AN EMPIRICAL ANALYSIS OF BUYER-SUPPLIER RELATIONS IN
HEALTHCARE CENTERS:
THE EFFECT OF DEMAND AND TECHNOLOGY UNCERTAINTY ON THE TRUST-PERFORMANCE AND COMMITMENT-PERFORMANCE RELATION
Reinee Huizinga Studentnumber: 1063243
University of Groningen Faculty of Business Administration
Msc. Operations and Supply Chain Management
Padangstraat 38 9715 CV Groningen
Tel: 06 24950105
E-mail: reineehuizinga@hotmail.com
AN EMPIRICAL ANALYSIS OF BUYER-SUPPLIER RELATIONS IN
HEALTHCARE CENTERS:
THE EFFECT OF DEMAND AND TECHNOLOGY UNCERTAINTY ON THE TRUST-PERFORMANCE AND COMMITMENT-PERFORMANCE RELATION
ABSTRACT
1. INTRODUCTION
This research focuses on the purchasing activities between various healthcare institutions and their suppliers in the Netherlands. Purchasing performance is an important determinant of a firm’s competitiveness (Noordewier, John, and Nevin, 1990, p. 80) and depends on how the supply management function coordinates relationships with external parties (Kaufmann & Carter, 2005, p. 653). The relationship between buyers and suppliers has been studied by several fields (see Claro et al., 2003, p. 703). Relation management is described as perhaps the most fragile and vague (Johnston et al., 2004, p. 24) and maintaining the relation is seen as an intimidating task (Chen & Paulraj, 2004, p. 122).
One observation that these fields have in common is that the relationship between partners has changed due to today’s competitive environment (e.g. Fynes et al., 2005, p. 3303; Heide & John, 1990 p. 260). Morgan and Hunt (1994, p. 34) identify commitment and trust as key mediating variables that contribute to overall network performance and have been viewed as a essential part of successful long-term relationships. Uncertainties can no longer be organized through hierarchical power, direct surveillance or detailed contracts (Edelenbos & Klijn, 2007, p. 26) and therefore trust and commitment as coordination mechanism becomes more important for coordinating buyer-suppler parties.
Many authors concentrate on trust and commitment and their relation upon performance, but environmental uncertainty is a critical predictor for these relations (see Noordewier et al., 1990; Fynes et al., 2005). Buyer-supplier relationships are described in terms of the uncertainty they face, and these authors state that the role of trust and commitment for deriving performance will differ in these relationships. This view is related to the contingency theory‘s assumptions (see: Gingsberg and Venkatraman 1985, p. 421).
buyer-supplier relations in service organizations (examples of articles that did empirical research in the service sector were articles by Doran, Thomas & Caldwell, 2005 and Eriksson & Sharma, 2003). Thereby, most of the authors that focus on business relations collect data either from the buyer (e.g. Kaufmann & carter, 2006, p. 653; Noordewier et al., 1990, p. 86) or the supplier (Fynes et al., 2005). This study combines these two perspectives and this has been identified as the dyadic perspective (Anderson & Narus, 1990, p.43). Johnston et al. (2004, p.35) call this a matched-paired approach. Our research focuses on single relations (a buyer firm and a supplying firm), and not on broader levels of analysis (as in e.g. Fynes et al., 2005).
The next paragraphs will elaborate on the concepts that were used for this project. Hypotheses will be formulated regarding the relationship of trust and commitment with performance. After that, the concept of uncertainty as well as its moderating effects will be explored and accompanying research questions will be formulated, including the conceptual model. The second section will focus on the instruments used for measurement and data analysis. The results of this analysis will be outlined in the third section. The last section will discuss upon these outcomes. This section will also discuss the shortcomings of this research as well as the possibilities for future research. The final paragraphs will present the implications for both research and management.
1.1 Trust and its relation with Performance
Although the concept has been studied on a regular basis, defining the concept is proved to be complicated and a universally accepted definition does not exist. In the rational choice based view of trust for example, actors are presumed to make rational, efficient choices. Researchers are critical towards this view (see: Kramer, 1999, p. 574) and state that trust must incorporate the social and relational foundations of trust-related choices in a more systematically way. This is why we adopted the image of trust as relational and not as rational choice behaviour.
In the relational choice perspective, trust is defined as a willingness to be vulnerable to the actions of another party based on the expectation that the other will perform a particular action important to the trustor, irrespective of the ability to monitor or control that party (Mayer et al., 1995, p. 712). This definition understands the trust concept as an internal state of the trustor with cognitive and affective components rather than an observable behavior (Riegelsberger et al., 2005: 386). Trust does not involve risk per se, but instead as a willingness to engage in risk-taking with the focal party (Mayer & Davies, 1999: 124). Trust is only a relevant factor in risky situations (Das & Teng, 1998, p. 494) and taking risk behaviours in the relation with the trustee is the outcome of trust (Mayer & Davies, 1999: 124).
toward another organization (Zaheer et al. p.143).
The trustor needs propensity to trust. Zaheer et al. (1998, p.143) characterize this as dispositional trust. Propensity is a personality characteristic (Das & Teng, 2004 p. 95) that is related to subjective trust. Propensity is a stable factor that can be understood as the general willingness to trust others and differs between people (Mayer et al., 1995, p. 715)1.
The characteristics of the trustee are covered using the concept of trustworthiness and this is not a stable factor as it differs between relations. This is also characterized as relational trust and focuses on a specific counterpart in the dyad (Zaheer et al., 1998, p.143). Mayer et al. (1995) claim that trustworthiness consists of three factors, earlier identified by Sako for inter-organizational trust (1991 and 1992) and these factors are ability, benevolence and integrity. These factors are not trust per se; these variables form the basis for the development of trust (Mayer et al., 1995, p. 709).
Ability is that group of skills, competencies and characteristics that enable a party to
have influence within some specific domain (Mayer et al., 1995, p. 717-718, Morgan & Hunt, 1994). Sako & Helper (1998, p. 388) refer to the term competence trust in this respect. Some authors use distinctive factors for ability, e.g. McKnight & Chervany (2000) who use the sub- constructs of competence, expertise and dynamism.
Integrity is understood as being the trustor’s perception that the trustee adheres to a set
of principles that the trustor finds acceptable (Mayer & Davis, 1995: 719). Sako & Helper (1998) refer to conformation to agreements and call this contractual trust. Riegelsberger et al. (2005: 399) use the term internalized norms for this factor of trustworthiness.
