The Impact of Retailer ’ s Economic and Social Satisfaction on Its Commitment , and the Moderating Effect of Manufacturer Brands ’ Strength

Available literatures have confirmed that there are unstable and complex relationship between manufacturers and large retailers especially in the growing strength of retail organizations in the global context. This study thus examines the impacts of retailer’s satisfaction dimensions on its commitment, and the moderating effect of the manufacturer brands’ strength on the relationships between retailer’s satisfaction dimensions, and its commitment. This quantitative study presents the empirical findings from a survey of 140 large retailers using partial least squares (PLS). The results indicate that retailer’s economic and social satisfaction have strong positive influence on retailer’s commitment, and surprisingly, the retailer’s social satisfaction is more influential on its commitment than retailer’s economic satisfaction. The study also found that the strength of manufacturer’s brands enhances the relationship between retailer’s economic satisfaction and its commitment, but not the relationship between retailer’s social satisfaction and its commitment. Theoretical and managerial implications are extracted from these findings.


Introduction
There has been a notable shift in the role of retailers since 1970s from being just a distributor to being a partner in a business relationship (Pelau, 2008) because of the transformation of the production markets to customer demand markets.The increasing competition and product homogeneity in several consumer categories that enabled retailers to choose manufacturers are also responsible in changing the role of the retailers (Mitchell, 2004).Coughlan et al. (2010) contend that upstream manufacturers are vocal about their aversion to being monopolized by large retailers despite the fact that majority of them enter into feasible partnerships with the latter.As a result, a notable shift in the behavior of manufacturer and retailer is noted (Pelau, 2008), urging the need to understand the retailer-manufacturer relationship and how this relationship shapes the retailer's satisfaction and commitment toward the manufacturer.
Securing and maintaining retailer's commitment is a core strategy employed by manufacturers.This is because commitment is viewed as the fundamental characteristic to any successful relationship with the desire for long-term benefits (Morgan & Hunt, 1994), such as, ensuring retailer's shelf space for the products of the manufacturer (Amrouche & Zaccour, 2007).Research on commitment has focused on Business to Consumer (B2C) and Business to Business (B2B) realms, but the latter is few and far between particularly in the context of non-durable goods, this may be due the limited understanding of the importance of commitment for maintaining long-term benefits and a positive relationship between parties of business relationship (Ali, 2011).In addition, many recent studies examined commitment as a consequence of overall relationship satisfaction (e.g., Davis-Sramek, Mentzer, & Stank, 2008;Sharma & Paul, 2000), but very little research has addressed commitment as a consequence of economic and social satisfaction (Goaill, Selvan, & Nor Azila, 2013).
Therefore, empirical study is needed to fill this gap.
Prior studies on relationship marketing have concentrated on various relational constructs like relationship satisfaction and commitment, which are deemed the most significant in developing and maintaining the relationship between manufacturer and retailer (e.g., Anderson & Weitz, 1992;Morgan & Hunt, 1994).They also were found to determine the quality and strength of the relationship (Kumar, Scheer, & Steenkamp, 1995) and the exchange performance (Palmatier, Dant, Grewal, & Evans, 2006).In the context of this background, we have considered different dimensions of relationship satisfaction (economic and social) and commitment as the relational indicators of the ongoing success of manufacturer-large retailer relationship.
A retailer is generally dependent on the manufacturer/supplier to supply products and to attract buyers (Anderson & Narus, 1990).Hence, the retailer has to have enough product lines to attract customers to the store specifically when it comes to consumer goods (such as food).This is why there are several items in a large retailer's store as the retailer represents the products of the manufacturer in a particular product line.Emerson (1962) describes this view of power as a dependence function.To him, the power of a manufacturer is higher because their products are significantly differentiated for consumers so that the retailer will be susceptible to offering them for sale and will not entertain similar products from other manufacturers.Brand strength is the manifestation of the manufacturer's power (Haines, 2007) and a strong brand is the core of this power (Anderson & Narus, 1990;Fein & Anderson, 1997).Brand strength also determines the manner the retailers and manufacturers tackle their business relationships (Glynn, Brodie, & Motion, 2012).Moreover, the manufacturer's strong brands have been found as a source of power or a pledge of long-term continuity in a channel relationship (Anderson & Weitz, 1992;Brown, Lusch, & Nicholson, 1995), and can enhance the relationship with the retailers (Frazier & Anita, 1995).
