به سوی درک بهتر از ویژگی های مشارکت : تجزیه و تحلیل اکتشافی از طبقه بندی نوع رابطه

In recent years the creation of partnerships has been espoused by academics and practitioners but despite extensive writing in the area of characterizing partnerships the nature and attributes of partnerships are still only poorly understood. This is due to a lack of empirical testing of the factors that distinguish partnerships from other relational forms and a limited research focus on characteristics that distinguish between polar relationship types. Using a framework developed from the political economy literature this exploratory study draws on an empirical base of 155 supplier questionnaires and uses multiple discriminant analysis to identify attributes of buyer–supplier relationships that most effectively discriminate between relationships classified by suppliers as having limited coordination, being highly coordinated or best described as a partnership. The results indicate that 8 of the 10 attributes of buyer–supplier relationships included in the framework differ significantly across relationship type.

Traditionally, inter-organizational linkages between firms have been arms-length and adversarial, with individual firms seeking to achieve cost reductions or profit improvements at the expense of their buyers and suppliers. However, the transfer of costs up and down the supply chain does not make firms any more competitive as ultimately all costs will make their way back to the final marketplace (Christopher, 1998 and Lamming, 1993). Increasingly, successful firms are recognising the limitations of the adversarial model of exchange and instead engaging in co-operative long term partnerships that help to improve the efficiency of the supply chain as a whole, for the mutual benefit of all parties involved (Christopher, 1998, Rinehart et al., 2004 and Spekman et al., 1998). Partnerships have been defined as ongoing relationships between two firms involving a commitment over an extended time period and a mutual sharing of information and the risks and rewards of the relationship (Ellram and Hendrick, 1995). Although academics and practitioners have advocated the adoption and creation of partnerships, and espoused the benefits that they can bring, little empirical work has been published on the attributes of such relationships (Ellram and Hendrick, 1995). Indeed, despite extensive writing in the area of characterizing partnerships, researchers state that the concept of partnerships and their exact nature is still only poorly understood (Lemke et al., 2003 and Siguaw et al., 2003). Instead there is an implicit assumption that we all share the same understanding of what constitutes a partnership or relational exchange (Zolkiewski, 2004). Ellram and Hendrick (1995) suggest that this lack of understanding is due in part to the fact that there has been limited empirical testing of the factors that distinguish partnerships from other relational forms. Previous research has also been criticized for its failure to provide a coherent picture of relationships, as the use of different constructs and definitions makes comparison of results difficult (Olsen and Ellram, 1997). Therefore, additional research is needed to establish if the partnering characteristics identified in the existing literature truly distinguish between partnering and non-partnering relationships, and to better conceptualize what relational exchange and partnerships actually entail (Ellram and Hendrick, 1995, Frazier, 1999, Lemke et al., 2003 and Moberg and Speh, 2003). A poor understanding of how partnerships differ from other types of co-operative relationships results from a limited research focus on characteristics that distinguish between polar relationship types. Specifically, the focus of most research has been to compare partnerships and relational exchange to discrete, transactional or adversarial arms-length dealings (i.e. Dwyer et al., 1987, Frazier, 1999, Lambert et al., 1996, Lusch and Brown, 1996, Moberg and Speh, 2003 and Mohr and Nevin, 1990). However, it is widely acknowledged that most firms have already achieved cooperation and coordination with key segments of their suppliers and customers (Spekman et al., 1998). Therefore rather than a dichotomous focus on discrete market transactions at one end of a relationship continuum and partnerships at the other, it is important that future research focuses on identifying the critical differences between partnerships and other types of cooperative relationships, as researchers state that these differences have been largely ignored (Golicic and Mentzer, 2005, Hausman, 2001 and Laing and Lian, 2005). Therefore, the aim of this paper is to address this gap in the literature by comparing partnerships to relationships where some coordination of activities is specified to exist, rather than comparing partnerships to transactional relationships based on infrequent or ‘one off’ market exchanges. As firms' beliefs about others identify their behavioral predisposition towards them (Donaldson and O'Toole, 2000) the premise of the study is that firms that perceive a relationship as a partnership will be motivated to engage in different behaviors and actions from those that don't. Specifically, this study compares relationships perceived by suppliers to be a partnership, to those perceived by suppliers as being highly coordinated or as having limited coordination of activities (Fig. 1). These categories of relationship type are consistent with descriptions of buyer–supplier relationships and interaction intensity identified in previous research (i.e. Spekman et al., 1998 and Webster, 1992). In line with previous research, inter-organizational relationships are positioned in-between continuum anchors of market transactions and vertical integration (Golicic and Mentzer, 2005, Nevin, 1995, Rinehart et al., 2004 and Webster, 1992). As the end points of the continuum are not relationships they are not studied in the paper. In addition, relationships such as strategic alliances, joint ventures, franchising and licensing are not identified as points on the continuum as the discussion in this paper is restricted to vertical market relationships between buyers and suppliers rather than horizontal non market relationships.