Benevolence reflects the will from a party to make unrestricted commitments and to
take initiatives for mutual benefit without taking advantage at the cost of its partners (Sako & Helper, 1998, p. 388). In the article by Sako & Helper, benevolence trust is referred to as
1 Some authors have doubts concerning the stability of this factor (e.g. Boon & Holmes, 1991: 196) but Das &
goodwill trust. It means that the trustee has some specific attachment to the trustor (Mayer et al., 1995, p. 718). Strong feelings of benevolence only develop in case of repeated episodes of trusting and fulfilling (Riegelsberger, 2005, p. 402). Ireland & Webb (2007, p. 484) mirror this and state that goodwill trust exists when partners are willing to act in ways exceeding specified contractual agreements.
The conceptualization of trust is not agreed upon in various disciplines2 (McKnight et al., 1998, p. 473) and conceptualizing trust is largely influenced by the context in which a study is conducted (Volery & Mensik, 1998, p. 989). Trustworthiness has also been conceptualized as consisting of two elements, honesty and benevolence (Kumar et al., 1995a, p.351). Morgan and Hunt (1994, p. 28) conceptualize trust as consisting of reliability, integrity and confidence. Zaheer et al. (1998) use confidence and predictability. McAllister (1995, p.25) use the dimensions affect- and cognition-based trust. For an extensive review on these issues, the article of Seppänen (2007) depicts the different dimensions of trust that were found in empirical research. The 1995 model of Mayer et al. has been used in distinct fields of research (Schoorman et al., p. 344) and therefore this conceptualisation is used for this research.
The relation between trust and performance has been researched theoretically, though empirical work on the link between trust and performance has been rare (Sako & Helper, 1998, p. 388 & Johnston et al., 2004). Trust is a significant predictor of positive performance within inter-organizational relationships and could also lead to a better overall buyers’ and suppliers’ performance (e.g. Johnston et al., 2004; Carr and Pearson, 1999;
2 social psychology (Rempel et al., 1985; Rotter, 1971), sociology (Shapiro, 1987), communication (Berlo et al.,
Ireland et al., 2007; Volery and Mensik, 1998; Sako & Helper, pg, 388 and Fynes et al., 2005). Theoretical evidence of this extends across multiple theories of organization research, such as organization sociology and business administration, transaction cost theory and social capital (Ireland & Webb, 2007, p. 483; Edelenbos & Klijn, p.26).
Trust has been characterized as an important psychological condition in inter- organizational relationships as it makes cooperation easier (Kanter, 1994) and these cooperative behaviors lead to higher perceived performance (Johnston et al., 2004, p. 36). It further reduces the need for contractual safeguards (Parkhe, 1993), lowers transaction costs in alliances (Das & Teng, 2001: 454; Sako & Helper, 1998, p. 388) and will result in a lesser need for vertical integration (Sako & Helper, 1998: 389). Because we are interested in the effect of trustworthiness alone and not in the resulting behavioural consequences, this research predicts a direct effect of trust on performance.
In our research, the performance of the supplier was measured from the perspective of both the buyer and the supplier. Performance was not measured using quantitative data; performance developments and their importance were measured using qualitative statements of both the supplier and the buyer. Our performance measure included both service and cost performance indicators, which stems from operations research where performance has been treated as a composite construct (e.g. Narasimham & Das, 2001, p. 596 & Johnston et al., 2004). Trust is also measured for buyer and supplier perspectives, but does not only focus on the supplier. The buyer rated the amount of trust the respondent has in the supplier. The supplier rated the amount of trust he had in the buyer. In short: we measure the influence of mutual trust on perceived supplier performance. The hypothesis regarding this relation is:
1a: Higher mutual buyer-supplier trust has a positive relation with perceived supplier
1.2 Commitment and its relation with performance
Commitment is defined as a channel member’s intention to continue the relationship (e.g. Geyskens et al., 1996, p. 304). Commitment implies a willingness to make short-term sacrifices to realize long-term benefits (Dwyer et al., 1987). This long-term orientation is based on an assumption that the relationship is stable (Anderson & Weitz, 1992, p.19). Meyer & Allen (1991) proposed a conceptualization that incorporates both attitudinal and behavioral approaches and is identified as multi-component approach. This conceptualization consists of three distinct factors, but relate to the intra-organizational form of commitment. Gundlach et al. (1995 and also Kumar et al., 1995b) look at the concept from an inter-organizational (or: exchange) view and also define a three-component construct for measuring commitment. The three factors of Meyer & Allen (1991) were used for our research and these authors have conceptualized commitment as consisting of affective, normative and continuance commitment.
Affective commitment refers to the partner’s emotional attachment to, identification
with, and involvement in the partner organization (Meyer & Allen, 1991, p. 67). It reflects the desire to continue the relationship because of positive affect toward the partner (Kumar et al., 1995b). The relationship is continued because they want this (Meyer & Allen, 1991, p. 67).
Continuance commitment is solely based on economic or extrinsic needs (Mavondo &
Rodrigo, 2001, p. 112). They state that commitment possesses an input or instrumental component (Gundlach et al., 1995, p. 79). Kumar et al. (1995b) use the term willingness to
invest to characterize instrumental commitment. This form of commitment is shown whenever
Normative commitment reflects a feeling of obligation to continue the relation and
partners feel that they ought to buy goods from or supply goods to the organization (adapted from Meyer & Allen, 1991, p.67). It derives from obligations caused by events before or after joining the partner. Meyer and Allen (1984) initially proposed that a distinction be made between affective and continuance commitment, but later suggested the existence of this third distinguishable component of commitment (Meyer et al., 2002, p. 21).
There are different research topics that use other factors for commitment. For example, channel commitment research has focused on one of two components of commitment, calculative and affective commitment (Geyskens et al., 1996, p. 304), like the older two-factor model for commitment proposed by Meyer & Allen (1984). Affective commitment is based on general positive and calculative commitment is based on a general negative motivation toward the exchange partner (De Ruyter et al., 2001, p. 272).
The commitment variable is also used differently in empirical and conceptual research models. It has been used as dependent and independent variable in empirical research (Reichers, 1985, p.465) and theoretical research (Swailes, 2002, p.156). It has also served as the dependent variable in many models including buyer–seller relationships (Mavondo & Rodrigo, 2001, p. 112) and channel theory as it is a good indicator of long-term relationships. In this research it is used as an independent variable3.
The commitment-performance relation has been researched, though the evidence for a strong link between commitment and performance is inconsistent (Swailes, 2002). Robson et al. (2006, p. 601) performed a literature-study in the context of international strategic alliances (ISA). 60% of the papers that he analyzed found a direct relation between commitment and performance.
