The sustainability of automotive supply chains in South Africa is important as this industry is substantially contributing 6.9% towards the country’s gross domestic product (AIEC 2018:5). Unfortunately, the automotive supplier relationships within the South African automotive supply chains seem to be strained, and this can result in elevated risk and possible non-sustainability (Naude, Ambe & Kling 2013:1; Sharma & Bhat 2014:45).

Supply chain theory emphasises the value of close buyer–seller relationships in the automotive industry (Cheng & Chen 2016:111). However, global pressures, coupled with suppliers that act opportunistically through late deliveries, poor service, quality defects and limited commitment, eventually culminate in strained relationships (Herko & Hanna 2017:87; Hiraoka 2013:2; Manzouri, Ab Rahman & Arshad 2015:85–86; Steinle, Schiele & Ernst 2014:124). These strained relationships result in a lack of trust between the supplier and the buyer (Steinle et al. 2014:124). Trust between the buyer and the seller is viewed as a vital part of successful automotive supply chain relationships, (Ambe 2014a:51; Wisner, Tan & Leong 2016:79). Trust has been motivated as vitally important within supply chain relationships (Chao et al. 2013:1060; Özer & Zheng 2017:3), not least because of the numerous benefits derived from it, which include sustainability (Chou, Techatassanasoontorn & Hung 2015:4).

In the absence of trust, the supplier might stand a chance of losing the contract with the buyer, thereby sacrificing sustainability (Tolmay 2018:80). Therefore, the benefits that derive from the presence of trust in the buyer–supplier supply chain relationship should not be underestimated (Vieira et al. 2013:265). Unfortunately, Naude et al. (2013:1) state that supplier conflict is hampering the sustainability of South African supply chains. The problem addressed in this article pertains to the erosion of trust because of conflict between suppliers in the South African automotive supply chains. This in return threatens to hamper sustainability of the supply chains (Chou et al. 2015:4). Because literature invites more research on trust (Akrout & Diallo 2017:159), specifically within supply chains (Batsaikhan 2017:3), the research reported on in this article investigated the research question: What are the antecedents of trust between buyer and seller within the South African automotive supply chains? The primary objective of the research was to determine the antecedents of trust, while the secondary objective was to utilise a path analysis to determine the antecedent(s) that have the strongest correlation with trust.

An investigation of ‘trust’ has become necessary as literature has suggested that trust leads towards sustainability within supply chains (Chou et al. 2015:4). The term ‘supply chain sustainability’ refers to the triple bottom line that includes the environmental, economic and social dimensions (Silva, Fritz & Nunes 2017:35). However, Silva et al. (2017:35) have added a fourth dimension to supply chain sustainability, namely governance, which encapsulates a holistic approach towards sustainability and includes strategies to enhance total quality management (TQM), risk management and supplier relationship management. Hence, sustainability, as referred to in this context, encapsulates the continued existence of automotive component suppliers, allowing them to undertake business for the foreseeable future.

Automotive component suppliers have limited parameters for differentiation through price, product quality and logistics (Yeh 2016:137). Therefore, suppliers must utilise the actual supply chain relationship with the buyer to differentiate themselves (Yeh 2016:137) and ensure sustainability through trust (Chao et al. 2013:1060; Özer & Zheng 2017:3). It is therefore important that automotive suppliers understand the importance of trust in buyer–supplier relationships as well as the governance thereof (Chao et al. 2013:1060; Özer & Zheng 2017:3).

Although literature has stated the benefits of trust in buyer–supplier relationships, more research relating to the conceptualisation of trust within automotive supply chains has been invited (Ashnai et al. 2016:137). This is because of trust being seen as a multifaceted and complex concept (Akrout & Diallo 2017:159; Eggert, Ulaga & Schultz 2006:20; Yaqub & Hussain 2013:436). Furthermore, literature has invited more research on trust in different environments and countries (Vieira et al. 2013:265), such as developing countries (Batsaikhan 2017:3; Niehaus, Feiboth & Goedhals-Gerber 2018:1), specifically within supply chains (Batsaikhan 2017:3).

