Optimal Trust, Strategy or Fantasy?
Matching Trust with Interdependence, Asset Specificity and Uncertainties and its Impact on Partner Performance
By
NADHILA QAMARANI
S2777975
University of Groningen Faculty of Economics and Business
MSc Business Administration – Organizational & Management Control
June 2017
Supervisor: Mr. A. Rehman Abbasi Co-assessor: R.B.H. (Reggy) Hooghiemstra
2
Acknowledgement
3
Abstract
Since the idea of optimal trust was first introduced, there have been abundance discussion and debates in the literature (Parkhe, A., & Miller, S. R, 2000; Bachmann & Zaheer, 2006; Gargiulo & Ertug, 2006). Wicks, Berman & Jones (1999) claimed an immense need of evidence to the theory they proposed. This paper aimed to provide this empirical grounding and close the empirical gap in the optimal trust literature and contributed to the discussion of the existence of optimal trust as a strategy to achieve optimal partner performance in interfirm collaboration. The basic framework introduced by Wicks, Berman & Jones (1999) that proposed the moderating role of interdependence to the relationship, is extended by adding two variables to the framework; asset specificity and uncertainty. 107 respondents were observed in the analysis. The result of this study did not find any significant evidence of curvilinear relationship between trust and partner performance. Possible signs of inverted U-shape relationship were detected, however the significant linear relationship between the two was confirmed instead. This study also confirmed the negative effect of market uncertainty towards the trust-performance relationship. No significant evidence found on the moderating effect of interdependence, asset specificity and behavior uncertainty. Although this study only supported a linear relationship instead of curvilinear, this study contributed to both perspective. It confirmed that linear relationship still exists in some cases and that there were possible signs of curvilinear relationship between the two variables.
Keywords: trust, optimal trust, transaction cost economics, interdependence, asset specificity, behavior
4
Table of contents
Acknowledgement 2 Abstract 3 Table of contents 4 1. Introduction 5 1.1. Research Questions 6 2. Literature Review 72.1. Transaction Cost Economics 7
2.2. Trust 8 2.3. Interdependence 10 2.4. Asset Specificity 11 2.5. Uncertainty 12 2.6. Partner Performance 13 2.7. Conceptual Model 13 3. Methodology 13 3.1. Data Collection 14 3.2. Measurements 15 3.3. Data Analysis 16 4. Results 17
4.1. Descriptive Statistics and Correlation 17
4.2. Validity and Reliability test 19
4.3. Regression analysis 19
5. Discussion and Conclusion 22
5.1. Implications 27
5.2. Limitations and Suggestions for further research 28
5
1. Introduction
The attentions and discussion about trust in business has been increasing among scholars and practitioners. Stream line of research has comprehensively integrated trust with the more developed and widely known theory of transaction cost economics, underlining the significance of trust (Chiles & McMackin, 1996; Zaheer & Venkatraman, 1995; Gulati, R,1995). The logic behind TCE is ensuring the minimal cost to ensure the desired quality and quantity within the exchange arrangements are achieved. Hence, managers are in effort to align the arrangements and features of the interfirm alliance to match with the exchange conditions and hazards of the collaboration (Wiliamson, 1985). Formal contracts and relational governance, such as trust, are the two arrangements that managers can craft to fit the situation accordingly, in which Poppo and Zenger (2002) discovered to function as complementary to each other rather than as substitutes. These make TCE and trust has more imminent relationship and has been combined in abundant research (Chiles and McMackin, 1996; Goshal and Moran, 1996; Bunduchi, 2005).
In his theory, Wiliamson (1993) acknowledged behavioral assumptions which enhances the adaptability and relevance of TCE theory to business decision makers. Wiliamson established extreme behavioral assumption, claiming that the sole cause of opportunism is the human nature and the only reliable safeguard to prevent this opportunistic behavior is through control mechanism. Goshal and Moran (1996) argued that relying on these ideas jeopardize a trusting relationship as it develops suspicions toward partner firm and harm the propensity to trust. Blind suspicion will lead to blunt and restrictive control and surveillance, enhancing negative feelings for both sides, and in the end damage the performance in the collaboration. Such risk can be minimized by implementing more trust in a relationship. Ring and Van den Ven (1992) believes these economic safeguards may over time be replaced by implementing more trust within an exchange. They claimed that the higher ability to rely on trust, then the risk inherited in a transaction is lessened. Nevertheless, it is critical to be aware that opportunism and other risks still do exist in an exchange, causing saintly trust or putting too much trust to be disadvantageous.
6
the risk of being taken advantage of. Gargiulo & Ertug (2006) highlighted the ‘dark side of trust’ in their study, underlining the potential detrimental effect from trust.
Back in 1999, Wicks, Berman & Jones initiated the idea that the right amount of trust for the right situation is extremely critical, since trust also comes with its own disadvantage. They highlighted that, “trust is good ―but conditional good”. They acknowledge that overinvest or underinvest in trust where both can be destructive to a business, in which they introduced the notion of optimal trust.
Though trust received a lot of attention in the business research, the notion of optimal trust is still
underappreciated and in need of empirical ground (Gargiulo & Ertug, 2006). Wicks, Berman & Jones
(1999) developed foundations for deeper understanding of the concept of optimal trust through testable propositions which has not been empirically tested and remains unanswered until now. Furthermore, this study also extended and enriched the basic framework from Wicks, Berman & Jones (1999), which only considered the indirect effect of interdependence, by including the additional two factors to the
framework; asset specificity and uncertainty. By acknowledging the existing literature gap, my purpose
is to contribute to trust literature in general and optimal trust in particular. This research also offers insights for the endless spiraling of distrusting relationship dilemmas found in TCE.