Despite its importance in the relationship between manufacturers and retailers, surprisingly, the effects of brands strength in this context have hardly been investigated.According to Glynn (2010), marketing channel relationships provide a unique opportunity to investigate the effect of the brand strength of the manufacturer in a B2B context.Paulssen and Birk (2007) also note that the potential moderating role of the brand strength has so far been neglected in the business-to-business context.Goaill et al. (2013) further suggest examining the influence of manufacturer brands' strength as a moderator in the relationship between retailer's satisfaction dimensions and its commitment.
In developing countries, there has been little behavioral research on manufacturer-retailer relationships (Ali, 2011), especially in the Arab countries' business environment, despite the fact that this relationship is characterized by complex and hostile at the present time (Coughlan et al., 2010).In particular, the relationship between manufacturers and large retailers in developing countries might be significantly different from those of developed countries in many aspects.There is a real need to assess the validity of existing marketing theories in the context of developing countries (Ren, Oh, & Noh, 2010).
The Middle East retail sector including Yemen is fast overtaking other emerging markets.According to a report on the Middle East retail sector, it is expected that the Middle East retail industry will grow at a compound annual growth rate (CAGR) of around 10% during 2011-2014 to reach over US$1,000 billion by 2014 (RNCOS, 2011).In addition, the report anticipated that the retail sales will also flourish in coming years with the increasing retail space in the region.Further, the report confirmed that the retail culture has evolved from traditional outlets to large shopping malls, hypermarkets, supermarkets, and organized chains (RNCOS, 2011).Especially in Yemen, retailer modernization was developing on a rapid scale in the last six years leading to an increased number of retail organizations through the modern super market system to around 100% (156 large retail organization in 2005 to 309 in 2011) (YCGSI, 2011), but there is a mutual relational gap between food manufacturers and large retailers in contrast to medium and small retailers (Al-Gaumey, 2010).In addition, the Yemeni Industrialists Association (2010) proposed that an investigation into the relationship between food manufacturers and retail organizations be conducted.
In the context of developing countries including Yemen, large retailers have equal or even greater power than manufacturers when it comes to the retailing stage of supply chains, and that can increase the complexity of their relationship.As a result, manufacturers find themselves in the position of finding new ways to remain competitive in the era of retail power.For majority of manufacturers, this indicates a need for a change from a mindset of obtaining consumer loyalty to "branding from the shelf" (Lincoln, 2006), or achieving a high level of relationship satisfaction to establish retailer commitment to manufacturer (Chiou, Wu, & Chuang, 2010;Davis-Sramek, Droge, Mentzer, & Myers, 2009).The question remains concerning the best concentration to achieve high levels of retailer's economic and social satisfaction and commitment to ensure appropriate shelf-space for the manufacturer's products in the context of a relational exchange.
Therefore, the present study examines the effect of economic and social satisfaction of large retailers on their commitment to food manufacturers.The study also aims to shed light on whether or not the strength of the manufacturer's brands further improves the retailers' economic and social satisfaction on their commitment.Toward meeting these objectives, this study is organized as follows.First, a brief literature review is conducted on the focal concepts.This is followed by hypotheses, method used, and results.Finally, theoretical and managerial implications of the results are highlighted and the study limitations explained.