Apart from the fact that further investigation has been called for by literature, the research reported on in this article was motivated by the desirable benefits, such as sustainability, that are derived from trust. The article is presented through the lens of social exchange theory (SET) (see Figure 1), and the role of trust within supply chain relationships is discussed, as well as the numerous benefits of trust, including sustainability. Finally, in an aim to better understand trust, the antecedents of trust are presented through a path analysis depicting the construct showing the strongest correlation (see Figure 1).

Relationships within the supply chains form part of the SET which originated from anthropology (Blau 1964; Cheng & Chen 2016:111). The SET covers many domains, including behavioural psychology, sociology, philosophy and economics (Chao et al. 2013:1059). Social exchange theory is particularly popular within the supply chain genre to evaluate buyer–supplier relationships and strategic alliances (Chao et al. 2013:1059). Vos, Pulles and Schiele (2016:3), for instance, emphasise the importance of positive buyer–supplier relationships and have suggested that trust between buyer and supplier forms the foundation of sustained social exchanges. According to Blau (1964), the relationship under SET will continue or terminate based on the ‘exchange’ experience between the two partners, be it positive or negative. According to Ashnai et al. (2016:129), trust forms the core of SET. Moreover, SET is explicitly popular within the supply chain genre to evaluate supply chain relationships that result in added value that benefits both parties (Chao et al. 2013:1059; Okdinawati, Simatupang & Sunitiyoso 2017:85).

The buyer–supplier relationship and associated aspects have been analysed in the literature from both the view of the buyer (customer) and the seller (supplier) (Ritter & Walter 2012:136). However, this article only focusses on the buyers’ perspective.

The study on which this article is based formed part of a larger quantitative empirical study. The study was explanatory and descriptive in nature, as it expected to determine relationships between constructs (Saunders, Lewis & Thornhill 2009:140). A deductive research strategy was followed to better understand a social science phenomenon (Saunders et al. 2009:125), in this case, identifying the most prominent antecedents of trust.

Before starting with the actual research, the researcher ensured that ethical compliance requirements were met in line with the ethics review committee of the applicable tertiary institution. The researcher submitted an application explaining the methodology and ethical compliance to the ethics review committee, and an ethical clearance certificate for the research study was issued.

In preparation for the actual research, the questionnaire was compiled by utilising the survey instrument designed by Ulaga and Eggert (2006:134) and Eggert et al. (2006:26) which is based on the three value drivers for relationships, namely, core offering, sourcing process and customer operations. A total of 16 items were utilised (see Table 3). The questionnaire utilised a seven-point measurement scale aligned with the seminal study by Morgan and Hunt (1994:35) and Eggert et al. (2006:26), where respondents had to indicate their perceptions – that formed the unit of analysis – using statements ranging from 1 = ‘strongly disagree’ to 7 = ‘strongly agree’. The questionnaire items from Morgan and Hunt (1994:35) (see Table 3) on trust and Eggert et al. (2006:26) (see Table 4) on value drivers were utilised, as presented by the respective studies.

Before distributing the questionnaire, a validation process was first conducted with three South African automotive industry experts. The purpose of this process was to test whether the questions from research previously undertaken in the United States (Morgan & Hunt 1994) and Germany (Eggert et al. 2006:26), respectively, were applicable to the South African automotive context. The validation process confirmed that the questions were stated clearly and were applicable to the South African automotive supply chains.

The questionnaires were then distributed to managers of all Tier 1 suppliers who are members of the National Association of Allied and Automotive Component Manufacturers (NAACAM). The Tier 1 supplier is the direct supplier to the OEM (original equipment manufacturer) such as Toyota. NAACAM was established in 1980, and 140 members represent the interests of the South African automotive component manufacturers. Almost 70% of South Africa’s first tier (Tier 1) component manufacturers are NAACAM members (AIEC 2015:7). An email invitation containing a link to the Web-based survey hosted by SurveyMonkey was sent to the sampling frame. Non-respondents were contacted telephonically, through a total of three waves over a 60-day period, and encouraged to participate in the research. The survey was directed at senior managers, namely, technical managers, procurement managers and chief executive officers. Of the potential 140 participants to the survey, 114 respondents participated. Hence, a response rate of 81.4% was achieved. However, not all respondents answered all the questions. The questionnaires received back from the respondents were cleaned and captured using SPSS 24.0. The findings are presented in Table 2.