Thus, I aim to further empirically tests the theory which will provide empirical evidence to the idea of optimal trust. This empirical evidence will provide theoretical contribution in two ways. Firstly, it adds empirical evidence on how trust influence partner performance in an interfirm relationship. Secondly, provides evidence on the moderating effect of interdependence, asset specificity, and uncertainty towards trust-performance relationship. The implications of this research finding will provide useful insight to address trust-related dilemmas in TCE. In practice, this empirical evidence will be useful for business practitioners as decision makers to adjust the arrangements and features of their interfirm collaboration to invest in more optimal level of trust in their strategy accordingly. Thus, the optimal net benefit and opportunities from collaborations can be achieved without corrupting partner performance and ruin the firm’s stability with om unnecessary risk or high agency cost.
1.1. Research Questions
Building from the literature gap identified and the research goals to be achieved, there are two research questions formed to be answered at the end of the research. First, the question focuses to find the empirical evidence on the relationship of trust and performance, leading to the first research question;
7
Next, is to find the empirical evidence of the notion of optimal trust by measuring the moderating impact
of interdependence, asset specificity, and uncertainty on trust and performance relationship, the second research question is;
RQ2: How does interdependence, asset specificity, and uncertainty moderate the relationship between
trust and partner performance?
2. Literature Review
2.1. Transaction Cost Economics
The transaction cost economic theory is widely known through the work of Wiliamson (1975, 1985) which was inspired by Coase’s work who first coined the theory back in 1937. This paper focuses on Wiliamson’s TCE since it is more developed as well as it is more relevant to business decision makers (Goshal and Moran, 1996). Goshal and Moran claimed that Wiliamson’s TCE predominates TCE research and applications, as it recognizes behavioral assumptions which offers logic closer to reality and normative purposes and enhances its practicality.
The main idea of Wiliamson’s TCE is managerial decision making, choice or predictions of governance structures assigned for the transaction. This theory addresses the three behavioral assumptions; opportunism, bounded rationality, and risk neutrality, which help to predict firm’s choice of governance structure. The inclusion of behavioral assumptions in this theory contribute to the subjective nature of cost in this theory. It also brings more imminent relationship between TCE and trust where these two ideas have been combined in abundant research (Goshal and Moran, 1996; Bunduchi, 2005). Chiles and McMackin (1996) has incorporated trust in TCE model and claimed the contributions of trust to the theory’s validity. Zaheer and Venkatraman (1995) claimed that incorporating noneconomic (sociological) factors such as trust within the traditional cost theory provide better understanding of interorganizational phenomenon.
8
begin to distrust their partner as consequences of their own surveillance, claiming them as unmotivated or untrustworthy. The partners on the other hand become demonstrably untrustworthy and eventually demands more intensive control and surveillance, destroying both trust and trustworthiness from both side (Enzle and Anderson,1993; Lepper and Greene (1975). These limitations call out attention to incorporate TCE with trust, and highlighted the importance in understanding the right amount of trust to invest in interfirm. collaboration
2.2. Trust
Many scholars in trust literature, who came from diverse background and disciplines, contributed their idea to define trust. Beccerra and Gupta (1999) claimed trust literature as ‘conglomerate of ideas without
a solid framework’. Ring and Van de Ven (1992) summarized existing trust literature in late 1980s into
two streams of trust definitions which are the risk-based view and faith-based view. In the first one, trust exists as "confidence or predictability in one's expectation". The second one described trust as "faith in the goodwill of others not to harm your interests when you are vulnerable to them".
The two streams of trust definition were then aligned in Wicks, Berman & Jones (1999) interpretation of trust. They recognized that both rational prediction and affect-based belief in moral character (goodwill) are significant since rational prediction helps preventing saintly trust while affect-based belief in moral character is critical to create and maintain trusting relationship. Thus, they summarized that trust starts when some element of deterrent (e.g., assured destruction) and affect-based belief in moral character are included in decision making until the point where it become a “blind faith” in others’ moral character. Their interpretation highlighted the existence of different level of trust.
Arrow (1974) convinced that utilizing trust in managing economic transaction can be the most efficient instrument. Without trust, more effort is needed in seeking partner resulting in an increase for the search cost. Preparing contracts will also consume more costs and time since more detailed contract is required to compensate the absence of trust in the relationship. Moreover, such detailed contract may become a
barrier to adapt in changing environment (Gulati, 1995; Nohria & Eccles, 1992). Trust improve firm’s
efficiency in market exchange (Smith, 1981). Trust also help economizing transaction costs result in more efficient transaction (Jones, 1995), since numerous costs such as searching, negotiating, monitoring, or enforcing can be lowered or even eliminated. When partners in relationship trust each other, they don’t need to invest in any costly incentive mechanism. With all the benefits offered, as firm invest more in building trust in their relationship, firm’s performance will increase.
