five years ........................................ . . . . 1 2 3 4 5 5 7 7. This franchise organization is highly interested in the welfare of its franchisees ..... 1 2 3 4 5 6 7 B. I am generally satisfied with my franchise Through it. I am able to meet all of my professional goals . ...... . .......................... 1 2 3 4 5 S 7 9. Franchisees have maior influence in the determination of policies and standards for this franchise organization . . . , . . . , . . ................... 1 2 3 4 5 6 7 As a group ....................................... 1 2 3 4 5 5 7 As individual franchisees ................................... . . . 1 2 3 4 5 5 7 10. Of the major fast food franchises. this svatem is the market leader: Locally, in the markets I operate in .................................. 1 2 3 4 5 5 7 Nationally .. . ........................................ 1 2 3 4 5 6 7 1 1. The rights and obligations of all parties concerned are clearly spelled out in the trenchiae contract . . .............. . . .................. 1 2 3 4 5 5 7 1 2. The corporate management of this franchise system is so far removed from my situation that their ideas often do more harm than good ............... 1 2 3 4 5 6 7 13. My ratio of rewards to effort is comparable to the franchisor's ratio of rewards to effort . . . ......................................... . . 1 2 3 4 5 6 7 14, The deciSion to switch to another franchisor or business venture would be based largely on the impact the switch would have on my lifestyle . . . ...... . . 1 2 3 4 5 5 7 1 5 My franchisor and l have very compatible business goals . . . 16. This franchise System requires more of my time and energy than I had originally expected ............... , . 20. 21 22. 23. 24. 25 26. 27 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. and promotional allowances ............ Franchisees are not allowed to provide input into the determination of standards The market my storels) is larel located in will decrease in population size . . . . in relation to my investment in time and money. the rewards received from my franchise are quite satisfactory ..... . . . ................... Given an opportunity to switch to another franChisor for another business venture) l wOuld switch My franchisor' a performance has been stable and predictable during- The last three years ...... . . . . . . The last five years .......... The deCision to switch to another franchisor lor another business venture) would be completely dependent on the monetary benefits of switching My actual monetary rewards compare favorably to my initial expectations . . The benefits would outweigh the costs of switching to another franchisor for another business venture) ...... . . The benefits received by switching to another franchisor or business venture would outweigh the personal effort required to switch ........... . This franchise system is very responsive to the competitive environment ..... Once they ve sold you the franchise. they lust forget all about you that is. until your fees are due again .......... My monetary rewards exceed my initial expectations . . . . . . . . . . . 1 am prowded sufficient guidelines and careful instructions on how to manage my franchise operations ............ . . . . . , . . . . . . . _ . My franchisor and l are very compatible on deCision strategy . Good ideas from franchisees often don‘t get passed along to franchise management ................................................ I am encouraged to use uniform procedures ................... There is a definite lack of support. coaching and feedback from the franchisor ...... My rewards compare favorably to my initial expectations ................... My franchisor's role performance has nOt met my initial expectations ........... My franchisor keeps me informed of all planned expansion projects and new stores in my market area ......................................... My franchisor's support and interface with our franchisee organization or its representatives is excellent ....................................... My franchisor explains how and why changes are occurring in the franchise system. and their effects on the local anwronment ........................ if I had it to do over. 1 would still become a franchisee of this system ............ PART II: EX'ECTATIONS W THE FUTURE i expect the franchisor to play a larger role in my success in the next five years ..... I expect the pizza segment of the fast food business to expand rapidly over the next five years ................................................ I expect this fast food system to become more competitive li.e.. gain market share) in the next five years: Locally, in the markets l operate in .................................. Nationally .................................................. NNNNN MN AGREEMENT Neutral 3 4 5 3 4 5 3 4 5 3 4 5 3 4 5 3 4 5 3 4 5 3 4 5 3 4 5 3 4 5 3 4 5 3 4 5 3 4 5 3 4 5 3 4 5 3 4 5 3 4 5 3 4 5 3 4 5 3 4 5 3 4 5 3 4 5 3 4 5 3 4 5 3 4 5 AGREEMENT m 3 4 5 3 4 5 3 4 5 3 4 5 3 4 5 (DU) 0! OQOOO MN] \JNJ 4. I expect to receive greater financial returns from this franchise in the next five years than I have in the past ........................... . . . ....... 1 2 3 4 5 6 7 5. lam extremely unsure regarding my expectations about the future . . . . . . . . . . . . _ 1 2 3 4 5 5 7 6. I expect consumer expenditures for fast food to increase by more than 10% annually over the next five years ................................ 