Literature Review and Hypothesis
2.1 Retailer's Commitment Moorman, Deshpande, and Zaltman (1993) define commitment as a lasting desire to keep a relationship that is positively valued.Similarly, Dwyer, Schurr, and Oh (1987) state that commitment is an implicit pledge of relational continuity between partners to the relational exchange, which indicates their desire for long-lasting benefits, even if the exchange entails sacrifices.Commitment is deemed to be crucial for successful relationships (Morgan & Hunt, 1994).In the context of a retailer-manufacturer relationship, retailer's commitment is an implicit promise to continue and support the relationship with the manufacturer.
Commitment is the key consequence of relationship satisfaction in channel relationship research.Wilson (1995) outlined a list of key variables that contribute to successful buyer-seller relationships and found that commitment was the most common dependent variable in that list.Commitment is also relevant to this research because retailers may be satisfied with the manufacturers, but it may not necessarily mean that they are willing to make sacrifices to accrue long-term benefits and have desire to maintain a positive relationship, if they do not show commitment to the manufacturers.

Retailer's Economic and Social Satisfaction
Majority of researchers have shed light on two satisfaction dimensions namely economic and social satisfaction in the manufacturer-retailer relationship (e.g., Chung, Huang, Jin, & Sternquist, 2011;Rodriguez, Agudo, & Gutierrez, 2006).These dimensions were confirmed in an earlier meta-analytical study by Geyskens, Steenkamp, and Kumar (1999).They revealed that relationship satisfaction has multiple dimensions.But only a few examined retailer's economic and social satisfaction as two distinct variables (Goaill et al., 2013).In this present study, both dimensions of satisfaction (economic and social) are considered because a retailer may be economically satisfied with a manufacturer but not socially satisfied with it and vice versa.

Retailer's Economic Satisfaction
Retailer's economic satisfaction is defined by Geyskens and Steenkamp (2000) as the evaluation of the retailer of the economic outcome that results from the relationship with the supplier.It describes the positive reaction from the supplier that stems from the economic rewards obtained like volume, profit margins and discounts.Geyskens et al. (1999) claim that a retailer that is satisfied economically views the relationship as being successful in light of goal attainment, relationship effectiveness, and productivity and financial results.But despite the fact that an economically satisfied channel member would view the relationship as a success, it is imperative to consider the notion that decisions made purely on economic criteria, is now considered unrealistic (Schellhase, Hardock, & Ohlwein, 2000).It is therefore necessary to achieve social satisfaction in the relationship.Mohr, Fisher, and Nevin (1996) define retailer's social satisfaction as a positive and affective reaction to the psychosocial elements of the relationship, in a way that its transaction with the manufacturer is filled with fulfillment, satisfaction and ease.A retailer, who perceives satisfaction in a socio-psychological way, shows appreciation for interactions with the supplier and finds satisfaction working with the latter on a personal level.The retailer is also convinced that the supplier is concerned, respectful and willing to exchange ideas (Geyskens et al., 1999).

Retailer's Social Satisfaction
Social exchange theory (Thibaut & Kelly, 1959) postulates that a retailer's social satisfaction stems from the evaluation of social outcomes in its interaction with the manufacturer.For example, if the manufacturer possesses a good social bond with the retailer, this may result in the retailer's positive social outcomes and the perception of a fulfilling interaction.Researchers have also addressed relationship satisfaction in its social aspect and described it as the evaluation of the interaction experiences (Gassenheimer & Ramsey, 1994;Scheer & Stern, 1992) and a reflection of the psychosocial well-being of the firm (Pelton, 2013).

Relationships between Retailer's Economic and Social Satisfaction, and Commitment
A thorough literature review validates Ganesan's (1994) claim on relationship satisfaction's role in determining commitment in the retailer-vendor relationship.Hennig-Thurau, Gwinner, and Gremler (2002) empirically revealed that overall relationship satisfaction is significantly related to commitment.Other studies also reported similar result (Johnson, Krapfel, & Grimm, 2001;Sivadas, Holmes, & Dwyer, 2012;Sharma & Patterson, 2000).Although insightful, the studies considered the role of overall relationship satisfaction in influencing commitment.On the basis of prior argument and the postulation of the social exchange theory that a modern notion of relationship satisfaction (economic and social) can result in commitment, this study developed the following hypotheses: H1.Retailer's economic satisfaction has a positive effect on its commitment toward the manufacturer.
H2. Retailer's social satisfaction has a positive effect on its commitment toward the manufacturer.