2.2.1. Overinvest and underinvest in Trust.
9
(performance) pointing out the dark side of trust and the detrimental effect of too much of a good thing (Bachmann & Zaheer, 2006; Gargiulo & Ertug, 2006). They argued that association of trust with its negative consequences was rarely discussed in the literature. In accordance with this, Wicks, Berman
& Jones (1999) addressed that, “trust is good ―but conditional good”. They elaborated two
unfavorable scenarios that may occur in employing trust in a business strategy which are underinvest and overinvest in trust. There is a possibility that in the end, firm may harm their business as they invest in inappropriate amount of trust as their strategy for particular relationship.
Firm can be exposed to the risk of overinvesting in trust when they put too much trust and effort in
building trust in a relationship. Flores and Solomon (1998) stated, “There is such a thing as too much
trust, and then there is "blind trust," trust without warrant, foolish trust”. This can be dangerous as firm
is exposed to unnecessary risk and misallocate their resources in the process. These may lead to decrease in cautiousness to partner's opportunistic behavior and risk of information leakage (Krishnan, Martin
and Noorderhaven, 2006). The second unfavorable scenario is underinvesting in trust and relies more
on protections and safeguards that often costly (Wicks, Berman & Jones, 1999). This exposes firms to the chance of missing out opportunities for cost savings in the relationship where enabling more trust can be better off for both sides. Furthermore, it leads to “pathological spiraling relationship” between
both firms, where the controllers’ distrust their partners more from the surveillance result while the controlee feel untrustworthy and unmotivated (Enzle and Anderson,1993).
2.2.2. Optimal Trust
By acknowledging the benefits, risks, and unfavorable scenarios upon trust implementation, we know that trust is valuable as much as it is costly and risky to create. Hence, Wicks, Berman & Jones (1999) introduced and developed foundations of the concept of optimal trust. They elaborated the basic idea of optimal trust based on Aristotle’s work of “golden means between extremes: excess or deficiency”. In the context of trust, golden mean represents the optimal trust, between the extremes which are naivete and cynicism. They provided their definition of optimal trust as as an embedded construct where trust is shaped and influenced by variety of actors, features, and structures, where the agents are committed to trust but also sensibly decide who to trust and its extent and capacity.
10
unnecessary risk such as the risk of partners’ opportunistic behavior, the lack of motivation from partner to make improvement or follow the alliance target as the situation is too comfortable for them and no incentive to improve or follow rules.
Hypothesis 1: The relationship between trust and partner performance is curvilinear (inverted-U shape).
Wicks, Berman & Jones (1999) provided a relevant way to operationalize optimal trust by linking level of trust to interdependence in a relationship to see how it enhance the benefit of trust towards partner performance. When there is a match between trust level and interdependence level, the trust level invested is the optimal level for the relationship, which will result in optimal benefit for the performance in a relationship. This study extended Wicks, Berman & Jones (1999) basic framework by including two additional factors that may have indirect relationship towards trust-performance relationship, which are asset specificity and uncertainty.
2.3. Interdependence
Interdependence represents how two parties are mutually dependent on each other to achieve certain goal and it may vary according to task or work division incorporated within the alliance (Calton and Lad, 1995; Gulati and Singh, 1998). The task and work division between the two parties can be simple
with minimal adjustment required but it can also be complex and overlapped. Wicks, Berman & Jones
(1999) noted that the higher extent of outcomes which are influenced by the action of other party is considered as interdependent while in the other extreme, if the relationship between the two parties have no significant dependencies, then the relationship is independent.
Figure 1.
11
Gulati and Singh (1998) stated that higher interdependence within a relationship entails higher expected coordination cost and greater challenge in task coordination and control. Hence, higher interdependence also comes with its more risk which increase the need and significance of higher level of trust in that context as coping mechanism towards those risks. On the opposite situation, when firm wish to enter independent relationship (such as market), they are not dependent on certain partner to obtain valuable resources. This may happen when the task is simple or there is strong market of alternatives available. Employing high level of trust in such case may outweigh the benefit gained from relationship as
building trust is costly and the partner firm may also feel betrayed and taken advantage of (Wicks,
Berman & Jones, 1999). Independent relationship doesn’t entail strong ties between parties, which makes it possible for firm to freely exit the relationship and contract another partner alternatives. Investing too much trust in this context is costly and binding, leading to conditions of fit and misfit represented in the Figure 2 below;
Thus, these conditions result in hypothesis 2 below;
Hypothesis 2: The performance effect of investing on trust is more positive for interdependent relationship than for independent relationship
2.4. Asset Specificity
Young-Ybarra & Wiersema (1999) elaborated asset specificity as a useful method in strengthening the ties between firms. The investment specific relationship will prevent parties involved to behave opportunistically as the investment made in the specific asset comes with substantial termination cost (Murray & Kotabe, 2005). The direct positive effect of asset specificity on trust and commitment has been proven by many authors, since asset specific asset sometimes require asset specific knowledge or strategy that is useless or not transferable for other purpose or creates the expectation of continuity (Kwon & Suh, 2005; Parkhe, 1993). Prior research also found positive relationship between trust and asset specificity as such specific investment increases firms’ credibility and bonding effects (Jeng and
Figure 2.
Trust, dependence and performance
Interdependent Relationship Independent Relationship T ru st Low Misfit Fit More positive effect on
performance
High
Fit More positive effect on
performance
12
Mortel, 2010; Dyer, 1996). However, whether asset specificity has moderating effect of on the trust performance relationship is yet to be tested.