3 Swailes (2002) directs more attention to the history of the concept and its factors and a further explanation will
From the perspective of transaction-cost analysis, performance is improved because commitment has been shown to reduce costs of opportunism and conflict between partners (Wu & Cavusgil, 2005, p. 81). From the resource-based view it can be derived that firm-specific resources can be integrated through collaboration. This will make both parties able to access complementary knowledge and skills that the company does not have or can not afford (Wu & Cavusgil. 2005, p. 87). When members develop commitment to their partners they are likely to spend more energy into achieving mutual goals (Clarke, 2006, p. 1186) Commitment is thus seen as a key attribute in order to achieve valuable outcomes (Morgan & Hunt, 1992, p. 23).
By measuring the direct relation between commitment and performance, we can analyse its direct impact upon performance. As was the case for trust, mutual commitment will be measured from the perspective of both the buyer and supplier. The same measure for performance will be used as the one used to analyze hypothesis 1a. Our hypothesis for the relation between commitment and performance is:
1b: Higher mutual buyer-supplier commitment has a positive relation with perceived
supplier performance.
1.3 Uncertainty and its moderating effect
to as contextual uncertainty (Eriksson & Sharma, 2003, p. 962).
Child (1972, p. 3) conceptualizes the environmental uncertainty construct as consisting of the frequency of changes in relevant environmental activities, the degree of difference involved at each stage (instability by Krishnan et al.,2005, p. 898) and the degree of irregularity in the overall pattern of change (unpredictability by Krishnan et al., 2006, p. 898). The model of environmental uncertainty proposed by Chen and Paulraj (2003, p.123) partially fitted the objectives of this study. They consider uncertainty in supply, demand and technology. Supply uncertainty is not included in the conceptual analysis in our study. Our analysis focuses on the downstream process (from supply to demand) and uncertainties in supply are not the issue. Environmental uncertainty is thus split up in both demand and technological uncertainty in our conceptualization.
Demand uncertainty is an important factor because firms face varying degrees of
unanticipated changes in their forecasted volume required and the mix of items needed by the firm (Noordewier et al., p. 82). Heide & John (1990, p. 28) characterize volume unpredictability as the inability to make an accurate forecast of volume requirements in the relationship. The construct for demand uncertainty in our research is adapted from Chen & Paulraj (2004, p. 123), because they also included uncertainty in mix requirements in their model. Our model thus measures demand uncertainty in terms of fluctuations (in volume) and variations in demand (mix) (Chen & Paulraj, 2004, p. 123).
Technological uncertainty is used to measure the rate of technological changes evident
instrument focuses more on the uncertainty caused by instability of technological developments, as used by Chen & Paulraj (2004) and Jaworski & Kohli (1993).
The moderating effect of uncertainty in both demand and technology is predicted to influence the relation of trust and commitment with performance. The relation of trust and commitment with performance is predicted as being dependent upon contingency conditions4 and a shift is proposed from the “is” question to the “when” question (Sharma and Patterson, 2000, p. 471). Some authors have already looked upon this phenomenon, but they have generated different outcomes. The main cause for this is that authors use different conceptualizations of their variables and research has been performed in different business situations (see Fynes et al., 2005; Noordewier et al., 1990; Krishnan et al., 2006 & Wu & Cavusgil, 2005). Because of these differences between studies, research questions instead of hypotheses were formulated to study the moderating effect of demand and technological uncertainty in our healthcare context.
Fynes et al. (2005) posit that firms functioning in a highly competitive environment have a greater need for close SC relationships than firms in stable markets (2005, p. 3306). Relational quality is used as a composite measure for relational variables and also includes trust as one of the measures of this construct. Noordewier et al. (1990, p. 85) refer to the term
relationalism to detect the location of a firm’s relation on the continuum with market and
hierarchies as polar extremes. They state that when there is high environmental uncertainty, a higher level of relational governance is present in that relation. Performance is thought to improve when more relational structures are introduced in response to high uncertainty (Noordewier, 1990, p. 85).
4 Contingency conditions could for example be the size of the firm (e.g. Wu & Cavusgil, 2005, p. 84) or
In a study by Krishnan et al. (2006) trust has been used as a distinct independent variable. Krishnan et al. (2006, p. 897) predict a weaker relation between trust and performance when environmental uncertainty (predictability and instability of the market) is high than when it is low. Inter-organizational trust results in inadequate responses to the challenges posed by an uncertain environment (Krishnan et al., 2006, p. 898). Thus, under high rates of environmental uncertainty, one group of theorists expects the trust-performance to strengthen (Noordewier et al., 1990 & Fynes et al., 2005) and another group expects the trust-performance relation to weaken (e.g. Krishnan et al., 2006). Environmental uncertainty is split up in demand and technological uncertainty in our conceptualization. The research question concerning these moderating effects of demand uncertainty on the trust-performance relation is:
2a: Does a high rate of demand uncertainty have an effect upon the relation between trust and performance of the supplier and does it make the trust-performance relation stronger or weaker?
2b: Does a high rate of technological uncertainty have an effect upon the relation between trust and performance of the supplier and does it make the trust-performance relation stronger or weaker?
The article of Wu & Cavusgil addressed uncertainty and the commitment-performance relation. The authors generate results that indicate that under high market uncertainty,5 the positive effect of commitment on alliance performance is strengthened. The reason for this effect is that commitment allows partners to work closely in modifying the alliance scope and address changing market needs by developing mechanisms for adaptation (Wu & Cavusgil, 2006, p.84). Heide & John (1990, p. 28) 6 also found that perceptions of volume
unpredictability increase expectations of continuity. To do so, the expectations of continuity should be increased by taking reprisals against opportunism.
Both arguments predict a stronger relation between commitment and performance with higher rates of demand uncertainty, but a weaker relation between commitment and performance could also possible in situations of high demand uncertainty. For instance, maybe institutions wish to be flexible in relations to their suppliers and cope with demand uncertainty by not choosing for long-term relationships but instead use multiple suppliers to address the uncertainties in demand. This could be the case when demand uncertainty is so high that organizations can not rely upon only one organization to meet both volume and mix demands7. The research question concerning these moderating effects of demand uncertainty
on the commitment-performance relation is:
3a: Does a high rate of demand uncertainty have an effect upon the relation between
5 This research used the measurement model for uncertainty developed by Jaworski & Kohli (1993, p. 59-60)
The items for the market turbulence scale assessed the extent to which the composition and preferences of an organization's customers tended to change over time and this is different from our measure for demand uncertainty. Technological turbulence items tapped the extent to which technology in an industry was in a state of flux. This measure mirrors our technological uncertainty construct.
commitment and performance of the supplier and does it make the commitment-performance relation stronger or weaker?