Moderating Effect of Strength of Manufacturer's Brands
Strength of brands represents the power of a manufacturer.It refers to the manufacturer's capacity to achieve prestige of brands, market share and customer loyalty to the brands (Iglesias & Vazquez, 2001).Among the studies of buyer-supplier relationship, Fein and Anderson (1997) found that the strength of the manufacturer's brand is a source of the unstable relationship between suppliers and retailers.In addition, strength of brands was found to determine the way retailers and manufacturers handle the B2B relationship (Glynn, 2010).According to Nevin (1995), building channel relationships is not the same in all manufacturer-retailer relationships as each relationship has governance mechanisms that support the manufacturer's brands.Hence, understanding the effect of brands strength on the retailers-manufacturers relationship is deemed to be a crucial issue for both parties and for academics.
A manufacturer with strong brands depends largely on strong consumer demand and may not have to opt for developing a close relationship with the retailer.On the other hand, a manufacturer with weaker brands is less capable of depending on the market demand and therefore it often uses deals to achieve the retailer cooperation (Curhan & Kopp, 1987).Retailers translate weak brands to high commitment to the manufacturers to promote the latter's product, which results in improving the store image of the former and achieving a more unique market demand (Webster, 2000).In this way, the retailer is able to gain benefits by associating itself with the manufacturer of weaker brands.The influences of inter-organizational resource for brands seem to collide with brand literature postulating that strong brands reflect greater trade cooperation, commitment and support (Keller, 2003).
Even though a manufacturer-retailer relationship has often been seen as adversarial, emphasizing inter-firm power and control (Gaskin, 1984), a manufacturer's power (strength of brands) can also enhance the relationship with the retailers (Frazier & Anita, 1995).This is because the strength of manufacturer's brands can be seen as a pledge of long-term continuity in a channel relationship (Anderson & Weitz, 1992).Thus, on the basis of the above arguments, the strength of manufacturer's brands is expected to positively moderate the association between retailer's satisfaction dimensions and its commitment to the manufacturer.The following hypotheses are thus postulated: H3.The strength of manufacturer's brands positively moderates the relationship between retailer's economic satisfaction and its commitment to the manufacturer.H4.The strength of manufacturer's brands positively moderates the relationship between retailer's social satisfaction and its commitment to the manufacturer.

Figure 1. Theoretical framework
The above theoretical framework in Figure 1 depicts the model that shows the network of variables to be tested in the present study.

Sample and Data Collection
This study focuses on large independent food retailers in Yemen (such as malls, hypermarkets, supermarkets and large grocery stores).A list of food retailers was obtained from the database of Yemen Company for Ghee and Soap Industry (YCGSI), which comprehensively details the relevant information such as the retailers' names, location and contact information.As of August 2011, there were 309 large food retailers in the country, which were evaluated based on the annual sales and their size > 70 sq.m (YCGSI, 2011).
As the level of analysis was organizational, sales managers or retail owners, who were responsible for merchandise inventory and dealing with food manufacturers' representatives, were approached as participants.According to John and Reve (1982), and Kumar, Stern, and Anderson (1993), key informants, i.e., those who have direct knowledge or influence on the buying or selling in the organization, can be used to provide organizational level attributes provided that the survey items are directed at the organizational level.
Because the population size was already small, the researchers decided to study the whole population rather than taking a sample for the study, following Zikmund's (2003) recommendation.In the final study, questionnaires via drop-off-survey were distributed to 279 retailers (30 retailers had already been considered in the pilot test).Of 279, only 151 retail companies completed the survey.But eight were excluded because they were not filled by either by managers or owners or assistant managers; instead they were filled by sales clerks or part-timers who clearly stated their positions as requested by the instructions given in the questionnaire.In addition, three questionnaires were discarded because more than 50% of them were incomplete, following suggestion by Hair Black, Babin, and Anderson (2010).All in all, the total usable responses for further analysis stood at 140, constituting 50.2% valid response rate.

Sample Characteristics
With respect to length of work experience in the retail industry, 51.7% of the participants had been working more than 10 years, while 30.8% between 5 and 10 years.Those who reported to have been working less than five years constituted 17.5% of the participants.Generally, more than 50% of the large retailers had more than 10 years of relationship with food suppliers in Yemen.This means that most of the responding retailers were in mature relationships and had accumulated experience in a channel relationship.In terms of position of the participants, slightly more than half of them (53.1%) were managers in large retailers, while only 6.3% were assistant managers.Owners of the retailer business constituted 40.6%.
Of the 140 participants, 21% had achieved annual sales turnover between RY40 million and RY60 million,  18.9% had between RY20 million and less than RY40 million, 18.2% between RY1 million and less than RY20 million, 14.7% less than RY1 million, and 10.5% more than RY100 million.Meanwhile, only 8.4% had achieved annual sales turnover between RY80 million and less than RY100 million, and between RY60 million and less than RY80 million.Overall, most participants enjoyed high sales, implying they had some degree of power in the relationship with their suppliers.