Hypothesis 3: The performance effect of investing on high trust is more positive for higher asset specificity than for the lower asset specificity
2.5. Uncertainty
Uncertainty exist when companies are unable to predict the future because of incomplete knowledge (Beckman, Haunschild & Phillips, 2004) Trust and uncertainty have an inevitably close relationship as the greater level of uncertainty leads to greater need of trust between firm as an effective governance mechanism (Madhok, 1995; Hansen, 1994). However, the higher behavior uncertainty is believed to exposes firm with unnecessary risks affecting partner performance in the alliance. Behavior uncertainty develop tension between firms in their collaboration as actions of partners becomes more difficult to anticipate or be understood (Sutcliffe & Zaheer, 1998). This can lead to many potential misunderstandings in alliance and easily starts conflict between partners, making it harder for both parties to coordinate and synergize (Park & Russo, 1996). When behavior uncertainty exist, firms will not be able to predict the intents of their partner, which drive firms to, sometimes without realization, detract their partners from fully contribute to the collaboration, lessening partner contribution and performance towards the relationship (Grindley, Mowery,& Silverman, 1994)
The external uncertainty such as market uncertainty is a more complicated issue since it is much harder to be controlled or predicted. When the uncertainty in the market is high, Mintzberg (1978) stated that another coordination problem might occur due to excessive information exchange in alliance. The unpredictable environment demands partners in alliance to constantly detect changes, adapt, and make decisions quickly. This may lead to biases in decision making among partners due to information unfamiliarity, which can result in significant errors (Park & Sheath, 1975; Ferrin & Dirks, 2003). It is also problematic since partner may try to adapt to the changes that comes with high market uncertainty, making major changes, and distract them from fully contribute to the collaboration (Harrigan, 1985). Not to mention, when firms invest too much in trust in such situation, the possibility of strategic blindness (Krishnan, Martin and Noorderhaven, 2006) may exist which limits the efforts from both partner to answer the challenge in the changing market. Both internal and external uncertainty is fundamental in business relationship.
Hypothesis 4: The performance effect of investing on trust is more negative when there was higher behavioral uncertainty than for the lower behavioral uncertainty
13 2.6. Partner Performance
Ensuring partner performance and contribution within an alliance by exerting surveillance frameworks and mechanisms is not a very profitable strategy since the effort and cost to do so often eat away the potential profits. Instead, it is beneficial to have mutual trust in the relationship as partner performance will be improved due to strong social foundation, less conflicts, and less monitoring cost (Cullen, 2000; Madhok, 1995). With trust invested in the alliance relationship, partner is allowed to function more effective and efficiently, increasing partners' quality of contribution, responsiveness and goal achievement and optimizing the overall cost.
2.7. Conceptual Model
The five hypotheses elaborated in the previous section were pictured in figure 3 below. This study proposed to test one direct interaction effect and four moderation effect.
3. Methodology
The main purpose of this research is to provide evidence-based answer that will help in defining relationship between trust and partner performance within an alliance. Secondly, this research also aims to find out how interdependence, asset specificity, and uncertainty affects relationship between trust and partner performance. Five hypotheses consisting of one main effect and four interaction effect hypotheses will be investigated with quantitative and theory testing approach which are suitable for this
research (Colquitt & Zapata-Phelan, 2007). In order to conduct this quantitative research, a survey was
organized through online questionnaire (Wright, 2005). This research will adopt steps for theory testing provided by Van Aken, Berends and Van der Bij (2012), which consist of; (1) addressing the business phenomenon and literature gap; (2) Generation of hypothesis and conceptual model; and the three part
14
that will be elaborated in this latter part which consists of (3) data collection; (4) data analysis; (5) result interpretation and implication.
3.1. Data Collection
Data collection is necessary to be done in this research as there is no existing data available. Primary data was obtained using survey instrument in the form of questionnaires, provided by the supervisor of this research and has been validated in previous studies. List of possible companies was made through media, network and personal contacts by adopting the stratified random method (Bowen et. al, 2009; Morales and Fernandez, 2009). This method is helpful to choose respondents with predetermined criteria based on the firm size (annual revenue, assets, and number of employees) and the age and nature of the interfirm collaboration where a contract was signed. This method also offers the advantage of consistency, simplicity and public acceptance (Brewer, 1999).
Respondents for this survey were managers of procurement function who were involved and have sufficient knowledge of interfirm collaboration within their company. Data collections were done through collective effort of seven master students. Eligible respondents were first contacted and reached via email or phone to inform them about the basic idea of the study and obtain their agreement to participate in the online survey. Upon their permission, the link for online questionnaire was sent via email. Reminders to complete the survey were sent via email and phone. The online survey facilitate access to the respondents, saves administration cost and time and avoids interviewer bias (Wright, 2005; Loijens, 2014). Some effort such as the predetermined criteria in choosing respondent and clear instructions to complete the survey, checking the accuracy of email address and keeping tracks of them, asserting confidentiality and anonymity, were done help to minimize the sampling issues, respondent errors and the lack of control for survey situations in online survey (Simsek and Veiga, 2000).
15 Table 1. Respondents Characteristics
Mean SD Min Median Max
Firm Age (years) 44.48 52.171 2 28 407
Firm Size (number of employee) 9891,98 33456.344 7 200 225,000
Alliance Age (years) 9.36 8.637 0.4 6 44
3.2. Measurements
The survey instrument collected information regarding the interfirm collaboration, trust, performance, asset specificity, interdependence, and uncertainties within a specific collaboration in which the manager possess the most knowledge, experience, and most familiar with. The unit analysis in this study is the interorganizational alliance or collaboration. The items are measured with seven-point Likert scale, which Nunnally (1978) claimed as scale with appropriate balance and enough degree of discrimination.
