Mobile Edition
Print
Get the Mag
Weekly UpdatesAbstract Relations between manufacturers and distributors have been the center point of the distribution channel's management. This study covers the effects of coercive, as well as non-coercive power on intermediary variables such as cooperation and conflict. It will also analyze the effects of cooperation and conflict on American car dealers' satisfaction and performance in Spain. Due to the small sample size (46 dealers), the model based on causal modeling compelled us to use the optimization method based on the partial least squares (PLS) regression techniques coupled with a bootstrapping to enable some generalization of the results.
Keywords Channels of distribution * Satisfaction * Performance
Partial Least Squares (PLS).
Introduction
At this point in time, when an important turning point is emerging in the relationship between manufacturers and dealers, we note the increasing power of negotiation of the latter vis-a-vis the manufacturers. There is a tendency for consolidation at the level of the dealers, creating quasi associations to achieve superior powers in purchasing, and novel forms of integrating their common interests. The bottom line for these changes is to improve the competitive position and the negotiating capacity of the dealers, to promote a significant increase in the growth sales through franchising (Stem and El-Ansary 1993); to increase the quality sought by the consumers (a feature of the franchising approach), and to improve the demand for custom quality goods on the domestic and export markets.
The franchising sectors, which have experienced the most dramatic increase in volume of sales, are those in services, and of course, clearly those associated with automobiles sales.
In Spain, the Detroit based companies have had a really good performance since the sixties; even today they do better than in North America. The relationship with the dealerships is premised primarily on the concept of power but more and more on those related to continuous quality improvements and the development of technically superior products (Maynard 2003).
Over the last few years, the American manufacturers have steadily lost market share in America to their surging foreign competitors: Japanese, Korean, European, and possibly soon, the Chinese, but this doesn't seem to be the case in the Spanish market where North America manufacturers hold a steady market share. In North America, the Asian competitors have practiced pressure from below strategies, where they enter the low end of the market and relentlessly start invading the higher margin sectors as the North American companies retreat. The Europeans on the other hand have been applying pressure directly in higher margin segments. Hence, the North American firms are caught in a classic squeeze play, forced to abandon the low ground, as they are receiving steady fire from above.
For example, although General Motors held 60% of the North American market in 1960 this has shrunk to 40% in 2003 and has eroded even faster since 2006. This is not the case for the Spanish market where the Asian car invasion has not begun yet.
Although superficially alike, the North American and the Spanish automotive markets show many differences. For example, in the American market, we can find Japanese and European models in the mid range and high end of the product offerings (luxury segment). This is not necessarily true in the Spanish market where the North American brands are present in middle segments of the market. The average Spaniard consumer also has the penchant to buy more compact vehicles than his North American counterpart (this is due to the price of gasoline at the pump but also to the price tag of cars), and his purchasing behavior is completely different.
Within this context, the objective of this article is to analyze the present relationships between manufacturers and Spanish dealerships selling American cars (GM-Opel-Chevrolet, Ford and Daimler Chrysler) by examining the level of satisfaction and performance (some of the most important factors in vehicle sales) between the dealerships and the manufacturers. Within the context of franchising, considered as a form of business bond between two organizations, it is necessary that there exist a congruence of objectives to maintain the relationship (Spinely and Burley 1996; Mc Neilly and Russ 1992). This co-dependence requires tasks of coordination by the two parties involved, with the goal of obtaining conjoint profits. In the interaction between these two actors, we analyze, from the perspective of the dealer, variables such as: the relative dependence between the two, the sources of power (coercive and non coercive), the exercise of this power between the franchisor and the dealer, the mutual cooperation, the conflict, and observed satisfaction and profitability.
This research analyses specifically the impact of dependence, the coercive and non-coercive powers, the cooperation, the conflict, and finally, the satisfaction and performance of the automobile dealers.