This study is one of the few that has empirically tested which aspects of buyer–supplier relationships effectively discriminate between relationships classified as partnerships, and other forms of cooperative or coordinated relationships. It also provides empirical evidence that buyer–supplier relationships can be placed on a relational continuum. This is a useful contribution, as Young et al. (1996) state that virtually no empirical data exists that provides confirmation or disconfirmation of a relational continuum across a range of relationship forms. 5.1. Implications for theory The results indicate that the theoretical framework used in this study offers a comprehensive method of classifying relationships, as 8 of the 10 variables in the framework differed significantly across relationship type. However, only 4 of these variables differed significantly when partnerships were compared against highly coordinated relationships, indicating that the variables in the framework were less effective at discriminating between these types of relationships. This highlights the need for the development of more comprehensive models of buyer–supplier relationships. In particular, the finding that both tangible and intangible aspects of relationships differed significantly between all types of relationships suggests that if greater understanding into the nature of partnerships is to be gained, attention must be paid to observable and tangible behavior, as well as to the softer aspects of relationships. This highlights the limitations of studies that have attempted to classify relationships using constructs such as relationship strength, which only focus on the sentiments that underlie the exchange such as commitment, trust and relational norms (Bove and Johnson, 2001 and Hausman, 2001). Similarly, relationship quality typically assesses relationships using a combination of sentiments such as commitment, trust, conflict and overall satisfaction with the relationship (Bove and Johnson, 2001, Kumar et al., 1995 and Walter et al., 2003). The limited scope of relationship quality as a purely attitudinal construct is highlighted by Woo and Ennew (2004) who suggest that the scope of relationship quality should be developed to capture both attitudinal and behavioral based dimensions of a business relationship. Similarly, Naude and Buttle (2000) suggest that the scope of relationship quality should be extended beyond the commonly used indicators of trust, commitment and satisfaction to include less frequently utilized relationship constructs such as integration, mutual understanding of needs, power and profit. The need for classification models to include a broader range of constructs is also recognized by Young et al. (1996) who investigated relational differences across a range of marketing relationships using measures comprised of elements and norms. In their study, the elements constituted the more tangible aspects of exchange interaction, and the norms constituted the expected and accepted behavior in the relationship. However, a limitation of their research was that they only looked at a selected set of governance mechanisms. As such they recommend that future research should include other variables such as conflict resolution methods and communication and influence strategies, which may be more capable of clearly distinguishing among different types of marketing relationships. Similarly, Donaldson and O'Toole (2000) suggest that to discriminate properly between different relationship forms, classification methods should consist of a combination of action and belief components, where beliefs capture the social bonding of partners, and action components capture the intensity of interaction between partners. Whilst they conclude that their combination of action and belief components offers broader insights into relationship type, they acknowledge that their construct of relationship strength should be further developed by adding to its action and belief components. Despite acknowledgements that more comprehensive frameworks of buyer–supplier relationship are needed to better differentiate between relationship types, recent studies have still only used a limited set of variables to identify differences between relationships (i.e. Rinehart et al., 2004). The political economy framework is capable of encompassing both action and belief components, as well as the norms and elements of exchange, and both attitudinal and behavioral components of relationships. Therefore it complements the existing literature and offers researchers avenues for the development of current classification models. The political economy paradigm also offer researchers an avenue to include performance in relationship models, something that O'Toole and Donaldson (2000) state is missing in many of the main theoretical models of relationships. Several implications for theory can also be derived from the results of the multiple discriminant analysis. For example, the finding that trust and relational