3.2.1. Control Variables
Control variables hold prominent role in quantitative research as it helps to set aside possible threats that might interfere the statistical testing by applying control within the designated research (Cook & Campbell, 1979). In selecting the control variables, Atinc et al. (2015) recommend several arguments. Selecting the relevant control variable can be done by choosing variables that has been confirmed by others with detailed justification of how the control variables relate to the study. This justification comes in the forms of causal relationship, correlations, or even from citation of empirical evidence from existing literature. In this study, three control variables were exerted in the research design, which are;
Firm’s size. The size of a firm tends to affects its power and resources to improve performance. Firm’s
size as control variable has frequently been applied especially in studies investigating performance and it’s often determined by number of employee within a firm (Tsai, 2001; Morales and Fernandez, 2009). More employee reflects how firms possess more resources, which will generate more power and strengthens their position. This situation affects partner's bargaining power, contribution, and eventually their performance within an interfirm collaboration (Pfeffer & Salancick, 1978; Harrigan 1986)
Firm’s age. The firm’s age is determined by number of years the firm has been operating. This control
16
Alliance age. The number of years a collaboration has existed is an important factor to be taken into
account in studying performance within alliance (Muthusamy & White, 2006; Lunnan & Haugland, 2008). As the longer period collaboration is established the more understanding and harmony will be created between partners hence conflicts are less likely to occur (Lin & Germain, 1998). Alliance age need to be controlled in this study as Simonin (1997) pointed out the more time both partner spent interacting with each other, the collaborative learning as well as efficiency will be developed and influence the contribution of both parties.
3.3. Data Analysis
The nature of this theory testing research is quantitative analysis with data retrieved from survey questionnaires and SPSS is the main tool used to analyze this data. Before getting to the statistical analysis, a series of data cleaning processes were conducted. The control variables, which are nominal data, have inconsistent formats in the database (e.g. contains text). Since several items were reverse coded, an effort had to be made to transform these items to avoid misinterpretation of data (Pallant, 2013).
The data analysis was initiated by testing construct validity by making sure that each item belongs to the construct they represent using principal component analysis. For an acceptable result of the analysis, correlation, Kaiser-Meyer-Olkin (KMO), sampling adequacy, and Bartlett's test of sphericity are necessary to be investigated. This analysis is a technique to reduce larger set of variables into the optimal number of components. It reveals how many constructs are formed, which items belong to specific construct, and if removal of an item is necessary (Cattel, 1966; Kaiser, 1974).
The reliability of remaining items was tested using Cronbach’s alpha, which calculate if several variables do have internal consistency to form a construct together (Bland & Altman, 1997). After ensuring the validity and reliability of the data, testing the hypothesis can be done using hierarchical multiple regression analysis. The main hypothesis of this research is to prove the existence of a curvilinear relationship between trust and partner performance. The moderation effect of interdependence, asset specificity, and uncertainty towards trust-performance relationship will be investigated also using hierarchical regression analysis. The regression model of each hypothesis can be found in table 2 below.
The first model only includes partner performance (PRTNRPER) as the dependent variable and the control variables (firm size, firm age, and alliance age). To investigate the curvilinear relationship, the
quadratic main effect need to be tested with following formula: Y=b0+b1X+b2X2 (Adobor, 2006,
17 Table 2. Regression Model
Model Regression
1 PRTNRPER= B0 + B1*FSIZE + B3*FAGE + B4*ALLAGE
2 PRTNRPER= B0 + B1*FSIZE + B2*FAGE + B3*ALLAGE + B4*TRUST
3 H1 PRTNRPER= B0 + B1*FSIZE + B2*FAGE + B3*ALLAGE + B4*TRUST + B5*TRUST2
4 H2 PRTNRPER= B0 + B1*FSIZE + B2*FAGE + B3*ALLAGE + B4*TRUST + B5*TRUST2+ B6*IND + B7*Trust*IND + B8*Trust2*IND
5 H3 PRTNRPER= B0 + B1*FSIZE + B2*FAGE + B3*ALLAGE + B4*TRUST + B5*TRUST2+ B6*ASPC + B7*TRUST*ASPC + B8*TRUST2*ASPC
6 H4 PRTNRPER= B0 + B1*FSIZE + B2*FAGE + B3*ALLAGE + B4*TRUST + B5*TRUST2+ B6*BHVUN + B7*TRUST*BHVUN+ B8*TRUST2*BHVUN
7 H5 B0 + B1*FSIZE + B2*FAGE + B3*ALLAGE + B4*TRUST + B5*TRUST2+ B6*MRKUN + B7*TRUST*MRKUN+ B8*TRUST2*MRKUN
The interactions of trust with moderators such as interdependence (IND), asset specificity (ASPC), behavioral uncertainty (BHVUN), and market uncertainty (MRKUN) were tested separately in model
4, 5, and 6 respectively, with the following equation: Y = b0 + b1X + b2X2 + b3Z + b4XZ + b5X2Z, to
see interaction effect in curvilinear relationship (Janssen, 2001; Aiken et al., 1991). In total, there are 7 models tested in the regression analysis.