In respect of the commitment-performance relation in situations of technological uncertainty, authors generated different results as well. Drawing from the resource dependency approach, buyers may want to secure the flow of their inputs to their organizations under critical conditions by forming long lasting relationships8 (Kaufmann et al., 2006, p. 657; Heide & John, 1990, p. 28). This argument predicts that under high technological uncertainty the commitment-performance will be stronger.
The relation between commitment and performance is also predicted to be weaker under conditions of high technological uncertainty. Kaufmann & Carter state that in dynamic situations it may be advantageous not to have entered into long-term relationships, in order to maintain flexible to capitalize into long-term relationships (2006, p. 669). Thus, under high technological uncertainty, the effect of commitment on alliance performance is weakened (Wu & Savusgil, 2006, p.85). In the results of both Kaufmann & Carter and Wu & Cavusgil a significant influence on the commitment-performance relation was not found. The research question is:
3b: Does a high rate of technological uncertainty have an effect upon the relation between
commitment and performance of the supplier and does it make the commitment-performance relation stronger or weaker?
The research model for this project is depicted in figure 2.1 below.
FIGURE 2.1
Research Model for this Study
Mutual trust Buyer-supplier
Dyad
Mutual commitment
Demand uncertainty Technological
2. METHODOLOGY
2.1 Data Collection and Procedure
This study is related to a PhD project by the Faculty of Business Administration of the University of Groningen (RuG). The hypotheses and research questions stated in the previous section will be tested using a questionnaire developed by this faculty. For this study, 45 chain relationships or dyads filled out a questionnaire9. A dyad consisted of a buyer and a supplier in the health sector and we are reasoning from a dyadic perspective (see Anderson & Narus, 1990, p. 43). Our research has its focus on the relationship between one buyer (with more respondents) and one supplier (with one respondent) and therefore, two different questionnaires were developed. They both contained the same, though reversely stated questions. The questionnaire was adapted to fit the context of the different health organizations that participated.
Data collection consisted of three phases. In the first phase health care institutions were identified by searching in databases on the websites10. In the second phase, the
purchasing managers of these Health Centers were contacted by phone. This was done to ensure that the person that completed the questionnaire was knowledgeable of and responsible for the relationship (Noordewier et al., 1990, p. 86). In that first conversation the purpose and value of the research and the operating procedure were explained shortly. If interested, supplementary information was sent by mail in order to provide the manager with more complete information. Subsequently, the managers received a second phone call one week later to ask if they were willing to participate in this project.
When the institution was willing to cooperate, the third and last phase began. The purchasing manager of the health care organization was asked to select three employees from
9 In the final analysis 38 relations were included because of outliers in regression analysis. This will be
highlighted in the result section
his own company. The three employees could be from another department than purchasing alone, as long as there was interaction between this respondent and the selected supplier. Respondents were able to participate when there was regular interaction between them and the supplier. The reason for using more informants is that we try to prevent single informant bias and by doing this, we generate more valid estimates of organizational properties (Anderson & Narus, 1990, p. 43).
The purchasing managers were also asked to select a supplier and were asked to recommend the research project to this supplier. The supplier that was selected by the health institution had to meet certain criteria. The supplier should provide goods and not services. Thereby, it should not be a company that supplies investment goods. The supplier should be a company that delivers products used in the process of caring for people, e.g. diapers, napkins and other products necessary for providing medical care and cure. The purchasing managers of the health institutions needed to provide us with the mail address and telephone number of the supplier’s respondent. The inclusion of a supplier was strictly necessary for this study, because the focus is on relations.
2.2 Respondents
The data were obtained from different types of health institutions, namely hospitals, care centers and rehabilitation centers. In total, 228 questionnaires were sent to 57 healthcare institutions and 153 were returned. This results in a response rate of 67%. Our respondents were mainly key respondents that were knowledgeable about their purchasing activities thus able to complete the questionnaire.
The number of employees that worked at health institutions and their suppliers ranges between 12 and 8000 and the mean number of employees is 1409 (SD = 1684,22). The distance between the health institutions and their suppliers varies between 1 and 1200 kilometers and the average distance is 136 kilometers (SD = 154.21). The length of the relation ranged between 1 and 132 years, with an average of 12.65 years (SD = 6.04). The dataset consisted of 83 male respondents and 45 female respondents. The ages of the respondents vary between 25 and 58. The average age of the respondents is 42 years (SD = 7.73).
2.3 Measures
The variable Trust was assessed using 12 items, based on Mayer and Davis (1999) and Cummings and Bromiley (1996). An exemplar question related to this construct is: “The operations of this supplying organization are always in line with arrangements made.” The questions that were used in Mayer et al., 1999 were included so all three factors of trust could be analysed independently. The instrument for measuring trust, developed by Cummings and Bromiley (1996), is the Organizational Trust Inventory (OTI). One Item of this instrument was incorporated, but we did not use the distinctive forms of trust that Cummings and Bromiley formulated in their instrument.
Commitment was measured using a twelve-item scale, based on Meyer and Allen (1991) and Kumar (1995). One question to measure this construct was: “We enjoy working with this organization and we would like to deliver goods in the future as well.” For this research we chose the article by Meyer and Allen, (1991: 67) where three different factors for commitment were identified. These items also consisted of questions used by Kumar et al. (1995) and this was especially the case for the items these authors used for measuring affective commitment. One of the items of Kumar et al. that we used is: “Our positive idea about this institution is an important reason to buy from this organization in the future.”
Technological uncertainty was measured using a five-item scale based on Chen & Paulraj (2004) and Heide & John (1988). In terms of Kaufmann (2006, p. 655) dynamism is included in our measure, but complexity is not. An exemplary question with respect to this measure is: “The goods our organization buys are characterized by a lot of changes in specifications.”
Performance was measured using eight items, adapted from Johnston et al. (2004). The items represented eight performance goals. The goals for this research were price, reliability, speed, mix flexibility, volume flexibility, costs, quality and innovation. Respondents first had to indicate the importance of eight provided performance goals on a five point Likert scale, rated from very unimportant to very important. Then, the interviewees had to indicate the improvements that were realized in respect of these eight performance goals, in a period of one year. The five-point scale rated improvements from improvements
are absolutely realized to absolutely no improvements were realized. The ultimate measure
for performance resulted from multiplying the importance and improvements scores of the performance goals.