norms had the greatest discriminating effect across relationship type supports the view that trust is a key characteristic of partnerships (Ploetner and Ehret, 2006) and should be included as one of the key discriminating characteristics when developing a robust typology of inter-organizational relationships (Laing and Lian, 2005). It is interesting to note that trust was not identified as a key indicator of partnerships in Lemke et al's (2003) qualitative study of partnership attributes. They state that this was a surprising finding and recommend that researchers investigate how trust relates to supplier partnerships in other industries to see if this finding can be generalized. The results of this study support their view that the finding was most likely context specific. However, in order to compare findings across studies it is vital that the multidimensional construct of trust is clearly defined, as a critical review of empirical research on inter-organizational trust found that the role and number of dimensions varied widely in the studies analyzed (Seppainen, Blomqvist, & Sudqvist, 2007). The second key discriminating variable for relationship type was economy factor three (the frequency and scope of communication and information sharing). This finding highlights that an accurate classification of partnerships should not be limited solely to an examination of relationship sentiments as it is very much an indicator of actual behavior that takes place in the relationship, consisting of questions that measure the frequency and type of communication that takes place, the information linkages that exist and the type and scope of information that is exchanged in the relationship. An interesting finding was the fact that this variable did not effectively discriminate between highly coordinated relationships and partnerships. This lends weight to the argument that traditional indicators of partnerships, such as computer-to-computer linkages, that characterize many JIT relationships, are not indicators of partnerships but instead a prerequisite of doing business (Joseph et al., 1995 and Turnbull et al., 1993). However, given the broad and exploratory nature of the scale used to measure communication and information sharing, and the growing body of evidence about the importance of communication in partnering relationships, (Holden and O'Toole, 2004, Lewin and Johnston, 1997 and Siguaw et al., 1998), it is recommended that the scale used to measure this construct be developed in future studies. In particular, as different facets of communication have been found to vary between difference governance structures (Holden and O'Toole, 2004), future research on relationship type classification would be well advised to draw on the communication literature that poses a link between communication and inter-organizational governance (Mohr and Nevin, 1990 and Mohr and Spekman, 1994). This would enable a more detailed investigation of communication and information exchange, which could uncover significant differences between partnerships and highly coordinated relationships. The third variable identified in the multiple discriminant analysis was interdependence, and appears to be the variable that most effectively distinguishes partnerships from other types of relationships. As such, the results provide support for the theory that interdependence, or mutual dependence, is fundamental in driving the nature of relationships, the type of partnering strategy employed and the extent of collaboration achieved (Frazier and Antia, 1995, Gattorna and Walters, 1996, Ploetner and Ehret, 2006 and Spekman et al., 1998). In addition, this finding highlights the importance of conceptualizing the economy and polity as separate constructs, as it provides researchers wishing to use the political economy framework with a tool to examine interdependence as a separate construct. This was not possible using previous empirical models such as those presented by Reve and Stern (1986) and Nidumolu (1995) and so is a useful contribution given the identification of interdependence as a key discriminating variable in relationship type classification. This is of particular relevance given that several researchers have recently stated that relative dependence (or power) and interdependence are still often overlooked in conceptual and empirical channel studies of buyer–supplier relationships (Caniels and Gelderman, 2007 and Hingley, 2005), and that few studies incorporate the construct of dependence together with other important channel constructs such as trust in a single empirical model (Yilmaz et al., 2005). 