1 2 3 4 5 5 7 7. I expect the Mexican segment of the fast food bushess to expand rapidly over the next five years .......................................... 1 2 3 4 5 5 7 5. I expect the hamburger segment of the fast food business to grow slightly over the next five years ......................................... 1 2 3 4 5 5 7 9. I expect the fish segment of the fast food business to expand rapidly over the next five years .......................................... 1 2 3 4 5 5 7 10. I expect inflation to remain under control over the next five years .............. 1 2 3 4 5 5 7 1 1 . I am very confident regarding my expectations about the future ............... 1 2 3 4 5 5 7 12. I do not expect any significant technological breakthroughs in the fast food business over the next five years ................................ 1 2 3 4 5 5 7 13. In the market my storels) is larel located in. I expect competition from company-owned stores to increase substantially over the next five years ......... 1 2 3 4 5 5 7 14. I expect the chicken segment of the fast food business to expand rapidly over the next five years ............................................. 1 2 3 4 5 5 7 1 5. I expect to receive excellent financial returns from my franchise in the future . . . . 1 2 3 4 5 5 7 PART II: IMPORTANCE OF SERVICES AND EVALUATION OF FRANCHISOR PERFORMANCE INSTRUCTIONS Listed on the following pages are various services often provided by a franchisor to its franchisees. This section involves three tasks. Each task will be explained separately. The first task is to rate the Importance of each of the services listed. Using the scale labelled IWORTANCE. "please circle the number which best expresses the Importance to you of receiving this service. If a service is nor important to you please circle number 1 (Not Important). A rating of 7 (Very Important) should be reserved for those some» that you believe are especially important. Please rate I of the aarvbes lated even though someservleasmeynotbeprovldedbywaranchlaor. The second not Is to evaluate your franchisor's ctwrent performenu as it relates to the services listed. Using the scale labelled PERCEIVED PERFORMANCE. please circle the number which best expresses your perception of the franchisor‘s current performance. If you perceive that the franchisor's performance is Poor circle a 1. Reserve a rating of 7 for Exceeant performance. If a service is not available from your franchisor please circle NA, NOT AVAILAILE. The third task is to indicate whether, in your opinion. the franchisor's performance is improving or declining. In the column labelled TREND. please Circle a e if performance is Improving. circle a - if performance is Deolnhg. and a a if performance remains Stable. IMPORTANCE PERCENEDPERFORMANCE m Example 1: f I .r .i‘ .4 3' . . 1 2 3 4 5 5 7 NA 1 2 3 4 5 I 7 - 1 Assistance in inventory control . . . . , , . . . . 1 2 3 ® 5 5 7 NA 1 2 3 4 5 Q 7 - . é Explanation: Moderately important to franchise 141 Performance received by franchisor is very good 15) Franchisor's perfOrmance is improv1ng in this area I +1 13. 14. 15. IS. 17. 18. 19. 20. 21. 22. 23. 24. 25. In general. franchisor support in the area of real estate and construction ......... . Length of promised order cycle Ileadl times on food items ......... Timely follow through after development of a business plan to assist with implementation ............. The ability of the packaging to help market the product ............... . . Competitiveness of the product line ......... Ability of the franchisor to meet prom1sed delivery dates on equipment ............ Your frenchiaor’ s assistance in building new stores ................. Provision of restaurant operations manual ..... Local media advertising ........... Your franchisor's ability to educate you regarding EEO guidelines ................ Your franchisor's assistance in terminating a lease ............................. Ability to minimize the time and effort required to prepare menu items for customers ........ The timing of new product rollouts .......... Your franchisor's assistance in repair and maintenance of equipment ............... Periodicuaeofretailpricedlecms.coupons. Your franchisor's assistance in negotiating tonne of a sale for a site or store ............ Advance notice on new products ........... Ability of a franchisor to meet promised delivery dates ........................ Your franchisor's assistance in kitchen design . . Length of promised order cycle lleadl timeaonpapersupplies ...... Product related support in general . . . Ability to minimize the time and effort needed to prepare raw products ....... Your ability to deviate from the company‘ a promotional program . . Your franchisor's asistance in negotiating a building contract: Putting it out to bid Supervisingconstruction . . . . . . . . . Development and introduction of new products .......... IMPORTANCE PERCEIVED PERFORMANCE 3‘ i f 43’ TREND 4 I J a: 4’ re' Iii 12345 7 111412345 7 .4 12345 7 NA12345 7 .4 12345 7 117112345 7 -4 12345 7 141112345 7 .4 12345 7 NA12345 7 -+ 12345 7 NA12345 7 .. 12345 7 NA12345 7 .. 12345 7 NA12345 7 =4 12345 7 "‘123‘5 7 ile- 12345 7 NA12345 7 =4 12345 7 NA12345 7 .4 12345 7 141112345 7 -+ 12345 7 NA12345 7 ... 12345 7 NA12345 7 -+ 12345 7 14412345 7 -4 12345 7 NA12345 7 .4 12345 7 M12345 7 .+ 12345 7 NA12345 7 -+ 12345 7 NA12345 7 .4 12345 7 NA12345 7 .4 12345 7 NA12345 7 =4 12345 7 141112345 7 =4 12345 7 NA12345 7 =o 12345 7 NA12345 7 =+ 12345 7 NA12345 7 -+ 12345 7 11412345 7 -¢ 12345 7 14412345 7 -+ 26, 27. 