According to demographical information obtained, a representative response spread was achieved. The respondents represented companies from all four provinces comprising automotive clusters, different company sizes, according to shareholding, turnover and number of employees per company. The respondents were also requested to indicate their position in the organisation. Of the 98 respondents, 51 (52.0%) held senior management or CEO positions, while 37 (37.8%) of them indicated that they were procurement or administrative managers, whilst 10 (10.2%) respondents stated that they were technical managers (see Table 2). The data analysis included an exploratory factor analysis (EFA), followed by a path analysis.

The research question under investigation was: What are the antecedents of trust between buyer and seller within the South African automotive supply chains? Therefore, trust (dependent variable) was established by utilising a three-item scale (see Table 3) obtained from the seminal research by Morgan and Hunt (1994). The focus of the study was on antecedents of trust, while trust, as depicted by Morgan and Hunt (1994), comprised one factor made up of three items (which is considered the minimum number of items in a scale and thus solid) (Castello & Osborne 2005:3; Henson & Roberts 2006:408).

The items relating to trust were (Table 3):

A high Cronbach’s alpha coefficient (0.917) was reported for the trust scale (see Table 3). This exceeds the generally agreed-upon lower limit of 0.70 for the Cronbach’s alpha coefficient for internal consistency, as recommended by Pallant (2011:303). It was confirmed that the respondents were in agreement with the importance of trust.

To assess the measurement model for the independent factors, as a first step, an EFA was conducted on the 16 items. The aim of this analysis was to identify the hypothetical construct factors emerging from the data and if these related to the original theorised factors and conceptual model. An orthogonal rotation method was utilised as the analytical procedure for the study, as it provides a separation of factors (Pallant 2011:185).

Inspection of the correlation matrix (see Table 5) produced from the EFA revealed that all the coefficients were higher than 0.5, motivating the suitability of the factor analysis (Pallant 2011:199). The Kaiser–Meyer–Olkin value was 0.898, which exceeded the recommended minimum value of 0.6 (Kaiser 1970, 1974), and the Bartletts’ test of Sphericity (Bartlett 1954) was statistically significant (p < 0.01), which further supported the factorability of the correlation matrix.

From the EFA, four factors were identified explaining a cumulative 72% of the variance in the overall scale response. Each of the items also showed strong loadings onto the identified factors (see Table 4). All items were therefore retained. The four factors were labelled as follows: product delivery and quality (F1), personal interaction (F2), product development support (F3) and improving time-to-market (F4) (see Table 4). The core offering, namely, product delivery and quality (F1), as well as the sourcing process with personal interaction (F2) corresponded with the factors of the study undertaken by Eggert et al. (2006). However, the customer operations function revealed two sub-factors rather than a single factor, namely, product development support (F3) and improving time-to-market (F4) (see Table 4).

A second phase of the model assessment involved conducting a confirmatory factor analysis (CFA). This phase more critically assesses the measurement model as it emerged from the EFA. The CFA model was found to be an acceptable fit supported by the following statistics: normed fit index (NFI) = 0.836 (Byrne 1994), Tucker Lewis index (TLI) = 0.847 (Hu & Bentler 1999), comparative fit index (CFI) = 0.889 (Bollen 1989), root mean square error of approximation (RMSEA) = 0.117 (Fan, Thompson & Wang 1999), PCLOSE (PCLOSE gives a test of close fit, while p gives a test for exact fit) = 0.000 and CMIN/DF (DMIN/DF is the minimum discrepancy divided by the degrees of freedom) = 2.605 (> 0.05) (Pallant 2011:176). Despite marginal fit indices, the factor loadings were higher than 0.8 on average (see Table 4).

Next, the Cronbach’s alpha coefficients were calculated to further assess the reliability of internal consistency of factors. Alpha values of above 0.8 were reported exceeding the minimum requirement level of 0.7 (Pallant 2011:6). The mean for each of the factors was henceforth calculated based on the summated average of all underlying items. This mean score for each factor served as the proxy measures of the factors and was utilised as the observed variables in the path analysis.