Measures
This study used reflective measurements based on previous research on distribution channels.All items were rated on a five-point Likert scale (1 = "strongly disagree", 5 = "strongly agree").Five items each was used to measure retailer's economic satisfaction and social satisfaction.The scales were adapted from Geyskens and Steenkamp (2000), and Lai (2007).Five items of retailer's commitment were taken from Nor Azila and Aziz (2012).Finally, strength of manufacturer's brands measure was adapted from Iglesias and Vazquez (2001), who used three items.All questions were translated into Arabic, and back-translated into English to check for consistency in meaning.
Before the final distribution of the survey instrument, a pilot study to check the reliability and validity of the constructs was carried out.A drop-off delivery method to 30 large retailers was used to distribute the questionnaire.The pilot study indicated that the internal reliability coefficients ranged from .86 to .95 for the main constructs.Exploratory factor analysis revealed that five items loaded on commitment, 10 items on economic and social satisfaction, and three items on strength of manufacturer's brands.The values of KMO were .714,.715and .733,respectively, exceeding the suggested value of .60 (Tabachnick & Fidell, 2007).The eigenvalues were 3.64, (5.18, 1.83), and 2.77, respectively, which were above 1 (Tabachnick & Fidell, 2007).The varimax rotation revealed that all items had factor loadings between .62 and .97.Based on the result, all items in the scale were used in the final study.

Analysis
Partial least squares-structural equation modeling (PLS-SEM) was utilized for data analysis through SmartPLS, as recommended by Ringle, Wende, and Will (2005).PLS-SEM has several features that result in its widespread use in marketing and management studies (Al-Dhaafri & Al-Swidi, 2014; Anderson & Swaminathan, 2011;Liu, Li, & Xue, 2010).PLS-SEM is primarily a soft-modeling approach (Wold, 1985) and is not as suitable as some methods to test complex theories owing to the lack of global optimization criterion that assesses the fitness of the model (Hair, Sarstedt, Ringle, & Mena, 2012).However, it is beneficial in comparison to the covariance-based structural equation modeling in the analysis of predictive research models that are in the infancy stages of theory development (Fornell & Bookstein, 1982).The latter mirrors the present research.Despite the fact that prior studies of exchange relationship have considered relationship satisfaction, commitment and brands strength, there is lack of research focusing on the interaction between brands strength, satisfaction dimensions, commitment that could enhance the exchange relationship.Hence, in the context of limited research works, PLS-SEM is a suitable data analysis technique.Secondly, PLS-SEM enables the researchers to examine the role of the moderating effect (Vinzi, Chin, Henseler, & Wang, 2010).Thirdly, it is appropriate for small sample sizes (Henseler, Ringle, & Sinkovics, 2009) as it reveals higher statistical power than the covariance-based SEM when employed on complex models with limited sample sizes (Reinartz, Heinlein, & Henseler, 2009).This is particularly relevant in the present study as the final sample size was 140 cases.

Results
A PLS model was analyzed and interpreted in two stages (Hair, Ringle, & Sarstedt, 2011;Valerie, 2012).First, the measurement model (outer model) was tested to ensure its validity and reliability.Measurement properties of multi-item constructs, including convergent validity and discriminant validity, were examined by conducting confirmatory factor analysis (CFA).Then the proposed structural model was analyzed by looking at the R square value, effect size, and predictive relevance of the model.Finally, bootstrapping was used to test the hypotheses.

Measurement Model
In assessing the measurement model, the researchers followed the guidelines suggested by Hair, Ringle, and Sarstedt (2011) and Valerie (2012) whereby construct, convergent, and discriminant validity was examined.The measurement model was assessed by examining the following criteria: (1) All items should be statistically significant with loading .70 or higher (Hair et al., 2011); (2) The consistency reliability (CR) and Cronbach's alpha must exceed .70(Hair et al., 2011); and (3) The average variance extracted (AVE) should exceed the generally recognized .50cut-off point (Hair et al., 2011).As shown in Table 1 and 2, after removing one item with loading of less than .7 (SociSat5 = .56),the consistency reliability and AVE values met the recommended threshold.All items loaded significantly on their respective construct at the level of .01.
The discriminant validity of the instrument was verified by examining the square root of the AVE as recommended by Hair et al. (2011).The square root of AVE represented as the diagonal elements in the constructs correlation matrix should be greater than the off-diagonal elements in corresponding rows and column.
In this study, none of the intercorrelations of the constructs exceeded the square root of the AVE of the construct (see Table 3).