2.4 Data Analysis
In order to test the hypotheses stated in the previous section, a hierarchical multiple regression analysis was executed. The dependent variable performance was not standardized. The analysis was done in three steps. In the first step the control variables were entered.
For our study the variables Length of Relation and level of conflict were selected as covariates. This study controls for this potential influence of relation length on performance, because prior history in the relationship could be an alternative explanation for our generated findings (Mohr et al., 1996, p. 110). The article of Dwyer et al. (1987) posits that in the course of time relations tend to evolve through different stages of relational development. We wanted to control for this effect upon our results.
Conflicts are also controlled for, because if conflicts are frequent and common, the performance of the supply relationship tends to be lower because the efforts of parties are focussed on resolving the conflict (Zaheer et al., 1998, p.146; Mohr et al., 1996, p.110). The two control variables have both a negative and a positive effect upon performance. Here it is stated that, over a longer course of time conflicts could also have more chance to appear and hence, can be of negative influence on performance.
In the second step the main effects of trust and commitment upon performance were entered. The last step involved the cross-product term of uncertainty (in demand and technology) and the main effects trust and commitment. This resulted in six interactions that were analyzed using six hierarchical regression models.11
11 See results section: 1. trust and demand uncertainty, 2. trust and technological uncertainty, 3. commitment and
3. RESULTS
3.1 Results for factor analysis
In table 3.1 the questions in the questionnaire for each variable as well as their loadings on each factor are depicted. These loadings were obtained using factor analysis with varimax rotation. For all variables recoding was not necessary, because propositions were not stated in opposite directions. For the trust variable, all the items loaded on one factor. The items measuring trust had a Cronbach’s alpha ( ) reliability value of .92 and no items were deleted. The three sub dimensions identified for trust were not found in the data set.
Commitment loaded on two factors; affective, normative commitment (further: affective commitment) loaded on one factor ( = .92) and continuance commitment loaded on the second factor ( =.69). The three factors for commitment identified in literature were not found in this dataset, but instead a two-component model for commitment resulted. The factor analysis resulted in a somewhat positive factor for commitment (affective, normative) reflecting a desire to maintain the relationship (Geyskens et al., 1996, p. 304). Also a more negative factor for commitment was found, reflecting a need to maintain a relationship (continuance, or calculative commitment in Geyskens et al., 1996, p. 304).
TABLE 3.1
Results of Component Analysis in SPSS of Demand Uncertainty, Technological Uncertainty, Trust and Commitment
factors
Items (from buyer perspective) 1 2 3 4 5
Demand uncertainty (alpha= .76)
The total amount of goods that this health institution buys has considerate weekly variations. -.05 -.02 .13 .69 .08
The mix of goods that this health organization buys has considerate weekly variations. -.10 .00 -.00 .73 .04
This health institution holds inventory in order to cope with variations in demand (DELETED). .04 -.14 -.20 .30 .26
The number of demanded goods of this institution is hard to forecast. -.21 -.05 .07 .71 -.05
The needed mix of goods for this health organization is hard to forecast .09 -.10 .19 .73 -.05
Technological uncertainty (alpha= .84)
The goods our organization buys are characterized by a lot of changes in specifications. .03 -.02 .74 .42 -.05
Our company has to change specifications in order to stay competitive. -.07 .07 .79 .04 .08
The speed of innovation is high in the industry that produces these goods. .05 .10 .76 -.12 -.11
The specifications of these goods change regularly. -.04 .03 .85 .16 -.16
The Product Life Cycle for these goods is short (DELETED). .07 .07 .47 .08 -.02
Trust (alpha= .92)
This supplier is competent in all things they do .80 -.05 .04 -.14 .03
This supplier is known for being successful in their work .72 -.18 .09 -.14 -.09
This supplier has a lot of knowledge concerning the work that has to be done .77 .02 .09 .01 .01
Our organization is very sure about how competent the health organization is. .75 .14 .07 -.12 .06
The workforce of this supplier is committed with respect to the performance of this organization. .72 .06 -.12 .06 .09
Personnel of this supplier is always concerned with what we think is important. .68 -.08 -.07 .23 -.06
Personnel of this supplier will do everything to help us. .75 -.20 .10 .15 .04
Personnel of this supplier will not do things that could be non-beneficial for our organization. .60 .17 -.07 -.10 -.26
This supplier will always comfort to liabilities. .75 .03 -.14 -.01 -.12
We never have to doubt whether this supplier comforts to arrangements made. .72 .34 -.02 -.02 -.12
This supplier never tries to break with liabilities. .75 .21 .04 -.23 -.20
TABLE 3.1 continued
Results of Component Analysis in SPSS of Demand Uncertainty, Technological Uncertainty, Trust and Commitment Factors
Items 1 2 3 4 5
Affective commitment (alpha=.92 )
It feels comfortable to work with this organization, so we will also work together in the future. .19 .79 .00 .06 .00
We want to keep this supplier because we enjoy to work together with this organization. .13 .84 -.07 .00 -.08
Our positive ideas about this institution are an important reason to buy from this organization in the future. .12 .72 .05 .16 -.07
We enjoy working with this organization and we would like to deliver goods in the future as well. .16 .85 .01 -.11 -.01
We would feel guilty if we would buy from another supplier. -.050 .71 .02 -.22 .28
We think it is inconsiderate to buy from another company, even it would be beneficial to our own organization. -.08 .82 .10 -.13 -.01 Even if we would get a better deal from another organization, we would find it inconsiderate to buy from another company. -.05 .82 .17 -.09 .04
We think it is important to be loyal to this supplier. -.02 .83 .13 -.08 -.03
Continuance Commitment (alpha=.69)
It would cost a lot of time, money and effort to find other suppliers. That is why we keep buying from this organization -.11 .17 -.12 -.14 .48
We are obliged to buy goods from this organization because there are no alternatives. .00 -.17 .02 .06 .87
Due to a lack of alternatives it is hard to change to another supplier and that is the reason for buying from this organization. -.10 -.14 -.01 .12 .84 We have done considerable investments in this organization and that is the reason to continue this relation. -.17 .24 -.11 .03 .61
Eigenvalue 7.33 5.30 3.37 2.48 1.93
TABLE 3.2
Results of principal Components Analysis of Performance
Factor
Items 1
Performance (alpha= ,870 )
Price of goods 0.63
High supply dependability 0.78
High supply speed 0.77
Volume flexibility 0.72 Mix flexibility 0.75 Low costs 0.73 Quality of goods 0.73 Innovation of goods 0.69 Eigenvalue 4.22
Percentage of variance explained 52.69
3.2 Correlations and descriptive statistics
In table 3.3 the means, standard deviations and correlations between the variables are depicted. Positive and significant correlations were generated between performance and the main effects trust (r = .40, p<.01) and affective commitment (r = .43, p<.01). The variable continuance commitment had a correlation with performance that was close to zero (r = - .05, n.s.).