The scope of the study is limited as only 46 dealers (out of a possible total of 400) returned properly completed replies to the survey. Nevertheless, this is an acceptable response rate given that the average survey copes with an 11.5% response rate. The analysis is based therefore on a regression analysis of latent variables using the optimization technique of the Partial Least Squares (PLS).
This study puts forth: a conceptual model, and a strategic framework with a set of hypotheses to be tested, a description of the employed methodology, an analysis of the results, and finally, the pertinent management conclusions along with a proposed future research course of action.
Conceptual PLS Model and Theoretical Framework
The PLS is a multivariate technique to test structural models (Wold 1985) and a general method to estimate path models with latent variables measured by many items. This tool has as a main objective, the causal predictive analysis where problems are complex and the theoretical support is very limited (Wold 1982). This technique is particularly suitable at a very early stage of development of the theory.
Basically, the objective of the PLS modeling is predicting dependent variables, latent and manifest, maximizing the Explained Variance ([R.sup.2]) of the dependent variables and minimizing the residual variance of endogen variables. The differences between the PLS method and models based on Structure Equation Modeling that lays out on its objectives (Barclay et al. 1995; Wold 1982; Levy Mangin and Varela Mallou 2006), suggests that the PLS optimization method is more suitable to predictive applications and theory development (Exploratory Analysis) while SEM methods will be more suitable to confirmatory analysis. Algorithms are also different, SEM use second order functions, like Maximum Likelihood, to maximize the parameter estimates while the PLS applies an OLS (Ordinary Least Squares) iterative sequence and multiple linear regressions analyzing a construct each time. The PLS method estimates the model parameters which minimize the residual variance of the whole model dependent variables (Chin 1998; Hsu et al. 2006).
The PLS does not require any parametric conditions, which are compulsory in the analysis of the SEE, so this technique is particularly appropriate in the case of small samples with non-normal data (Chin 1998). These PLS characteristics are different from those of the SEE models based on covariance analysis, which requires a minimum sample of 100 (Hair et al 1999) to 150 cases (Bollen 1989) due to the sensitiveness of the Chi-square test.
Due to the condition of this research and to the scope limited to 46 dealers, the use of an exploratory method was seriously considered and motivated us to discard the use of a more confirmatory SEM method.
Channels of distribution are economic and social systems (Stern and Reve 1983) that are managed as super organizations, which include dependence (Emerson 1962; Frazier 1983a; Stern and El-Ansary 1993; Skinner et al. 1992) and the basis of power (Hunt and Nevin 1974; Gaski 1984 and Gaski and Nevin 1985).
The concepts of dependence and the basis of power constitute the fundamental structure of the model, which is represented by the nature of the relations between different parts (Anderson and Narus 1984, 1990) that could lead to dealer's satisfaction (Robicheaux and El-Ansary 1975; Frazier 1983a, 1983b; Anderson and Narus 1984, 1990; Lopez et al. 2002; Yieh et al. 2007).
Conflict and cooperation correspond to dominant feelings, which regulate exchanges between manufacturer and dealer. The first one refers to all moves that impede the dealer to attain the identified objectives, and the second refers to all actions that favor the attainment of these objectives. These concepts are distinct but not necessarily opposed; both of them lead to the interdependence of exchanges. They are moderating latent variables to satisfaction (Stern and Reve 1983; Brown and Day 1981; Lusch and Brown 1982; Dant and Gijndlach 1998; Gassenheimer et al. 1996).
The conflict, the satisfaction and the performance concepts have been used in many researches (Etgar 1979; Lusch 1976; Dwyer 1980; Brown and Day 1981; Brown et al. 1983); some of them, like cooperation and dependence, have received a more discrete attention in the marketing literature. It was a supplementary reason to include them to our model.
Our structural model has seven constructs issued from these concepts, and twenty hypotheses have been generated from the relations of these seven constructs (HI to H20).
This model could be laid out with the next structural equations (1):
Dependence = [[gamma].sub.1] Non coercive power + [[zeta].sub.1]
FEED