Structural Model
After analyzing the measurement model, the next step was to evaluate the structural model i.e., by analyzing the inner model.To do this, the researchers depended on the recommendations of by Chin (2010), Hair et al. (2011), andValerie (2012) to include the R² values, effect size (f²), predictive relevance of the model, and the level and significance of the path coefficients.Finally, bootstrapping was employed to test the study hypothesis.
According to Hair et al. (2011), the most crucial criterion for evaluation of the structural model is the R² measures.This is because the aim of the prediction-oriented PLS-SEM method is to clarify the endogenous latent variables' variance.Therefore, the level of R² main target constructs' level has to be high.The determination of what exactly is high R² hinges upon the kind of research discipline.For instance, R² value of .20 is deemed high in disciplines like consumer behavior while an R² value of .75 is deemed high in success driver studies.Generally speaking, in marketing research studies, R² values of .75,.50, or .25 for endogenous latent variables in the structural model may be established as the rule of thumb and described respectively as substantial, moderate, and weak (Hair et al., 2011).Based on the results from running PLS algorithm, the R² value of retailer's commitment was .575,suggesting that 57.5% of the variance in retailer's commitment can be explained by economic and social satisfaction, brands strength, brands strength*economic satisfaction, and brands strength*social satisfaction.The R² value of retailer's commitment was in the moderate range.
Effect size (f²) is used to determine the effect sizes of specific latent variables' impact upon the dependent variables with the help of f² analysis, which is complementary to R² (Chin, 2010).The f² effect size is calculated as it is not automatically provided in PLS.The researcher manually calculated it with the help of the formula: f² = (R² included-R² excluded)/(1-R² included).The f² values of .02,.15 and .35,respectively, were used to guide for small, medium and large effect sizes of the predictive variables, as recommended by Cohen (1988).The effect sizes of specific latent variables and the moderator's role can be evaluated by the same formula proposed by Cohen (1988).
Table 4 shows a small moderating effect size of brands strength on the relationship between economic satisfaction and commitment (f² = .026)and a very small moderating effect size of brands strength on the relationship between social satisfaction and commitment (f² = .001).Also, there was a medium effect of social satisfaction on commitment with f² values of .244while there were small effects of economic satisfaction and brands strength with f² values of .108and .135,respectively.Based on the recommendation of Hair et al. (2011), cross-validated redundancy was assessed by the PLS-SEM estimates of both the structural model and the measurement models to predict data, which perfectly fit the PLS-SEM approach.If an endogenous construct's cross-validated redundancy measure value (i.e., Q²) for a certain endogenous latent variable is larger than zero, its explanatory latent constructs exhibit predictive relevance.The Q² is a criterion to evaluate how well the model predicts the data of omitted cases which is referred to as predictive relevance (Hair et al., 2011).To use blindfolding to obtain Q², Hair et al. (2011) recommended that the number of cases in the data must not be a multiple integer number of the omission distance d "otherwise the blindfolding procedure yields erroneous results", and they suggested to choose a value of d between 5 and 10.Therefore, this study used 9 as a value for d to obtain cross-validated redundancy measures for the dependent variable.As suggested by Hair et al. (2011), the model will have predictive quality if the cross-redundancy value was found to be more than zero; otherwise the predictive relevance of the model cannot be concluded.The result indicated that the obtained cross validated redundancy for commitment was 0.440.This result supports the claim that the model has an adequate prediction quality.
The final step was to test the hypothesized relationships by running PLS algorithm and Bootstrapping algorithm.
Based on the recommendation by Hair et al. (2011), we used 5000 samples with number of cases equal to our observation of 140.Our critical t-values for a three-tailed test were 1.65 (with a significance level of 10%), 1.96 (with a significance level of 5%), and 2.58 (with a significance level of 1%).Table 5 contains the path coefficient and the bootstrapping result.Table 5 shows that the moderating effect of manufacturer's brands strength on the relationship between economic satisfaction and commitment was supported (H3).SPSS was used to detail the relationship by drawing a line graph.Figure 2 show two different curves of the relationship between economic satisfaction and commitment that represent strong and weak brands of food manufacturers.The curves of strong and weak brands were always positive for the relationship between economic satisfaction and commitment, but the curve of the strong brands was found to be faster than the curve of the weak brands (Dawson, 2013).This means the strong brands could enhance the retailer's commitment towards the manufacturer when it is economically satisfied more than weak brands of the manufacturer could.In other word, when a manufacturer has strong brands, the retailer's economic satisfaction will further enhance its commitment towards the manufacturer.