28. 29. 30. 31. 32. 33. 34. 35. 35. 37. 35. 39. 40. 41. 42. 43. 44. 45. 45. 47. Your franchisor's assistance in controlling labor cost ........................... The overall contribution or promotion dollars to sales volume .................. Training support in general ............... ailing from the franchisor: Accurate ........................... Timely ............................. Ability or iranchisor to introduce additions totheproductlinewhichcanbeaccom- modatad with existing stored. space ....... The introduction of fully tested and proven newproducts ........................ Your iranchisor's assistance in adding a drth ........................... Your franchisor's assistance in developing security policies ...................... A new product testing program that encom- pasaasallphasesoioperationsandmarketing .. Your franchisor's assistance in developing starting requirements ................... Your franchisor's assistance in labor scheduling .......................... Availability or in-stora promotional materials ior new products ...................... The extent to which the packaging keeps the product at correct serving temperature . . . . Accessibility of franchisor management overall ............................. Corporate management ................. Regional management .................. District Franchise representative ............. Ability to use the same raw ingredients in more than one finished product ........... Your franchisor's assistance in developing incentivelmotivation mama ............ Assistance in developing advertising and promotions for new store openings .......... Your franchisor's assistance in remodeling Reasonable storage requirements for promotional items ..................... Your ability to preview new advertising and promotions ....................... Extent to which new products tit the product line ......................... Provision of program ior crew training ........ ”ROMANCE ! ‘ 1 i .5: A; a; 147 1 2 3 4 5 5 7 NA 1 2 3 4 5 5 7 - n s 1 2 3 4 5 5 7 NA 1 2 3 4 5 5 7 - a + 1 2 3 4 5 5 7 NA 1 2 3 4 5 5 7 - n + 1 2 3 4 5 5 7 NA 1 2 3 4 5 5 7 - u + 1 2 3 4 5 5 7 NA 1 2 3 4 5 5 7 -- u 4 1 2 3 4 5 5 7 NA 1 2 3 4 5 5 7 - u + 1 2 3 4 5 5 7 NA 1 2 3 4 5 5 7 - a 4- 1 2 3 4 5 5 7 NA 1 2 3 4 5 5 7 - u 4- 1 2 3 4 5 5 7 NA 1 2 3 4 5 5 7 - . + 1 2 3 4 5 5 7 NA 1 2 3 4 5 5 7 - u + 1 2 3 4 5 5 7 NA 1 2 3 4 5 5 7 - u + 1 2 3 4 5 5 7 NA 1 2 3 4 5 5 7 - a + 1 2 3 4 5 5 7 NA 1 2 3 4 5 5 7 - a + 1 2 3 4 5 5 7 NA 1 2 3 4 5 5 7 - . + 1 2 3 4 5 5 7 NA 1 2 3 4 5 5 7 - - 4» 1 2 3 4 5 5 7 NA 1 2 3 4 5 5 7 - - + 1 2 3 4 5 5 7 NA 1 2 3 4 5 5 7 - - + 1 2 3 4 5 5 7 NA 1 2 3 4 5 5 7 - n + 1 2 3 4 5 5 7 NA 1 2 3 4 5 5 7 - - 4 1 2 3 4 5 5 7 NA 1 2 3 4 5 5 7 - s 4- 1 2 3 4 5 5 7 NA 1 2 3 4 5 5 7 - - + 1 2 3 4 5 5 7 NA 1 2 3 4 5 5 7 - - + 1 2 3 4 5 5 7 NA 1 2 3 4 5 5 7 - u 4- 1 2 3 4 5 5 7 NA 1 2 3 4 5 5 7 - - + 1 2 3 4 5 5 7 NA 1 2 3 4 5 5 7 - I 4 1 2 3 4 5 5 7 NA 1 2 3 4 5 5 7 - - 4- 1 2 3 4 5 5 7 NA 1 2 3 4 5 5 7 - u 4- 1 2 3 4 5 5 7 NA 1 2 3 4 5 5 7 - n 4 45. 49. 50. 51. 52. 53. 54. 57. 50. 51. 52. 53. 54. 55. 55. 57. 55. Ability of the franchisor to deliver 100 percent of the items ordered: Food ........... Equipment .................... . Paper supplies ....................... Provides a single point of contact for: Ordering ........................... Settling claims ............... The extent to which a quality assurance program estabiishes and enforces standards Length of promised order cycle llaadl times on equipment ........................ Ability of franchisor to meat promised delivery dates on food items ........... Use of merchandising programs for new products ........................... Your franchisor's assistance in developing quallty standards ...................... Theehelflifeofraw'ingredients ............ Oualityofaseistancerecaivedfromthe franchleerepreaentativa ................. Your franchisor's assistance in negotiating lease prices ......................... Theextenttowhichtheproductsrepreeent agoodvaluetoconeumars ............... Advancednoticereceivedfromyou frenchieorondeletionofproducta .......... Co-operative advertising hands provided by yourfranchisor ....................... Provision of guidelines for terminating emuoyees .......................... Advance notice of price changes ........... Langthofpromisedordercycle ileedl timesonprornotionalmaterials ............ Theeheiflifeoffinishedproducte ........... Equipment provided for use in preparing: Physical distribution support in general ....... National TV advertising ................. Your franchisor's assistance in developing a 3-5 year business plan ................. lnnovativenass of new products li.e.. consumers perceive thorn as unique and desirable) ...................... W'ONTANCE WED PERFORMANCE TREND [ i i i a: i 47 447 1 2 3 4 5 5 7 NA 1 2 3 4 5 5 7 4 1 2 3 4 5 5 7 NA 1 2 3 4 5 5 7 4 1 2 3 4 5 5 7 NA 1 2 3 4 5 5 7 4 1 2 3 4 5 5 7 NA 2 3 4 5 5 7 4. 1 2 3 4 5 5 7 NA 1 2 3 4 5 5 7 4 1 2 3 4 5 5 7 NA 1 2 3 4 5 5 7 4 1 2 3 4 5 5 7 NA 1 2 3 4 5 5 7 4 1 2 3 4 5 5 7 NA 1 2 3 4 5 5 7 4 1 2 3 4 5 5 7 NA 1 2 3 4 5 5 7 4 1 2 3 4 5 5 7 NA 1 2 3 4 5 5 7 4 1 2 3 4 5 5 7 NA 1 2 3 4 5 5 7 4 1 2 3 4 5 5 7 NA 1 2 3 4 5 5 7 4 1 2 3 4 5 5 7 NA 1 2 3 4 5 5 7 4 1 2 3 4 5 5 7 NA 1 2 3 4 5 5 7 4 1 2 3 4 5 5 7 NA 1 2 3 4 5 5 7 4 1 2 3 4 5 5 7 NA 1 2 3 4 5 5 7 4 1 2 3 4 5 5 7 NA 1 2 3' 4 5 5 7 4 1 2 3 4 5 5 7 NA 1 2 3 4 5 5 7 4 1 2 3 4 5 5 7 NA 1 2 3 4 5 5 7 4 1 2 3 4 5 7 NA 1 2 3 4 5 5 7 4 1 2 3 4 5 5 7 NA 1 2 3 4 5 5 7 4 1 2 5 7 NA 1 2 3 5 5 7 4 1 2 3 4 5 5 7 NA 1 2 3 4 5 5 7 4 1 2 3 4 5 5 7 NA 1 2 3 4 5 5 7 4 1 2 3 4 5 5 7 NA 1 2 3 4 5 5 7 4 1 2 3 4 5 5 7 NA 1 2 3 4 5 5 7 4 1 2 3 4 5 5 7 NA 1 2 3 4 5 5 7 4 1 2 3 4 5 5 7 NA 1 2 3 4 5 5 7 * 1 2 3 4 5 5 7 NA 1 2 3 4 5 5 7 4 MRTANCE i " r i i r a; r .1“ if PERCEWED PERFORHANCE TREND 70. 