5.2. Managerial implications As relationship type was classified by managers, rather than by researchers through the use of cluster analysis (i.e. Rinehart et al., 2004 and Cannon and Perrealt, 1999), it was possible to identify the fact that many practitioners appear to have misconceptions about the exact nature of the relationship they have with a customer. This is indicated by the finding that a higher percentage of relationships described by suppliers as partnerships were incorrectly classified as highly coordinated relationships, rather than correctly classified as partnerships. As O'Toole and Donaldson (2002) suggest that perceptions of relationship type will impact on expectations of performance, any misconception about relationship type will likely result in unmet or unrealistic expectations for the relationship. The classification results in Table 4 also indicate that these misconceptions work in the opposite direction, with relationships perhaps being undervalued by managers. For example, 36% of relationships described by managers as having limited coordination were actually classified by the MDA as highly coordinated relationships, while 19% of relationships described and perceived by managers as being highly coordinated were actually classified as partnerships. Therefore the results indicate that some relationships are not being developed to their full potential. As such firms could be overlooking the opportunity to work more closely with a customer to improve the effectiveness and efficiency of the relationship in question. Therefore in addition to having realistic expectations for the relationship, an accurate assessment of the status of a relationship is important for managers looking to develop appropriate interaction strategies. In particular, a manager who understands what type of relationship they have will be better placed to know when to develop the potential of relationships through the investment of resources to support the relationship (Rinehart et al., 2004). This highlights the importance of developing frameworks that can effectively discriminate between relationship types and help managers identify gaps between perceptions and reality (Holden and O'Toole, 2004). Firms looking to move along the relational continuum should take note of the importance of trust and relational norms, as its key role in discriminating between all relationship types highlights the importance of ensuring a companies workforce, particularly boundary spanners, adopt the right attitudes. The importance of attitudes is highlighted by Christopher and Juttner (2000) who suggest that firms struggle to implement partnership programmes, as although they rationally accept the benefits of partnership approaches, they do not pay enough attention to the soft side of the strategy (i.e. the people issues). More recently, Spekman and Carraway (2006) state that managerial attitudes are still a major barrier to successfully making the transition to, and sustaining, more collaborative buyer–supplier relationships. The finding that interdependence is a key indicator of partnerships highlights the fact that firms looking to be promoted, or even selected for business in a highly competitive market, must review their marketing strategy and seek ways to add value to the relationship. This is because relationship value is often linked to the replacability or criticality of the buyer or seller (Krapfel et al., 1991). The core offerings of a relationship, such as product quality, price and on time delivery offer the weakest source of differentiation in business relationships (Ulaga and Eggert, 2006), and are increasingly perceived as order qualifiers (Lemke et al., 2003), or as entry tickets for relationships (Ploetner and Ehret, 2006). Therefore suppliers must develop more unique and defendable product or service capabilities that they can use to better serve retailers needs and differentiate their service from competitors (O'Keeffe, 2001 and Tuominen, 2004). One area suggested by researchers is superior product and market knowledge as knowledge is a source of value that is difficult for other firms to imitate (Batt and Purchase, Fearne and Hughes, 1999 and Ulaga and Eggert, 2006). Finally, it should be recognised that partnerships are not always advocated or desirable (Spekman and Carraway, 2006). If they are, not all relationships may be capable of progressing along a relational continuum due to the fact that the power structure in the industry can determine the type of relationship that can develop, or is most appropriate to develop (Brennan et al., 2003 and Cox, 1999). In particular, vertical partnerships may only be feasible in circumstances where dominant buyers exist as these have the power to gain vertical cooperation from suppliers (Cox, 2004 and Hingley, 2005). 