3.3 Testing the Hypotheses and research questions
The hypotheses and research questions are tested using the results of the multiple hierarchical regression analysis (Cohen & Cohen, 1983). Seven relationship dyads were canceled during the analysis. In regression analysis, standardized residuals were compared with the predicted values (Janssen & Van Yperen, 2004, p.376). Relation 9, 21, 34, 63, 66, 68, and 69 were detected as being outliers for performance and were left out of the regression analyses.
In Figure 3.1 depicts the results for the six models that were analyzed. In the first regression step the control variables were tested and did not have significant associations with performance in all of the models experimented with.
In the second step the analysis focused on the first set of hypotheses. Significant results can be observed for the association between trust and performance in the first (b= .61,
p<.05) and the second model (b=.68, p<.05). In these models, higher trust relates to higher performance. No significant results were found for the relation between affective commitment
TABLE 3.3
Univariate Statistics and Pearson Correlations Among the Variables
Variable Mean s.d. 1 2 3 4 5 6 7 1. Length of relation 13.2 8,06 2. Conflicts 1.64 .58 .13 3. Trust 3.81 .57 .03 -.38 ** 4. Affective commitment 3.12 .99 .05 -.34 ** .13 5. Continuance commitment 1.90 .70 -.00 .28** -.27** .05 6. Demand uncertainty 2.43 .67 -.03 .10 -.15 -.14 .03 7. Technological uncertainty 2.21 .76 .06 .18 .01 .03 -.09 .12 8. Performance 14.46 1.82 -.05 -.10 .40** .43** -.05 -.24* * .25**
* = Correlation is significant at the 0.05 level (2-tailed). **= Correlation is significant at the 0.01 level (2-tailed).
in model 3 (b =.41, n.s.) and model 4 (b =.34, n.s.). The same holds true when continuance commitment was tested in model 5 (b = -.05, n.s.) and model 6 (b =.07, n.s.) Trust has significant associations with performance, but affective commitment and continuance commitment do not. Hypothesis 1a will be accepted, but hypothesis 1b will be rejected.
Research questions 2ab and 3ab are answered by studying the moderating effects of demand (2a & 3a) and technological uncertainty (2b and 3b). The moderating effect of demand uncertainty on the trust-performance relation was tested in the first model (see table 3.4) and generated no significant interaction effects (b = -.32, n.s.).
Significant interaction effects were found when technology uncertainty was tested (b = -.69, R² = .16, p<.05) in model 2. The moderating effect of technological uncertainty on the affective commitment-performance relation is depicted in figure 3.1. High technological uncertainty will have a very small effect on the relation between performance and trust. There is a stronger relation between trust and performance when technological uncertainty is low.
When the moderating effect of demand uncertainty was tested between affective commitment and performance, no significant moderating effect was found (b =. 08, n.s.).The moderating effect of technological uncertainty did generate significant results (b = 1.63,
p<.01) had significant values. The adjusted R² had a lower significance (adj. R² = .13, p<0.1), but this significance was sufficient for the adjusted R². In order to answer the research question 2b the regression plots that are depicted in figure 3.2 were examined. In this model, it can be seen that the effects are different from the technology uncertainty and trust model. When technological uncertainty is higher, the relation between affective commitment and performance becomes stronger. In case of low technological uncertainty the relation between affective commitment and performance weakens.
significant ( R² = .01; adj. R² = .03, n.s.). The same holds true for testing model 6 were no significant moderating effect was found (b = -.20, n.s.). Model 6 had negative figures for adjusted R² in the third step ( R² = .04; adj. R² = -.10, n.s.). For research questions 3 a & b this means that there is no significant effect of demand and technological uncertainty for the relation between continuance commitment and performance.
TABLE 3.4
Results of Hierarchical Multiple Regression Analysis
Perceived performance
Step and Variables 1 2 3
Model 1 TR*DU:
1. Length of relation .02 -.09 -.11
Conflicts -.13 .16 .17
2. Trust .61* .68*
Demand uncertainty -.68* -.71*
3. Trust × demand uncertainty -.32
R² .00 .25* .03 Adjusted R² -.05 .17* .17* Model 2 TR*TU: 1. Length of relation .02 -.07 -.06 Conflicts -.13 .19 -.04 2. Trust .68* .68* Technological uncertainty -.04 .16
3 Trust × technological uncertainty -.69*
R² .01 .13 .16* Adjusted R² -.05 -.03 .18* Model 3: CO1*DU 1. Length of relation .02 .03 .04 Conflicts -.13 -.06 -.06 2. Affective commitment .41 .40 Demand uncertainty -.78* -.76*
3 Aff. Commitment × demand uncertainty .08
R² .01 .20* .00 Adjusted R² -.05 .11 .08 Model 4: CO1*TU 1. Length of relation -.02 .05 .01 Conflicts -.13 -.10 -.09 2. Affective commitment .34 .53 Technological uncertainty .09 -.29
3 Aff. commitment × technology uncertainty 1.63**
R² .01 .04 .20**
Adjusted R² -.05 -.06 .13†
TABLE 3.4 continued
Results of Hierarchical Multiple Regression Analysis
Performance
Step and Variables 1 2 3
Model 5: CO2*DU
1. Length of relation .02 -.02 -.02
Conflicts -.13 -.11 -.08
2. Cont. commitment -.05 -.06
Demand uncertainty -.74* -.71*
3 Cont. commitment × demand uncertainty .32
R² .01 .14 .01 Adjusted R² -.05 .05 .03 Model 6: CO2*TU 1. Length of relation -.04 -.04 -.05 Conflicts -.14 -.19 -.10 2. Cont. commitment -.08 -.06 Technological uncertainty -.11 -.02
3 Cont. commitment × technological uncertainty -.38
2-way Interaction of technological Uncertainty and trust 10,0 11,0 12,0 13,0 14,0 15,0 16,0 17,0 18,0 19,0 20,0 Low High pe rf or m an ce TU high TU low Trust
2-way Interaction of technological Uncertainty and Commitment1
10,0 11,0 12,0 13,0 14,0 15,0 16,0 17,0 18,0 19,0 20,0 Low High pe rf or m an ce TU high TU low affective commitment FIGURE 3.1
4. DISCUSSION
The first set of hypotheses analyzed the trust-performance, the affective commitment-performance and the continuance-commitment-performance relations. The first hypothesis predicted a positive relation between trust and performance. This direct positive relation between trust and performance was found and the hypothesis was accepted. This mirrors the results of different studies that analyzed this relation (e.g. Zaheer et al., 1998; Fynes et al., 2005 & Krishnan et al., 2006). Johnston (2003, p. 25) is critical towards the direct link between trust and performance because behaviours are intervening between trust and outcomes. The positive relation between trust and performance proves that trustworthiness of the partner can lead to higher performance as proved by both the dependence and TCA theory12.