71. 72. 73. 74. 75. 75. 77. 78. 79. 80. 81. 82. 83. 84. 35. 87. 90. 91. 92. 93. 94. 2 3 4 8 8 7 NA 1 2 3 8 8 - 4 Your franchisor's assistance in record- keeping ....................... 1 2 3 4 8 8 7 NA 1 2 3 4 8 8 7 - 4 Your ability to influence the company's promotionalstrategyandprograms ........ 1 2 3 4 8 8 7 NA 1 2 3 4 8 8 7 - 4 Ability of franchisor to meet promised delivery datesonnewproductintroductions ......... 1 2 3 4 8 8 7 NA 1 2 3 4 8 8 7 . 4 Thequalityorgradeofrawingredients ....... 1 2 3 4 8 8 7 NA 1 2 3 4 8 8 7 - 4 Your franchisor's assistance in developing customaraervicest .............. 1234887 NA1234887 .4 Your franchisor's sssistance in developing hiringstandards ...................... 1234887 NA1234887 .4 Merchandising programs lpremiums. games.contestel ..................... 1234887 NA1234887 -+ Your franchisor's assistance in controlling variableespenses ..................... 1234887 NA1234887 -4 Ability of in-store promotional material tostimulateadditionaisaleeofspecificitems... 1 2 3 4 8 8 7 NA 1 2 3 4 8 8 7 - 4 Availability of the franchise consultant ....... 1 2 3 4 8 8 7 NA 1 2 3 8 8 7 . 4 Promotionalsupportingeneral ............. 1 2 3 4 8 8 7 NA 1 2 3 4 8 8 7 - 4 Your franchisor‘s abillty to educate you regardingunamploymentclaims ........... 1 2 3 4 8 8 7 NA 1 2 3 4 8 8 7 - 4 A new product testing program consistently appliedtonewproducts ................. 1234887 NA1234887 -+ Provisionofatrainingcenter .............. 1 2 3 4 8 8 7 NA 1 2 3 4 8 8 7 - 4- Number or approved suppliers for purchase of: Food .............................. 1 2 3 4 8 8 7 NA 1 2 3 4 8 8 7 - + PeperSuppliee ....................... 1234887 NA1234887 .4 Equipment .......................... 1234887 NA1234887 -+ in general. franchisor support in the area ofpersonnel ......................... 1 2 3 4 8 8 7 NA 1 2 3 4 8 8 7 - 4 Advertisingsupportfornewproducts ........ 1 2 3 4 8 8 7 NA 1 2 3 4 8 8 7 - 4 New productpricingrecommendations fromthefrenchieor .................... 1 2 3 4 8 8 7 NA 1 2 3 4 8 8 7 . 4 Your franchisor’s sssistanceindeveloping aeleryguidelines ...................... 1 2 3 4 8 8 7 NA 1 2 3 4 8 8 7 . 4 Your franchisor's assistance in inventory control ............................ 1 2 3 4 8 8 7 NA 1 2 3 4 8 8 7 - 4 General pricing recommendations from thefrenchisor ........................ 1234887 NA1234887 n4 Provisionofrranchiseeoperationsupdates 1 2 3 4 8 8 7 NA 1 2 3 4 8 8 7 - 4 From franchisor. convenient and clearly communicated handling procedures for your complaints and claims ............... 1 2 3 4 8 8 7 NA 1 2 3 4 8 8 7 - 4. Your franchisor's assistance in controlling costofsales ........................ 1 2 3 4 8 8 7 NA 1 2 3 4 8 7 a + Competitivenesaofretailpricas ............ 1 2 3 4 8 8 7 NA 1 2 3 4 8 8 7 - '4 95. 96. 97. 98. 100. menace m M! TREND f I a: - s: f i 0?: NA1234887 - Length of promised order cycle llssdl timesonmerchandisingmeterials .......... 1 2 3 4 8 8 7 NA 1 2 3 4 8 8 7 - - 4 Assistsnceindevslopingadvertisingand promotionalprogramsforspsciallocalevents .. 1 2 3 4 5 8 7 NA 1 2 3 4 8 8 7 - - 4 Provisionofexplanationsforpricechangss.... 1234587 NA1234887 - - 4 Your franchisor's assistance in evaluating: . Existingstores ....................... 1234887 Na1234887 --4 Newetorssites ....................... 1234887 NA1234887 --4 Abilitytogetthsloweetpricsongoode purchassdthroughthefranchisor: Food .............................. 1234587 NA1234887 =4 PaperSupplies ...................... 1234887 NA1234587 -.4 Equipment .......................... 1234587 NA1234587 --4 Your franchisor's assistance in obtaining buildingpermits ...................... 1234887 NA1234567 -=4 rmms‘ruronosormm INSTRUCTIONS: Please provide the following information with respect to the levels of performance that you require or expect from your franchisor. 1. 10. Please indicate the new» number of totd ths you require from your franchisor representative during a typical year. _Numbsr of visits Please indicate the m number of franchisee business W you expect the franchisor to offer during a typical year. _Numbsr of meetings What is the m and new number of new products you expect the franchisor to introduce during a typical year? _Minimum nurnbsr _Meirimum nurnbsr - What is the m return you expect a new product to generate? Return equals profit before tax divided by dollars invested. ___% Minimum return What is the m acceptable bless-even (cost rscovsryl timefreme for you on a typical new product? _Maximum break-even timsframe expressed in months will smhmsnt chsnga _Maximum break-even timeframe expressed in months without m ohsnge Howmuchadvanoenetlosdoyoursouireonatypicalnewproaict? _Deys I ___Weelrs withsmhmsnt orange _Deys/_Wesltswllhoinmohsnge Howmuchadvanosnstloedoyoursquirsonenewadvanhhgcsmpslm7. _Deys/_Weslts Howmuchedmensdosdoyounoifinonanewprmidormigm? _Deys/__Weelts Howmuchedvanoenodosdoyoureouireonsprlosdisnga? _DeysI__Weelts Under normal conditions. what is the average order cycle (advance notiosl you must give your franchisorldistributor when ordering lfrom day youordsrimtldayyoureceivethsordsrl: Food _Deys Pse- Iirsslss _Deys W _Dey- Promotlonsl W Days 11. When ordering from your franchisor/distributor. what