5.3. Limitations and future research directions In assessing this study's findings, several limitations should be noted. The first is that information was only collected from one side of the dyad. Whilst the suppliers' view of the relationship offers insights into the characteristics of partnerships it would be fruitful for future research to try and collect information from matched pairs. By having matched pairs, where both the buyer and the supplier classify and comment on the same relationship, comparisons could be made between relationships where the perception of relationship type was the same, and where perceptions differed. In addition, by gaining information from both sides of the dyad a more accurate measure of the interdependence structure could be gained, as the retailer, rather than the supplier would have completed the questions about customer dependence. However, while many researchers advocate collecting information about the same relationship from both sides of the dyad, most acknowledge that in practice this is very difficult due to the operational difficulty of getting responses from both sides and an adequate sample size. In addition, researchers of buyer–supplier relationships often collect data on very sensitive issues. Suppliers may therefore be unwilling to identify the customer who they are talking about for fear of commercial reprisals. For these reasons future empirical research is likely to continue collecting information from just one side of the buyer–supplier dyad. The use of a single industry also impacts the generalizability of the results to other populations and some of the findings in this study may be context specific, as has been the case in other studies (i.e., Lemke et al., 2003). Indeed, the context for this study may be a key reason why two variables (dependence asymmetry and exclusive offerings) did not differ significantly between categories of relationship type. This is because in the food industry, retailers are in a dominant position of power relative to suppliers and so are able to demand a certain level of exclusivity from their suppliers in terms of product and/or service (Fearne and Hughes, 1999 and Hingley, 2005). Although, several researchers suggest that the vast majority of channel relationships are imbalanced (Hingley, 2005 and Kumar, 1996) and that strategic partnerships are not necessarily characterized by symmetric power positions (Caniels and Gelderman, 2007) cross industry investigations should be encouraged to enable greater generalizations of findings. Generalization of the results of this study is also limited by the sample size, which constrained the group sizes for some relationship types, particularly those classified as having limited coordination. As noted in the study by Young et al. (1996) another limitation arises from the fact that only a selected set of constructs is used to distinguish between relationships. Although these variables were capable of differentiating between the categories of relationship type investigated in this study, the discriminating power of the framework could be improved through further development and refinement of the scales. In particular, the results indicate that the internal economy scale, used to measure tangible elements of relationships, could be usefully developed to incorporate more facets of communication and influence strategies as these have been identified as mechanisms that may be more capable of differentiating between different types of exchange relationships (Young et al., 1996 and Boyle et al., 1992). In addition, whilst all scales had adequate reliabilities for exploratory research, the lower alpha scores of some measures in the study indicate a need for further refinement of scales. Further work is also needed to improve scale validity as the conflict resolution and performance scales were bi-dimensional, whilst the scale used to measure trust did not load as expected, loading instead with the scale used to capture relational norms. These refined scales should be assessed through confirmatory factor analysis on a larger sample. Furthermore, although distinct differences are discernable between the relationship types investigated in this study, and they fall as predicted on the continuum, it is important to recognize the difficulties of making empirical distinctions between relationship types and identifying at what point on a continuum a relationship is considered a particular type of relationship. Therefore as noted by Young et al. (1996) one cannot assume that exchange categories are mutually exclusive or exhaustively cover the range of relationships that exist. This highlights the need for more continuous variable research that could enable greater delineation of relationship type that fall between the categories of relationships with limited coordination and partnerships. Finally, further research could employ a multidisciplinary perspective and identify differences across a greater range of relationships as only vertical buyer–supplier relationships were examined in this study. Future research could also establish a link between specific relational types and performance as researchers such as O'Toole and Donaldson (2000) state that variations in performance across relationship types remains less understood. In addition, given the episodic nature of many relationships greater emphasis should be given to longitudinal studies as cross sectional data only provides a snapshot of relationships and so does not fully capture the dynamics of relational exchange (Krapfel et al., 1991 and Siguaw et al., 1998).