4. Results
4.1. Descriptive Statistics and Correlation
Descriptive statistics helps in summarizing and describing the data by showing the central tendency, the spread of the data, deviation, and many other measures (Hinton et al., 2014). In table 3, the result of the means and correlations of each of the items are displayed. The mean for items in the partner performance construct were significantly higher than other variables with smaller dispersion of scores as shown by lower standard deviation compared to the others. This can be translated as the more positive responses received in terms of partner performance compared to the other variables. The low mean value of interdependence and behavior uncertainty also deserve mentioning. This may suggest that alliance participating in this study did not commit in a highly interdependent collaboration or encounter major behavioral uncertainty.
18
There is a negative correlation between trust and behavior uncertainty (r = -0.346, p < 0.01) indicating trust may decrease when the behavior uncertainty is higher since higher tension exist because of difficulty in anticipating partner's action (Park & Ungson, 2001).
Table 3. Descriptive Statistic and Correlations
Trust and Interdependence surprisingly have negative correlations contrary to the literature (Wicks, Berman & Jones, 1999). The possible arguments for this is that with higher interdependence in a relationship, the need for trust is less. It is also possible that as partners become so dependent and attached, dependence problem (e.g. vulnerability, overly dependent, excessive control and involvement) causing the nature of trust to shift into more calculative based trust (Premus & Sanders, 2008; Poppo et al., 2015).
Partner performance and market uncertainty have negative correlations (r = -0.367, p < 0.01). This shows that high market uncertainty may leads to lower performance from partner due to greater risks and volatility, which is also supported by previous studies (Adobor, 2006; Hansen 1994). A Significantly positive correlation is found between market uncertainty and asset specificity which reflects how high uncertainty in the market may contribute to the more asset specificity within an alliance (r = 0.288, p < 0.01). The unpredictability that comes with market uncertainty may be minimized and controlled by investing in asset specialization which also one of the strategy implemented by Japanese automotive firms (Dyer, 1996). A positive correlation between asset specificity and interdependence exists (r = 0.347, p < 0.01). This is consistent with the literature as the more asset specialization is applied the more partners become interdependent to one another (Das and Teng, 2003). Mean S.D. 1 2 3 4 5 6 7 8 9 Firm Age 44.48 52.17 1.000 Firm Size 9891.98 33456.34 0.162 1.000 Alliance Age 9.358 8.637 0.021 0.029 1.000 Trust 4.765 1.047 0.094 0.080 -0.194* 1.000 Interdependence 3.350 1.738 -0.129 -0.310** 0.033 -0.304** 1.000 Asset Specificity 4.015 1.445 0.032 -0.228 * 0.080 -0.170 0.347** 1.000 Behavior Uncertainty 3.009 1.433 -0.047 -0.196 -0.157 -0.346 ** 0.133 0.140 1.000 Market Uncertainty 4.399 1.285 -0.029 0.078 0.074 -0.002 0.031 0.288 ** -0.019 1.000 Partner Performance 5.407 0.932 -0.033 0.007 0.078 0.324 ** 0.042 -0.032 -0.367** 0.018 1.000
19 4.2. Validity and Reliability test
Validity analysis using principal component analysis (PCA) makes sure that each item is loaded correctly in their respective constructs and there is no cross loading of items (Cattel, 1966; Kaiser, 1974). Before proceeding to the result, the data was first checked to ensure that assumptions to perform PCA are met. First, all variables are required to have at least one correlation coefficient greater than 0.3. Then, the assumption of sampling adequacy was tested using a measure of Kaiser-Meyer-Olkin (KMO) measure of 0.666, an appropriate value according to Kaiser (1974). Bartlett's test of sphericity was also statistically significant (p<.0005), concluding that PCA can be performed in this dataset. The result of this validity analysis confirms six principal components after removing 2 items due to low individual KMO.
The remaining item after PCA were then proceeded into the next step, which was calculating reliability analysis using Cronbach alphas. Cronbach's alpha is a commonly used tool for reliability analysis (Hinton et al., 2014; Bland & Altman, 1997). The values of 0.7 or higher shows a very appropriate internal consistency but the value above 0.5 is considered acceptable (DeVillis, 2003; Kline, 2005). The Cronbach’s alpha of interdependence was not measured in this analysis because there was only one item in interdependence construct.
Table 4. Principal Components and Cronbach Alpha
NA = Not Applicable
4.3. Regression analysis
Hierarchical multiple regression analysis was used to test the main and interaction effect in the five hypotheses in this research. Hierarchical regression analysis provides the statistical answer of how the addition of independent variables can explain the variation in the dependent variable (Gelman & Hill, 2007). In order to ensure hierarchical regression is the appropriate analytical tool the data must be investigated whether it fulfills seven required assumptions.
First, the Durbin-Watson statistic test was done to ensure the independence of observations. The Durbin-Watson statistic for this analysis varies from 1.57 to 2.04 which are sufficient as it’s close to 2. Hence, it is accepted that there is independence of residuals (errors) in the data (Gujarati & Porter, 1999). The second assumption tests for linearity using collective scatterplot of studentized residuals against the predicted values and a partial regression plot was executed to test the linearity of dependent variable with each of the independent variables (except the control variables).