Discussion
Our study shows that retailers' economic satisfaction and social satisfaction have a strong positive influence on their commitment to food suppliers.This indicates that large retailers receive economic and social benefits from their supplier and are willing to reciprocate the suppliers' effort as a result.In other words, the retailers' economic and social satisfaction with the supplier is instrumental in increasing their desire to maintain the relationship.The finding of this study is consistent with that reported by Nor Azila, Selvan, and Zolkafli (2010) who found that economic and social satisfaction has a significant impact on commitment in the dealer-supplier relationships in the context of durable goods.In general, the finding also is consistent with other studies that found a positive relationship between overall relationship satisfaction and commitment (e.g., Garbarino & Johnson, 1999;Sivadas, Holmes, & Dwyer, 2012;Sharma & Patterson, 2000).
More importantly, the present study found that retailers' social satisfaction is more influential than economic satisfaction in enhancing their commitment toward the supplier.This indicates that the retailers' positive and affective reaction to the psychosocial elements of the relationship is able to generate higher commitment than their positive response to economic rewards derived from the relationship with national manufacturers.This finding supports the idea of Cook and Emerson (1978), who argue that commitment is central in distinguishing social from economic exchange.
The study also examines whether strength of food manufacturer's brands moderates the relationship between large retailer's economic and social satisfaction, and commitment.Two hypotheses were formulated toward that end.As shown in Table 5, brands strength positively moderates the relationship between the retailer's economic satisfaction and its commitment to the food manufacturers, thus providing support for H3.But strength of brands has no significant moderating effect on the relationship between the retailer's social satisfaction and its commitment to the food manufacturers, hence providing no support for H4.The result indicates that the strength of brands plays an important role in enhancing the relationship between economic satisfaction and commitment but it does not play a significant role in enhancing the relationship between social satisfaction and commitment.
The non-significant result may be because the relationship between the retailer's economic satisfaction and its commitment is more likely to be economic than social.As the brands strength can add economic value to the relationship, commitment of retailers to such relationship is therefore increased.
Even though the present result indicates the important role of brands strength on the relationship between economic satisfaction and commitment, some scholars argue that weak brands may give more advantage to the relational exchange than strong brands because manufacturers having weak brands can better build relationships with retailers as they are more reactive to the individual retailer's needs (Anselmi, 2000;Glynn, 2010).In order to assess the validity of the argument, the study uses the graph lines produced by SPSS to explain the relationship and identify which brands are more enhancing.Figure 2 show two different curves in the relationship between economic satisfaction and commitment that represent strong and weak brands of the food manufacturers.Even though both brands appear to be able to enhance the retailer-manufacturer relationship, it seems that the strong brands have a more enhancing value (the dotted curve) than weak brands (the solid curve) (Dawson, 2013).This suggests that the higher retailer's economic satisfaction is derived from the benefits it gets from the manufacturer's strong brands, which result in the former's commitment towards the manufacturer.In other word, when a manufacturer has strong brands, a retailer that is economically satisfied will be more committed towards the relationship than if the manufacturer carries weak brands.
The finding of the current study seems to be consistent with brand literature that suggests that a strong brand is akin to greater trade cooperation and support (Keller, 2003).Also, this finding is consistent with that of Frazier and Antia (1995), who demonstrated that a manufacturer's power (strength of manufacturer's brands) can enhance the relationship with retailers.However, the finding of this study does not support the study of Glynn (2010), who found that the effect of reseller relationship satisfaction on commitment is more significant for minor brands (weak brands) than for major brands (strong brands).The likely reason for the inconsistent finding may be because the retailers in the present study perceive that they receive more economic benefits from a manufacturer's strong brands than from a manufacturer with weak brands.For example, a high profit margin through the fast rate of sales turnover, greater discounts, and the provision of the needs and desires of their customers may lead to a high level of commitment towards the manufacturer.This may lead to continued relationships with companies that have strong brands, as reflected in increased economic satisfaction and commitment.