The positive relation between our commitment factors and performance was not found in our study and different and this could be caused by different factors. Factors like geographic location and business type had strong effects upon the commitment-performance relation (Robson et al., 2006, p.599). Clarke (2006, p. 1195) posits that research fails to satisfactorily operationalise the multidimensional nature of commitment and Robson also identified this feature as a cause for inconsistent results with respect to the commitment-performance relation (Robson et al., 2006, p. 600).
In our conceptualization, commitment was split up in the factors affective and continuance commitment. Actually, the affective commitment variable consisted of both affective and normative commitment. Clark (2006, p. 1195) found that performance has a stronger relation with the affective commitment variable. This author found that normative commitment has a positive relation upon shared values and not upon performance. The normative component that was included in our affective commitment variable could be of
influence on the insignificant outcomes with respect to the relation between commitment and performance.13
The relation between affective commitment and performance was stronger than for the continuance commitment variable. Results of Swailes (2002, p. 166) indicate that not-for-profit organizations rely heavily on satisfaction and affective commitment since there is less opportunity than in the profit-seeking sector to stimulate continuance commitment.
Our research questions concerned the effect of demand and technological uncertainty on the relation of our behavioral variables in combination with performance. In the trust-performance relation, trust in high technological uncertainty leads to neither stronger nor weaker performance and instead low technological uncertainty strengthens the trust-performance relation. Trust does not seem to be an effective coordination mechanism between partners when technological uncertainty (environmental uncertainty in Krishnan, 2006) is high. Trust functions as a simplifying mechanism that may produce biases (McEvily, 2003, p. 100). Trust can also limit knowledge exchange or can be costly when it causes strategic blindness (McEvily, 2003, p.100, Krishnan et al., p. 898). Edelenbos & Klijn (2007, p. 34) also point out that there is a risk of trust. This is especially the case when there is too much trust, which can lead to unreasonable relaxed attitudes from partners.
In our sample, in situations of low technological uncertainty, performance and trust have a positive stronger relation. Higher trust seems to be effective for deriving improvements in performance. This is in line with statements of Krishnan et al. (2006, p. 898) because the limiting effects of trust are less relevant because complete and accurate environmental scanning and adjusting to the environment is less critical.
A higher amount of technological uncertainty will have a strengthening effect upon
13 We have to be cautious generalizing the outcomes of Clark (2006) to our context. Clark (2006) also analyses
the affective commitment-performance relation. This outcome can be understood by adopting the resource dependence approach (Kaufmann & Carter, 2006, p. 657). Following this approach partners want to secure the flow of goods to their organizations under technological uncertainty by forming stable, long-lasting relationships (Kaufmann & Carter, 2006: 657). Through affective commitment cooperative behavior is smoothened and in these stable relationships partners are willing to exchange complementary knowledge and skills (Wu & Cavusgil. 2005, p. 87). Krause et al. (2007, p. 534) state that through multiple interactions buyers and suppliers develop a common language for discussing technical and design issues which could effective for coping with technological uncertainty.
In the previous part, the results of the plots depicted in figure 3.1 were discussed upon. We will now look at the two plots in combination. Two overall results were generated; demand uncertainty did not have significant interactions with the relation between our behavioral variables and performance. Under conditions of high technological uncertainty trust seems to have neither a positive nor a negative relation with performance. Affective commitment relates to performance improvements positively under technologically uncertain conditions.
The effect of demand uncertainty was not significant upon both the trust-performance and the commitment-performance relation. In our regression analysis demand uncertainty had a direct negative relation with performance. Trust and commitment do not seem to be of influence in coping with demand uncertainty. When unpredictability and instability in demand are high the health organization can limit the negative performance outcomes, for instance by holding inventory.
as easily as for the disposable goods. A greater amount of goods exchanged in our sample can be classified as medical disposable goods. These goods can be kept in stock, for instance bandage material is not subjected to client specificity and these goods are not perishable. We conclude that this could be a reason for finding no significant moderator effects for demand uncertainty in our model and we expect that other mechanisms are used to cope with this uncertainty factor.
TABLE 4.1
The Products That Were exchanged in the Healthcare Buyer-Supplier Relation
The reason why the moderating effects were found in case of technological uncertainty is that addressing this factor could trigger more interaction. In our context, a lot of different products are used in the care process. Knowledge and skills need to be exchanged between the partners, for example when instructions for usage are needed because of the product specifications that were changed. In case of high technological uncertainty, commitment seems to be more effective than trust in the sense of performance outcomes.
These different results regarding the trust and affective commitment variables could be caused by the context of our study. Clarke (2006, p. 1199) states that health organizations are functioning in a regulatory framework, even more than is the case for other public
Goods supplied Number of suppliers Examples of products supplied
Medical disposables 24 cover material for operation room, bandage material, incontinency materials, plaster materials Medical investment goods 2 Probe feeding installations, spare parts and diverse installations
Disposables and investment goods 3 see examples for medical disposables and medical investment goods Specialized medical usage goods 9 implants, intervention materials, artificial limbs
organizations14. We think that the healthcare institutions are not acting in a hostile environment like manufacturers that operate in the competitive arena. For example, deriving profit is not the main issue for healthcare institutions and also competition is weak in this strongly regulated sector15. Risk-taking behaviors are limited by government control and thus also the behavioral manifestation of trust will be limited (Mayer et al, 1995, p. 724). Partners must take a risk16 in order to engage in trusting action. Commitment is a long-term orientation that is based on the assumption that the relationship is stable (Anderson & Weitz, 1992, p.19). When there is a stable affective relationship commitments can be implemented because both partners expect the relation to continue and partners do not have a high propensity to leave the relation. In a high risk environment, exchange partners could choose to break up the relation and choose a new partner. In these risky environments partners will not invest in affective commitment, because investments into the relation will not be returned.