is the range in order cycle time you find acceptable (maxim-n delay in delvery you find tolerable). Food _Deys Delay Paper Supplies _Deys Delay 8m _Deys Delay Promotional Materials/MM _Deys Delay What is the mhimum order fl rate you expectldssh'e from your franchisor/distributor (96 of order included in initial shipment): Food _.95 Paper Supplss _% W —* Promotional Materiels/Merchandse _% What order fl rats do you patently receive from your franchisor/distributor on: Food —% Pepsr 8m _95 Equipment _% Promotional Materials/Merchants _% How much advance notice from vaur franchisor/distributor would you prefer prlor to the arrival of: Food _Deys Notice Paper Staples _Deys Notice Equipment _Deys Notice Promotional MoterlslslMerchsndss ___Days Notice What is the minimum acceptable return lprofit before tax divided by dollars invested) for you on a promotional campaign or merchandising program? _% What is the m number of approved staplers you find acceptable for: Food _lsupolim maul“ _rSuoolisrs Em _rsiippliars PromotionalMssarlalslMsrchsndes __aSupoliors PART IV: EVALUATION OF GENERAL PERFORMANCE RELATED FACTORS AND FUTURE PLANO 1 . My posmonltltle is: Sole OwnerIFranchises (Go to 2) One of Multiple Owners (Go to 2) Business Manager lGo to 1al Store Manager (Go to lei _____Other (Go to lei Please answer all questions as best you can regarding any questions that pertain to the owner/franchisee as if you were the owner 7 franchisee lf y0u are unable to answer the question. simply leave it blank on both this section (Part IV) and the following section (Part V). Please mark a 00m! anywhere on the line below that best expresses your level of satisfaction with your franchisor's overall performance. If vOu are extremely dissatisfied with your tranchisor's performance. a merit should be placed very near the left end of the line llabelled Poorl. If you are exceptionally pleased with your franchisor‘s performance. a marl: should be placed very near the right end of the line llabelled Excellent). A midpomt has been placed on the line to correspond with a 8atlsfectory performance level. OVERALI. FRANClNSOR PERPORMANCE Poor 8atbfsotory Elm L l J I I 1 Please distribute l 00 paints among the three factors listed below. Assign the number of points which best expresses your evaluation of the importance each of these factors played in the achievement of your financial results. A. Your performance as an ownerrmaneger B The franchisor's performance C Other situational factors. such as economic conditions TOTAL I 100 points As a percentage of 1984 sales. what level of sales do you expect to achieve for 1988-87? (1994 - 100%) A. Expected 1985sales (as a 94 of 1984 sales) 8. Expected 1988 sales (as e 95 of 1984 sales) C. Expected 1987 sales (as a 96 of 1984 sales) Please distribute 100 points among the three factors listed below. Assign the number of points which best expresses your evaluation of the importance each of these factors wl have in the achievement of your 1985-1987 finanCial results. A. Your performance 9. The franchisor's performance C. Other situational factors. such as economm conditions TOTAL '- 100 paints Given an opportunity. would you expand the number of stores you currently operate. (Please check (w) one). A. ___Yes. I would expand 8. ___No. I would not expand Please indicate by checking (.4) one answer whether or not you currently have an expansion proiect underway. A. __Yes. I am expanding my existing storels). (Please indicate by checking (a) the amine items that best express the focus of your expansion protect): 1)___Adding more seats in a dining area 2)__Adding e drive-thru 3)___Adding more seating outdoors 4)__Adding more parking spaces 8)__Bemodeling the kitchen 8)__Point-of-sale cash registers (computerized) 7)_Othsr lplsass describe WM 8. _Yes. I am adding a new store or stores. C. ___No. I will not expand at the present time. What is your major goal for the next year? (Check (I) one of the following) increase franchise profitability ._Devalop other business interests ___Spend less time at work _Other (Please specify) Expansion of current business The risk to the franchises in this franchise operation is: Comparable to other maior franchises ___Less than other mayor franchises More than other maior franchises ___Don't know Please indicate how much you have reinvested or plan to reinvest in your franchise as a percent of profits before tax. A. 96 of profits reinvested during 1984 8. ___‘lb of profits you plan to reinvest during 1988 C. __‘lo of profits you plan to reinvest during 1988 D. 96 of profits you plan to reinvest during 1987 Please distribute 100 percent among the following nine factors. Assign the percentage in the left column based on your evaluation of the franchisor's current efforts in each of these areas. In the dirt column. assign the percentage based on how you think your franchisor’s efforts should be distributed. (Example: Currently, 20% of the franchisor‘s effort is spent on product related activmas. but you prefer that 30% of the franchisor's effort to be spent in this area. Mark 20% in the column labelled Durant Distribution of Effort and 30% under the column labelled Preferred Dhtrblrtlon of Effort. . Current Prefer!“ Dbtrhutlon Dbtrltudon of Effort of Effort A. Product ..................... 8. Pricmg ...................... C. Promotion . . , ............... D. Distribution/Customer Service . . . . . ____ E. Operations .................. F. Training ............. . , 6. Real Estate 8 Construction ........ H. Personnel ................. TOTAL - 100 percent 100 percent Please indicate by checking (a) the appropriate items which of the following you have implemented. A .. _Drive-thru B. ___—New menu boards PART V: USER CHARACTERIITES AND MWTlON DATA 11. 15. 16. How many years have you been a franchisee with this franchise system7 ___Years What are the first two numbers of your zip code? Approximately what percentage of your total annual income comes from your franchise? _% of total annual income Approximately what percentage of your total net worth does your franchise contribute? __‘lb of total net worth Please indicate your approximate liquid net worth (cash. marketable securities) as a percent of your total net worth. 96 Liquid net worth (as percent of total net worth) Please indicate the fatal number of managers you employ and their average years of managerial experience in the restaurant business. Number of management employees AverlOO Years of experience What percentage of your restaurant managers have completed training at the franchisor‘s training center? ___‘lti managers trained What is the average age of your restaurant managers? ...—Under 28 __30-40 —25-30 __Over 40 Please indicate number of stores you operate. Total number of stores _Number of stores in suburban locations Number of stores in rural locations _Number of stores in urban locations (population less than 50.000) (inner city) Please indicate which franchise system you belong to by checking (I) one of the following: A. __Taco Bell F. __Wendy's B. __Chi Chi's 6. __Kentucky Fried Chicken C __8urger King H. __Arby's 0 ___Long John Silver's l. __Herdee's E. Pizza Hut J. _Other (please specify) Please indicate how many years of experience you have in the fast food business‘ . ___ Years Were you employed by this chain before becoming a franchisee? ___Yes ___.Number of years employed No Are y0u in a test market for new products for your franchise system? Yes No How many days per week do you personally work at this franchise? 1-2 days ___8 days _3-4 days _7 days ___5 days How many hours do you personally devote to your franchise in a week? __Hours Please indicate the number of stores in your market area that are (that is the area within which you compete for customers); ___Company operated Franchise units Have you owned a business prior to the purchase of this franchise. or were yOu gainfully employed but not self-employed? Yes. I owned my business ___Yes. I was gainfully employed ___No 18a. 19 20. 21 22, 23. Do you currently own another business. or are you gainfully employed in another occupation? Yes. I own another business (Please answer 18a) Yes. I am gainfully employed (Please answer 18s) ___.No (Go to 19) ll yes. please indicate how many hours per week you typically devote to this business or gob. _ Hours Please indicate your level of education High school graduate __Some graduate education Some college ___.Graduate degree ___—College degree _Other (please specify) Please indicate the year of your birth. Please indicate your mental status. ___..Single ___—Married with children living at home Divorced ___Married with grown children Separated __Married with no children Do you operate your own distribution center? _Yes _No Please indicate by checking (w) the appropriate column where you get each of the following supplies. If you obtain a given item leg. food) from more than one source, please indicate what percentage of your supplies is provided by each source: National (may be a SuppIar separate (may be another Local (may also supply subsidiary) francnisee) Sinpier the franchisor) Food ....................... 96 % 96 “it: Paper Supplies ............ 96 95 oh qb Equipment ........... 96 "6 3ti 0’0 Promotional Materials/Merchandise 96 96 96 qt) Thank you for ypur particrpation and cooperation in completing this survey. Your time and effort are sincerely appreCiated. Please return the questionnaire in the envelope prowded or mail to; DOUGLAS M. LAMBERT. Ph.D. 0 M. CHRISTINE LEWIS. M.8.A. Department of Marketing and Transportation Graduate School of Business Administration Michigan State University East Lansing, Michigan 48824-1 121 BIBLIOGRAPHY BIBLIOGRAPHY Alderson, Hroe (1954), “Factors Governing the Development of Marketing Channels,“ in R.M. Clewett, (ed.), Marketing Channels for Manufac- tured Products (Homevood, Ill.: Richard D. Irwin, Inc.), 5434. (1957), Marketing Behavior and Executive Action (Homewood, Ill.: Richard D. Iriln), Chapter XI; Armstrong, S.J. and T.S. Overton (1977), “Estimating Nonresponse Bias in. Mail Surveys," Journal of Marketigg Research, (August), 396-402. Anderson, J.C. and 0.8. Gerbing (1982) "Some Methods for Respecifying Measurement Models to Obtain Unidlmensional Construct Measurement," Journal of Marketing Research, 19 (November), 453-60. Assael, Henry (1968), “The Political Role of Trade Associations in Distributive Conflict Resolution,“ Journal of Marketing, 32 (April), 21-28. (1969), "Constructive Role of Interorganizational Conflict,“ Administrative Science Quarterly, 14 (December), 573-582. Bagozzi, Richard P. (1980), Causal Models in Marketing, New York: John Riley & Sons, Inc. Beier, Frederick J. and Louis N. Stern (1969), “Power in the Channel of Distribution," in Distribution Channels: Behavioral Dimensions, Louis N. Stern, ed., Boston: Houghton Mifflin, 924116; Blau, Peter (1964), Exchange and Power in Social Life, New York: Riley. Bowersox, Donald, leby Cooper, Douglas Lambert and Donald Taylor (1980), Management