Theoretical Contribution
The empirical testing of consequences of economic and social satisfaction and the moderating effect of brands strength on the consequences of the relationships is the major theoretical contributions of this research, particularly in the manufacturer-retailer relationships in the retail and consumer goods industries.
The study also provides more evidence to the literature of the important role of the satisfaction dimensions in influencing commitment.Achieving high levels of economic and social satisfaction are essential to increase commitment of retailers to successful and continued relationship with manufacturers.
Another theoretical contribution of the present study is with regards to the moderating effect of the strength of manufacturer's brands on the relationship between retailer's relationship satisfaction (economic and social) and its commitment, adding evidence to the existing body of relationship marketing literature (Glynn, 2010;Goaill et al., 2013;Paulssen & Birk, 2007).The findings of the research strengthen the role of the manufacturer brands' strength in enhancing the relationship between manufacturer and retailer to achieve high levels of economic satisfaction and commitment which helps to stabilize the exchange relationships.Thus, the study contributes to the theory of social exchange by providing a broader theoretical perspective concerning the role of power (strength of brands) in the relationship between economic and social satisfaction, and commitment in the context of business-to-business exchange relationships.
This study is one of the very few studies conducted in developing countries especially in the Arab countries' business environment to investigate the manufacturer-retailer relationships in the context of B2B.This is because very few empirical works has been conducted to examine the effect of economic and social satisfaction on commitment in the context of channel relationships.In addition to that, this study has expanded the boundary of the current literature as it investigated the role of the strength of manufacturer's brands in explaining how different economic and social satisfactions enhance commitment.In sum, this study contributes to an expanding research stream on relationship marketing currently dominated by Western research works by adding the Arab perspective.

Managerial Implications
This study provides an insight into how managers of food manufacturers in developing countries can ensure the long-term success of their mutual business relationship with large retailers in light of the growing power of the latter.This can be achieved by understanding how to satisfy the retailers economically and socially so that they are committed to the relationship.
The findings of this study can raise awareness among marketing and sales managers of food manufacturers on the importance of the strength of brands in promoting retailers' economic satisfaction and their commitment to the suppliers.Moreover, the results also suggest that strength of brands is one of the main factors for manufacturers to achieve a strategic competitive position in the emerging markets.In the current business environment in the Middle East, food manufacturers should build strong brands in order to enhance the stability of the manufacturer-retailer relationships by ensuring that the retailers are satisfied and committed to their supplier.Taking the advantages of these findings, owners and managers of food manufacturers should put in place effective plans to enhance their business relationship with large retailers and facilitate commitment.In this context, the result contributes modestly to literature in a B2B relationship in developing countries, particularly in the Middle East.

Limitations and Future Research
The findings of this study and consequently the conclusions drawn are based on the perceptions of retailer's representatives at one point of time; this study does not consider the psychological and attitudinal changes they may have experienced over time as a result growing pressure and competition in the marketplace.The second limitation of the present study is associated with the generalizability of the findings.Because the present study was conducted in a consumer goods industry (non-durable goods), the findings may not be generalizable to other industries.To expand the external validity of the findings, future research should consider different types of industries/sectors in the context of B2B relationship such as durable industries (furniture, appliances, electronic equipment, cars etc.), service sector (banking, tourism, travel etc.), and the like.

Conclusion
A manufacturer's relationship with retail companies will remain one of the major issues in light of strong competition and product homogeneity in many consumer categories.In this situation, many retailers have switched from one manufacturer to another, forcing many manufacturers to compete for retail business.The enhancement of the overall relational performance of manufacturers with retail companies has been the concern of all decision makers in developed and developing countries.
The present study has examined the effect of economic and social satisfaction of large retailers on their commitment to food manufacturers.It has also looked into how and why the strength of manufacturer's brands is a source that strengthens the relationship with retailers (Frazier & Antia, 1995) and determines how retailers and manufacturers approach the business-to-business relationship (Glynn et al., 2012).
The study showed that retailer's economic and social satisfaction had a strong positive influence on their commitment towards the supplier.Furthermore, the present study revealed that the strength of manufacturer's brands enhances the relationship between economic satisfaction and commitment but not the relationship between social satisfaction and commitment.
To sum up, this study has shown that a strong relationship between manufacturers and retailers should be maintained for mutual benefits.By achieving high levels of retailer's economic and social satisfaction that leads to willingness to make sacrifices to accrue long-term benefits and maintain a positive relationship.The use of strong brands is very important and significant to the manufacturers as it maintains and strengthens the manufacturer-retailer economic relationship.

Figure 2 .
Figure 2. The Moderating effect of strength of manufacturer's brands on economic satisfaction-commitment relationship

Table 1 .
Loadings and cross loadings

Table 2 .
Result of convergent validity analysis

Table 4 .
Effect sizes of the latent variables

Table 5 .
Result of hypotheses testing