Trust is not an unimportant behavioral variable and it results in a positive orientation towards the exchange partner that is reflected in affective commitment (De Ruyter et al., 2001, p.282). In cases where there is no trust it is highly unlikely that affective commitment is present and thus trust remains of importance.
4.1 Shortcomings of this research
The number of respondents was low and there are some explanations for this. Our four questionnaires could cost the institutions a lot of time. In the analysis we wanted to look at suppliers and buyers individually, but this was not possible because the number of suppliers
14 We must note that suppliers in our sample are situated in a more business-like environment. The risk in the
environment of healthcare suppliers is thought to be lower than in situations where products are delivered to organizations that are active in a competitive environment
15 Although regulated competition has recently been implemented in the Dutch healthcare sector
16 This concept is different from the uncertainty concepts used. For an illustration we refer to the article of Das &
in this study was not sufficient. For certain reasons17 our data were less valid because single informant bias was not cancelled out.
In the first hypothesis the relation between trust and performance was analyzed and a significant relation between these variables was found. From this it can not be concluded that higher trust leads to higher performance. Higher performance could also lead to more trust in time. Correlation and multiple regression analysis can not suggest that causality exists between the outcomes found.
The factors of trust identified by Mayer et al. (1995) were not found in the dataset used and these factors could lead to different or more accurate conclusions. This was also the case for our commitment measure where two instead of three factors were identified and performance, that consisted of one factor instead of more performance constructs.
Performance was measured by evaluating improvements and their importance for the past year. Trust, commitment and demand and technological uncertainty were measured in the contemporary situation. The performance of the dyad could be sufficient in the long run, but could be insufficient for this year. Because of this our results could be different when performance over the last five years was measured. Performance improvements could be more prevalent in earlier periods. Maybe investments in improvements are not done yearly, but for example every five years.
Uncertainty was conceptualized as consisting of demand and technology. Authors are inconsistent when measuring this concept and including these forms of uncertainty makes the uncertainty concept much clearer and easier to measure. Though environmental uncertainty was measured, effects of behavioral (or internal) uncertainty in the relation could lead to
17 In most instances, only one or two personnel members were able to fill in the questionnaire because of the
different conclusions18. Perceptions of uncertainty may be unique to individual observers (Heide & John, 1990, p. 34 & Eriksson and Sharma, 2003, p. 962) and implicitly some internal uncertainty is measured.
4.2 Future research
Future research could strive to collect more data from different health institutions so that these institutions can be looked upon individually. Hospitals are different with respect to for example the predictability of demand. In a care centre the number of patients is relatively stable because patients tend to live in these care centers for a long time. In a hospital a lot of patients stay a few days, so demand could vary more over time. It could be that in the context of hospitals different results are generated, especially with respect to demand uncertainty. The products used in hospitals are more specialized than those used in the care centers. We also think that technologies therefore develop more rapidly and occur more frequently in the hospital environment. This could lead to different outcomes, especially with respect to technological uncertainty.
In order to prove the causal relation between trust and performance, longitudinal research should be conducted to show how trust, commitment and performance are developed in time. These changes could be analyzed and so a clearer picture will be derived with respect to the direction of the relationships between the variables included in our studies.
Fruitful conceptual approaches to develop a deeper theoretical understanding of the trust phenomenon is still rare (Bachmann, 2001, p. 339), but including propensity and the factors of trust would make a considerable improvement for measuring it. Future research that concentrates on single relations should incorporate propensity to trust in the measurement of trust. The distinct commitment factors should also be incorporated in future research in order
18 For instance, the article of Krishnan et al. (2006: 896) predicts that the relation between trust and performance
to create conceptual consistency.
In this particular context more research is also needed to examine both the internal and external forms of uncertainty. Trust could enhance performance, but only under conditions of high behavioral (internal) uncertainty (Mcevily et al., 2003, p. 92). Using this form of uncertainty as contingent condition could generate different outcomes concerning the trust-performance and the commitment-trust-performance relationship.
Although we controlled for length of the relation, the fact that the mean relational length was 13 years indicates that overall we are dealing with relations that have been interacting with each other for a longer time period. Dwyer et al., (1987) devoted an article to relationship building and moves through different stages. Trustworthiness is particularly demonstrated and of importance in the earlier stages of relational development (Johnston, 2004, p. 37).
4.3 Scientifically expectations
This research was executed in the service sector and so results of this study can not be generalized to for example more production orientated companies or commercial services. In a production situation the purchasing function is more strategically embodied than in service organizations. In service organizations, providing a service is the core business and goods in the service process are needed to provide the service. In production organizations most of the expenses are made by the purchasing function, while personnel costs are the most important cost factor of service organizations. Healthcare organizations are controlled by legal regulations that could be of influence on our outcomes.
Scientifically, constructs of uncertainty should be more consistent because in this study both demand uncertainty and technological uncertainty have different moderating effects. Thereby, in many articles the trust and commitment constructs are combined and second order factors like relational quality are used. This research stressed the importance of analyzing these constructs in isolation, because in uncertain situations commitment and trust both behave different upon performance. Operations management and Supply Chain research can learn from channel management, sociology and marketing literature. Trust and commitment behave differently and this research proved this feature. Especially in operations management articles tend to use composite measures for trust and commitment (in combination with other behavioral concepts) that undermines the behavior of the individual constructs.
4.4 Managerial implications
Our research has identified some important insights for managers. Trust is an important condition, but trust can be overrated. Managers should invest in their relations, not only by formal long-term contracts, but also by investing psychological and implicit contracts. Environmental uncertainty is usually a motivation for investing in relational closeness, though demand uncertainty can easily be met by simple operational procedures. Investing in trust and commitment will therefore be inefficient in certain situations.
In our sample the affective commitment variable is especially important for rendering higher performance in technological uncertain situations. Commitment and trust are less needed when technological and demand uncertainties can be easily addressed. In situations of uncertainties that require a lot of interaction, requiring technological knowledge and implicit knowledge, commitment is an important feature in business relations. Explicit contracts can only be used in stable situations and have certain drawbacks, leading to lower performance.
4.5 Conclusions
This research attempted to use the behavioral concepts of trust and commitment in relation with performance under contingency conditions. The contingency factors used for this study were both demand and technological uncertainty. The sample consisted of buyers and suppliers in context of health institutions and single relations were analyzed. In this sample, higher trust had a positive relation with performance. Affective commitment did not have a significant relation with performance. The relation between continuance commitment and performance was weaker than with the variable affective commitment.