Trust Interdependence Specificity Asset Uncertainty Behavior Uncertainty Market Performance Partner
Cronbach’s
20
The data was then checked for multicollinearity. The data in this study has no multicollinearity issue as the VIF values were less than 10 and the tolerance values were greater than 0.1. Afterwards, the data was checked for multicollinearity. The data in this study has no multicollinearity issue as the VIF values were less than 10 and the tolerance values were greater than 0.1. An important exclusion need to be
made for one of the variable, Trust2. There is high multicollinearity between Trust and Trust2 because
Trust2 is the quadratic term of Trust and both have high correlations and happen. This was expected and
happens in many research concerning curvilinear relationship (Miller & Schwieterman, 2013; MacCallum & Marr, 1995; Shepperd, 1991). Furthermore, Ganzach (1997) claimed that the inclusion of quadratic terms will generate a better and more representative model rather than excluding them and only use the linear terms.
Afterwards, significant outliers and high leverage points are identified and removed from the analysis. There were 4 outliers detected with standardized residual value greater than ±3, leverage value greater than 0.2, or Cook's distance values less than 1 (Cook and Weisberg, 1982). These outliers were later removed from the regression analysis. There were 3 outliers detected with standardized residual value greater than OO3, leverage value greater than 0.2, or Cook's distance values less than 1 (Cook and Weisberg, 1982). Lastly, the residuals (errors) were confirmed for normal distribution by Normal Q-Q Plot of the studentized residuals
The first model showed all three control variables have no significant relationship with partner
performance. The variance that can be explained (R2) by the inclusion of control variables were very
low and not significant (R2 = 0.009). Therefore, it can be concluded that in this study, the firm age, firm
size and alliance age were not variables that could explain the variance in partner performance. The next step added linear term for trust to the second model and leads to a statistically significant model
and significant increase in R2 by 21.9% (R2 = 0.228, F(4,99)= 7.927, p < 0.001). Trust appeared to be
positively related to partner performance (b = 0.382, p < 0.05).
The third model tested H1 by entering the quadratic term, Trust2, to the regression model. The result
shows that there is indeed a curvilinear relationship between trust and partner performance by comparing the b value between Trust (b = 0.807) and Trust2 (b = -0.042). The negative coefficient for
the interaction of Trust2 indicated how trust positively affect performance but to a certain point this
trend goes downward (inverted U-shape). However, this result was not significant (p > 0.05). The slight
improvement on R2 by the inclusion of Trust2 to the model was also not significant. Thus, H1 was not
22
The interaction effects were tested in the next four models. Model 4 analyzed the moderation effect of interdependence toward trust and partner performance relationship. The model was statistically significant (F(8,95) = 3.920, p <0.01), however the increase in variance explained by adding interdependence to the model is not significant. There was no support of positive moderating effect from interdependence, hence, H2 is not supported. The moderating role of asset specificity investigated in model 5 also leads to nonsignificant result, while the overall model was statistically significant (F(8,95) = 3.975, p<0.01). No positive interaction found between asset specificity and trust-performance relationship. H3 is not supported from this analysis. In model 6, behavior uncertainty was entered and yielded in another nonsignificant result of the moderation effect and rejected H4. However, there was significant result for the main effect between behavior uncertainty and partner performance where behavior uncertainty negatively affect performance (b = -0.259, p<0.01).
The last model explored the interaction effect of market uncertainty and the significant result was shown. The inclusion of market uncertainty as moderator significantly improve variance explained in the model by 6.6% (R2 = 0.300, F(8,95) = 5.101, p<0.05). The overall model is statistically significant and significantly negative moderating relationship was discovered (b = 1.344, p < 0.05). Hence, hypothesis 5 is supported.
5. Discussion and Conclusion
This research was initiated to provide an empirical support towards the theory of optimal trust elaborated by Wicks, Berman & Jones (1999) who stated the major need of empirical testing to ground and enrich this theory. The total of 107 observations were explored in this study to test hypotheses developed mainly from Wicks, Berman & Jones (1999) and relevant research concerning trust and interfirm collaborations. From all five hypotheses, one hypothesis is not confirmed, three hypotheses were rejected, and one hypothesis is supported.
H1: The relationship between trust and partner performance is curvilinear (inverted-U shape) - Not supported.
23
this study. This part of the regression results alone, confirmed how performance begins to increase along with the increase in trust, yet up to a certain threshold, the performance effect of trust will diminish. For easier visual interpretation, a simple slope test (Aiken & West, 1991) was applied to illustrate and compare both linear (simple slope: b = 0.382, t = 5.297) and quadratic relationship (simple slope: b = -0.427, t = -0.882) and presented in figure 4.
As depicted from the figure 4, there is a sign of inverted U shape relationship (the dotted line) in trust and partner performance relationship. This may be in accordance with the idea proposed by many authors in the literature that trust is good but conditional good and too much of a good thing can lead to a detrimental effect (Wicks, Berman & Jones, 1999; Bachmann & Zaheer, 2006). Investing trust within a relationship may at first bring many advantages and positive effect to partner performance such as cost efficiency, more efficient transaction process, or better coordination and synergy with partner which increase partner’s responsiveness to the alliance (Arrow, 1974; Smith, 1981; Jones, 1995). Nevertheless, up to a certain point when a firm put too much trust to their partner there are some behavioral consequences such as; risk of malfeasance, opportunistic behavior or incompetence from partner due to the lack of monitoring, the inability to detect the decline in partner performance, the lack of motivation from partner to make improvement or follow the target of the alliance due to the excessive trust or blind faith they have earned and put them in a comfortable zone might occur and responsible for the downward trend of curve (Gargiulo & Ertug, 2006).