in Marketing Channels, McGraw-Hill Book Company. Brown, James R. (1977), "Toward Improved Measures of Distribution Channel Conflict," in Contemporary Marketing,Thought, Barnett A. Greenberg and Danny M. Bellenger, eds., Chicago: American Marketing, 385-389. and Gary L. Frazier (1978), "The Application of Channel Power: Its Effects and Connotations," in Research Frontiers in Marketing: Dialogues and Directions, Subhash C. Jain, ed., Chicago: AmeriCan Marketing, 266-270. Burt, R.S. (1973), "Interpretational Confounding of Unobserved Variables in Structural Equation Models,“ Sociological Methods and Research, 5 135 Cadotte, Ernest R. and Louis N. Stern (1979), "A Process Model of Inter- organizational Relations in Marketing Channels,“ Research in Marketing, 2, 127-158. Campbell, D.T. and o.w. Fiske (1959), "Convergent and Discriminant Validation by the Multitrait-Multimethod Matrix," Psychological Bulletin, 56, 81-105. Churchill, Gilbert, Jr. (1979), "A Paradigm for Developing Better Measures of Marketing Constructs,“ Journal of Marketing Research, 16 (February), 64-73. Dickson, Peter R. (1983), "Distributor Portfolio Analysis and Channel Dependence Matrix: New Techniques for Understanding and Managing the Channel," Journal of Marketing, 47 (Summer), 35-44. Dwyer, F. Robert (1980), "Channel-Member Satisfaction: Laboratory Insights,“ Journal of Retailing, 56 (Summer), 45-65. El-Ansary, Adel I. (1975), " Determinants of Power- Dependence in the Distribution Channel,“ Journal of Retailing, 51 (Summer), 59-74, 94. and Louis N. Stern (1972), "Power Measurement in the Distribu- tion Channel,” Journal of Marketing Research, 9 (February), 47-52. Emerson, Richard M. (1962), “Power-Dependence Relations,“ American Socio- logical Review, 27 (February), 31-41. Etgar, Michael (1976a), "The Effects of Administrative Control on Effici- ency of Vertical Marketing Systems," Journal of Marketing Research, 13 (February), 12-24. (1976b), “Channel Domination and Countervailing Power in Distri- butive Channels,” Journal of Marketinngesearch, 13 (August), 254-262. (1977), “Channel Environment and Channel Leadership," Journal of Marketigngesearch, 14 (February), 69-76. (1978a), “Intrachannel Conflict and Use of Power," Journal of Marketing Research, 15 (May), 273-274. (1978b), "Selection of an Effective Channel Control Mix,“ Journal of Marketing, 42 (July), 53-58. (1979), "Sources and Types of Intrachannel Conflict," Journal of Retailing, 55 (Spring), 61-78. Fornell, Claus and David F. Larcker (1981), "Evaluating Structural Equa- tion Models with Unobservable Variables and Measurement Error," Journal of MarketinggResearch, 18 (February), 39-50. 136 Frazier, Gary L. (1983a), “0n the Measurement of Interfirm Power in Channels of Distribution,“ Journal of Marketing Research, 45 (May), 58-66. (1983b), "Interorganizational Exchange Behavior in Marketing Channels: A Broadened Perspective," Journal of Marketing, 47 (Fall), 68-78. and John 0. Summers (1984), "Interfirm Influence Strategies and Their Application Within Distribution Channels,“ Journal of Marketing, 48 (Summer), 43-55. French, John R.P. and Bertram Raven (1959), “The Bases of Social Power," in Studies in Social Power, Dorwin Cartwright, ed., Ann Arbor: University of Michigan Press. Gaski, John (1984), “The Theory of Power and Conflict in Channels of Distribution,“ Journal of Marketing, 48 (Summer), 9-29. and John Nevin (1985), “The Differential Effects of Exercised and Unexercised Power Sources in a Marketing Channel,” Journal of Marketing Research, 22 (May), 130-142. (1986), “Interrelations Among a Channel Entity's Power Sources: Impact of Reward and Coercion on Expert, Referent, and Legitimate Power Sources," Journal of Marketinngesearch, 23, 62-77. Guiltinan, Joseph P., Ismail B. Rejab, and William C. Rodgers (1980), "Factors Influencing Coordination in a Franchise Channel," Journal of Retailing, 56 (Fall), 41-58. Homans, George (1974), Social Behavior: Its Elementary Forms, New York: Harcourt, Brace, JovanoviCh, Inc., revised editiOn. Hunt, Shelby D. and John R. Nevin (1974), "Power in a Channel of Distri- bution: Sources and Consequences," Journal of Marketing Research, 11 (May), 186-93. Hunter, J.E. (1973), "Methods of Reordering the Correlation Matrix to Facilitate Visual Inspection and Preliminary Cluster Analysis," Journal of Educational Measurement, 10 (1), 51-61. and S.H. Cohen (1969), "PACKAGE: A System of Computer Routines for the Analysis of Correlational Data.“ Educational and Psycho- logical Measurement, 29, 697-700. and D.H. Gerbing (1982), "Unidimensional Measurement, Second Order Factor Analysis and Causal Models,“ in Research in Organizg; tional Behavior, B.M. Staw and L.L. Cummings, eds., Greenwich, CT: JAI Press, voTume IV. 137 , , S.H. Cohen and T.S. Nicol (1980), "PACKAGE 1980: A System of FORTRAN Routines for the Analysis of Correlational Data,“ prepared with the support of Academic Computing Services, Baylor University, Waco, Texas. Joreskog, K.G. and D. Sorbom (1981), LISREL V: Analysis of Linear Structural Relationships by Maximum Likelihood and Least Squares Methods. Chicago: National Educational Resources. Kelly, J. Steven and J. Irwin Peters (1977), “Vertical Conflict: A Comparative Analysis of Franchisees and Distributors," in Contemporary Marketing Thought, Barnett A. Greenberg and Danny N. Bellenger, ed., Chicago: AmeriCan Marketing, 380-384. Kenny, David (1979), Correlation and Causality, New York: Wiley. Lambert, Douglas M. (1978), The Distribution Channels Decision, New York: National Association of Accountants; and Hamilton, Ohtario: The Society of Management Accountants of Canada. Lusch, Robert F. (1976a), "Sources of Power: Their Impact on Intrachannel Conflict,“ Journal of Marketing Research, 13 (November), 382-390. (1976b), "Channel Conflict: Its Impact on Retailer Operating Performance,“ Journal of Retailing, 51 (Summer), 3-12, 89-90. (1977), “Franchisee Satisfaction: Causes and Consequences,“ International Journal of Physical Distribution, 7 (February), 128-140. .