The first factor, product delivery and quality (F1) (core offering) relates to the suppliers’ technical performance that includes reliability, product quality and delivery performance. Not surprisingly, F1 was the most highly rated factor (mean of 5.791, SD = 0.988), as this is after all the prime reason why suppliers are appointed in the industry (see Table 4).

Secondly, personal interaction (F2) (sourcing process) (mean of 5.697, SD = 0.968) relays the personal relationship that the automotive supplier has with the buyer, and it includes aspects such as the sharing of information and how well the client (or buyer) is treated (see Table 4).

Thirdly, product development support (F3) (customer operations) (mean of 5.285, SD = 1.269) comprises the extent to which the supplier assists the buyer with the development of their products.

Fourthly and finally, improving time-to-market (F4) (customer operations) (mean of 5.044, SD = 1.270) relates to the extent to which the supplier assists the buyer with the speed of introducing products into the market (see Table 4). The respondents, in general, seemed to have less agreement with the last two factors (F3, F4). The standard deviation for the four factors revealed similar variations across the means.

The preliminary analysis furthermore revealed no serious violations of the assumptions of normality, homoscedasticity and linearity. The correlation coefficients calculated between factor scores (F1, F2, F3, F4) (see Table 5) were all lower than 0.9 and did not suggest a concern of multicollinearity (Pallant 2011:151).

Thereafter, a path analysis was conducted through AMOS version 24 software to determine the strength of the four factors (independent variables) as predictor variables, and their relationship in explaining the variation in trust (dependent variable). It was found that all correlations measured above 0.5, and are therefore considered strong (Pallant 2011:134). While these results provide insight into the strength of the underlying bivariate relationships, the aim of the study was to identify those variables that contribute to explaining the variation in trust from a multivariate perspective.

The results of the path analysis, with the applicable standardised regression coefficients (β coefficients), are depicted in Figure 2 as:

The relationships between the independent variables on the left of the path analysis (F1, F2, F3, F4) (see Figure 2) revealed acceptable discriminant validity (r < 0.90) (Pallant 2011:151), and this is indicative of sufficient independence between the factors.

The implications of the above findings are discussed in the following section.

Globalisation brought about various challenges for South African automotive component suppliers. One of the challenges is strained supplier relationships that diminish trust, which in turn threatens sustainability. The problem addressed in this article pertains to the erosion of trust because of conflict between buyers and suppliers in the South African automotive supply chains. As literature has suggested that trust can add value and enhance sustainability, the aim of this study was to answer the research question: What are the antecedents of trust between buyer and seller within the South African automotive supply chains?

The primary objective of the research was to determine the antecedents of trust. To accomplish this, literature was reviewed, and it was found that supply chain theory stresses the value of close buyer–seller relationships in automotive supply chains (Cheng & Chen 2016:111). According to Ulaga and Eggert (2006:128) and Eggert et al. (2006:21), value can be added through three drivers, namely, a core offering, the sourcing process and customer operations. Literature further states that supply chain buyer–supplier relationships with elevated value result in trust (Tolmay & Venter 2017:3; Wisner et al. 2016:79; Yeh 2016:140). The research reported on in this article utilised the three value drivers (core offering, the sourcing process and customer operations) and tested their relationship with trust as possible antecedents (Eggert et al. 2006:21).

The secondary objective was to determine the antecedent(s) showing the strongest relationship with trust within the South African automotive supply chains, through a path analysis. After processing the data from the survey, it was found that the respondents were in agreement with the importance of trust within the automotive supply chains. Heinrich et al. (2011:227) state that theoretically the value drivers identified by Eggert et al. (2006) will result in trust, but the author of this article could not find evidence that this model was tested specifically on trust. Thus, the research reported on in this article tested the relationship of four value-adding antecedents (Eggert et al. 2006) with trust. These include product quality and delivery (F1) as part of the core offering, personal interaction (F2) as part of the sourcing process, product development support (F3) and improving time-to-market (F4). The latter two antecedents both represent customer operations.

It was found that product quality and delivery (F1) (β = 0.56, p < 0.01), as part of the core offering, revealed the strongest coefficient with trust, followed by personal interaction (F2) (β = 0.44, p < 0.01), which forms part of the sourcing process (see Figure 2).

As product quality and delivery (F1) (as part of the core offering) revealed the strongest coefficient (or positive relationship) with trust, it is implied that the biggest enhancer of buyer–supplier trust is the timely delivery of products by the supplier to the buyer, according to the correct specifications and quality (see Table 4).

When Ulaga and Eggert (2006:13), Sun et al. (2014:80) and Tolmay and Venter (2017:6) tested the coefficient (or positive relationship) of the three value drivers (Eggert et al. 2006) on relationship value (and not trust), they found contradicting outcomes as the sourcing process revealed a higher coefficient with relationship value. All three studies revealed that the sourcing process seems to add the most value to relationship value, whereas the research reported on in this article found that the core offering (product quality and delivery) seems to be the largest contributor towards trust.

This might be the phenomenon because in a quality-conscious environment, such as the automotive supply chain, there is no room for defects, as they result in financial implications, which negatively influence the buyer and the seller. Failing to comply with product quality and delivery might result in the decrease of trust within the buyer–supplier relationship. Suppliers are advised to guarantee that product quality and delivery are consistent, and according to the requirements of the buyer in order to secure trust.

The second most important antecedent that enhances trust seems to be personal interaction (F2) (β = 0.44, p < 0.01) (as part of the sourcing process), which entails the knowledge of the supplier and also the ability to provide the buyer with appropriate information when required. Ambe (2014:50) and Vieira et al. (2013:266) also confirm the importance of knowledge and information sharing within automotive supply chains. F2 also encapsulates good working relationships with the buyer, and includes treating the buyer as being important (see Table 4). It suggests that if the personal interaction between the buyer and the seller is not enhanced, trust in the relationship will be hampered, which will result in negative consequences. Suppliers are encouraged to prioritise personal interaction activities as part of the buyer–supplier relationship in order to elevate trust.

The two factors that form part of customer operations, namely, support for product development (F3) and improving time-to-market (F4) did not reveal a significant correlation coefficient with trust. Both these factors represent research and development, and it can be argued that this function is usually undertaken by the core Tier 1 company (in the mother country), and is therefore not required by Tier 2 suppliers situated in South Africa. Therefore, these two factors do not seem to affect trust between the buyer and the seller.

Literature invites more research on the antecedents of trust, and this article makes valuable theoretical and managerial contributions. The theoretical contribution presents a better conceptualisation of trust, by depicting antecedents of trust specifically in automotive supply chains. In an industry characterised by strict quality requirements, it seems that product quality and delivery is the construct that contributed the most towards trust (F1) (as part of the core offering). Therefore, theoretically, buyer–supplier trust is enhanced when the supplier delivers products to the buyer on time, according to the correct specifications and quality.

The managerial implications include inviting managers to focus on product quality and delivery, as well as personal interaction, to enhance trust and ultimately to add value within supply chains. In line with theory promoted by Yeh (2016:138), automotive component suppliers might consider introducing a relationship governance strategy where trust, specifically, is incubated through product quality and delivery (core offering) and personal interaction activities (sourcing process). A relationship governance strategy focusses on how value can be added through the relational exchange with the client or the buyer. It is important that managers define the strategic objectives for the relational governance approach on how specifically to add value, that results in trust through product quality and delivery as well as personal interaction.

The failure to enhance trust through supplier relationships might result in increased risk, such as negative financial implications, which might hamper the sustainability of suppliers. As part of the relational governance strategy, a trust-centric approach is advised where both parties collaborate through the sharing of knowledge and innovation, and where both parties are willing to invest in each other by means of the sharing of expertise (know-how), while investing in the relationship and problem-solving.

The study invites more research, especially relating to the supply chains of various industries in different countries. Further research on the expansion of the conceptualisation of supplier relationships, trust and other relational sustainability enablers is invited. As a limitation, the research only approached Tier 1 companies (buyers) who are members of NAACAM, and the results cannot necessarily be generalised to all buyers in the South African or global automotive supply chains. In addition, the research reported on in this article only tested the one-way perception of buyers (Tier 1) regarding suppliers (Tier 2), and not that of suppliers regarding buyers. Future research into two-way perceptions might provide interesting results, such as perception asymmetry, and might deliver additional insightful findings.