________ (1973). “Intrachannel Conflict and Use of Power: A Reply," Journal of Marketinngesearch, 15 (May), 275-276. and James R. Brown (1982), “A Modified Model of Power in the Marketing Channel,“ Journal of Marketing Research, 19 (August), 312-323. Mallen, Bruce (1963), “A Theory of Retailer-Supplier Conflict, Control, and Cooperation," Journal of Retailing, 39 (Summer), 24-32, 51. (1973), “Functional Spin-off: A Key to Anticipating Changes in Distribution Structure,“ Journal of Marketing, 37 (July), 18-25. McVey, P. (1960), "Are Channels of Distribution What the Textbooks Say?“, Journal of Marketing, 28 (January), 61-64. Michie, Donald A. (1978), "Managerial Tactics: An Alternative Explanation of Warranty Satisfaction in a Channel of Distribution," in Research Frontiers in Marketing: Dialogues and Directions, Subhash C. Jain, ed., Chicago: American Marketing, 260-265. Nie, Norman H., et al. (1975), SPSS: Statistical Package for the Social Sciences, 2nd ed., New York: McGraw-Hill Book‘Cb. 138 Nunnally, J.C. (1978), Psychometric Theory, 2nd ed., New York: McGraw- Hill Book Co. Pearson, Michael M. (1973), “The Conflict-Performance Assumption,“ Journal of Purchasing, 9 (Februany), 57-69. Phillips, Lynn W. (1981), "Assessing Measurement Error in Key Informant Reports: A Methodological Note on Organizational Analysis in Marketing,“ Journal of Marketing Research, 18 (November), 395-415. Pondy, Louis R. (1967), "Organizational Conflict: Concepts and Models," Administrative Science Quarterly, 12 (September), 296-320. Raven, Bertram H. and Arie W. Kruglanski (1970), “Conflict and Power,“ in The Structure of Conflict, Paul Swingle, ed., New York: Academic Press, 69-109. Reve, Torger and Louis W. Stern (1979), "Interorganizational Relations in Marketing Channels," Academy of Management Review, 4 (no. 3), 405-416. Robicheaux, Robert and Adel I. El-Ansary (1975), "A General Model for Understanding Channel Member Behavior,“ Journal of Retailing, 52, 4 (Winter), 13-30. Roering, Kenneth J. (1977), "Bargaining in Distribution Channels," Journal of Business Research, 5 (March), 15-26. Rosenberg, Larry J. (1974), "A New Approach to Distribution Conflict Management,“ Business Horizons, 17 (October), 67-74. and Louis W. Stern (1971), "Conflict Measurement in the Distribution Channel," Journal of Marketing Research, 8 (November), 437-442. Rosenbloom, Bert (1973), "Conflict and Channel Efficiency: Some Concep- tual Models for the Decision Maker,“ Journal of Marketing, 37 (July), 26-30. Ross, Robert and Robert F. Lusch (1983), "Logistical and Promotional Services and Wholesaler-Broker Satisfaction," International Journal of Physical Distribution and Materials Management 13, 1, 56-64. Ruekert, Robert and Gilbert Churchill (1984), "Reliability and Validity of Alternative Measures of Channel Member Satisfaction," Journal of Marketing Research, 21 (May), 226-234. Sheridan, J.E. and J. Slocum, Jr. (1974), "The Direction of the Causal Relationship Between Job Satisfaction and Work Performance," Organizational Behavior and Human Performance, 14 (October), 159-72. Sheth, Jagdish and David Gardner (1982), "History of Marketing Thought: An Update," in Marketing Theory: Philosophy of Science Perspectives, R. Bush and S. Hunt, eds., Chicago: AmericanTMarketing Association, 52-58. 139 Stern, Louis W. and Adel I. El-Ansary (1977), Marketing Channels, Englewood Cliffs, NJ: Prentice-Hall, Inc. and Ronald M. Gorman (1969), “Conflict in Distribution Channels: An Exploration,“ in Distribution Channels: Behavioral Dimensions, Louis W. Stern, ed., Epston: HOughton Mifflin, 156-175. , Brian Sternthal, and C. Samuel Craig (1973), "Managing Conflict in Distribution Channels: A Laboratory Study," Journal of Marketing Research, 10 (May), 169-179. and Torger Reve (1980), "Distribution Channels as Political Economies. A Framework for Comparative Analysis," Journal of Marketing, Vol. 44 (Summer), 52-64. Vroom, Victor (1964), Work and Motivation, New York: Wiley. Walker, Jr., O.C. (1972), "The Effects of Learning on Bargaining Behavior,“ in 1971 Combined Proceedings, F.C. Allvine, ed., Chicago: American Marketing, 194-199. Walker, Orville, Gilbert Churchill, Jr., and Neil Ford (1977), "Motiva- tion and Performance in Industrial Selling: Present Knowledge and Needed Research," Journal of Marketing Research, 15 (May), 156-168. Wanous, J.P. (1974), “A Causal-Correlational Analysis of the Job Satisfaction and Performance Relationship," Journal of Applied Psychology, 59, 139-44. Wilkinson, I.F. (1974), "Researching the Distribution Channels for Consumer and Industrial Goods: The Power Dimension,“ Journal of the Market Research Society, 16, 1, 12-32. _________ (1979), “Power and Satisfaction in Channels of Distribution,“ Journal of Retailing, 55 (Summer), 79-94. (1981), “Power, Conflict, and Satisfaction in Distribution Channels--An Empirical Study,“ International Journal of Physical Distribution and Materials Management, 11 (no. 7), 20-30. "ITHIIII'IIIIIIIIIIWIII