Still, this curvilinear effect is not significantly proven in this study. There are several possible arguments for this result. First, it is possible that the trust level of invested in the alliance of respondents in this study has not yet passed the optimum level. Judging by the mean and standard deviation of trust score
given by the respondents (x̅ = 4.765, s.d. = 1.047) which is high but not particularly extreme in the case
of 7 Likert scale items, it can be assumed that the trust invested in the alliance still yield in the upward Figure 4.
24
part of the curve. Gargiulo and Ertug (2006) stated that only after certain threshold is passed, optimal trust become excessive trust, and yield in damaging effects on partner performance. The length of alliance period may also be another contributing factor to this result as almost 50% of the samples have the alliance age below 5. Possibly most of these alliances still at the stage of building trust (Curall & Epstein, 2003), hence another prediction can be made is that the long-term effect of trust investment is not yet detected or the net benefit of trust is not yet calculated. Since the test only confirmed the linear relationship, the curvilinear relationship is therefore not supported in this study.
H2: The performance effect of investing on trust is more positive for interdependent relationship than for independent relationship - Not supported.
The prediction of moderating role of interdependence in the trust-performance relationship was rejected in this study. This shows that the degree of dependency between firms in an alliance do not affect the relationship between trust and partner performance in anyway. There are several possible causes to this result. Although some authors believed the positive moderation role of interdependence (Wicks, Berman & Jones, 1999; Krishnan, Martin and Noorderhaven, 2006), in his research, Campion (1993) found no significant relation between interdependence and productivity, satisfaction or judgement within a work group. Hence, it is possible that in this study, interdependence also did not explain the productivity and satisfaction of the benefit in investing trust to the partner performance. The low mean value of interdependence score obtained from respondents in this study (3.350) may be another reason behind this result as it indicates alliance in general did not involve in a highly dependent relationship (Sambasivan et al., 2013).
Hypothesis 3: The performance effect of investing on high trust is more positive for higher asset specificity than for the lower asset specificity - Not supported.
25
Hypothesis 4: The performance effect of investing on trust is more negative for higher behavioral uncertainty than for the lower behavioral uncertainty - Not supported.
The regression analysis did not provide any significant support the existence of moderating effect from behavioral uncertainty towards the main effect of trust and partner performance relationship. However, a significant direct interaction of behavior uncertainty towards partner performance is discovered from the analysis. This research found that behavior uncertainty negatively influences partner performance within an alliance. This interaction is illustrated in figure 5. In their research, O'Driscoll & Beehr (1994) expressed that uncertainty is associated with substantial concerns such as psychological strain, lack of satisfaction, or intention to leave organization, or in this case, leave the alliance. Parkhe (1993) stated that behavioral uncertainty can lead to higher risk of opportunism, agreement violations, or cheating from partner, which explain why the higher behavioral uncertainty will negatively affect the way partner contribute and perform in the interfirm collaboration.
Hypothesis 5: The performance effect of investing on trust is more negative for higher market uncertainty than for the lower market uncertainty - Supported
The results of regression analysis provided a significant support towards the moderating role of market uncertainty where hypothesis 5 was accepted. The market uncertainty coefficient of interaction of b = 1.344 towards the quadratic term of trust and partner performance relationship confirmed that market uncertainty negatively influences trust effects on partner performance. Figure 6 interpreted how this interaction is portrayed. The solid line in figure 6 resembles the situation where market uncertainty is low. As depicted on the graph, the curve line is in inverted U-shape, almost similar to the predicted relationship between trust and partner performance in a normal condition, ceteris paribus. In the low market uncertainty, when there is less risk, unpredictability and volatility, the effect of investing in more trust is advantageous and contributes to the increase in partner performance. However, once there is a high market uncertainty happening, this curve line drastically changes leading to a totally contrast result
Figure 5.
26
on partner performance. As pictured by the dotted line, with high uncertainty in the market it is better to invest in low trust and preferably relies more in more formal controlling instrument to maintain desirable and expected performance from partner and avoid unnecessary risk that comes with market unpredictability. If a firm keep investing on more trust in such case, the benefit of trust in partner relationship will decline as shown by downward slope in the dotted line. Since partner firm feels like they’ve gained trust in a high market uncertainty, it will create room for opportunistic behavior and other behavioral consequences.
Nevertheless, an important note need to be made concerning existence of a critical threshold or optimal point in this interaction. When market uncertainty is low, low trust result in lower partner performance and as the trust grows, the line showed an upward slope, signaling that partner performance is improved due to better coordination and harmony between firms. However, this only apply up to a certain threshold, then this effect will decrease as the slope goes downwards and turns into a detrimental effect to partner performance. On the other hand, in a situation with high market uncertainty, low trust result in better partner performance, yet, as firms invest more in building trust, the positive effect on partner performance diminishes and goes downward. After a certain point however, the benefit of investing in trust will outweigh the consequences that comes with it in the beginning, resulting in a net benefit of trust that increase partner performance. For instance, when the market uncertainty is high and a firm has successfully build an optimal amount of trust, indulging in a close interfirm alliance will becomes an effective mechanism to deal with market uncertainty.
Referring back to the two main questions that this research aimed to solve in the beginning, after the whole process of data gathering, data analysis and interpretation, some conclusions can be drawn to answer this research question. Firstly, this research intended to define the relationship between trust and partner performance and see if this relationship, as predicted by many authors in the literature, is indeed curvilinear with inverted U-shape. Based on the result of this study, it can be summarized that there is
Figure 6.
27
a positive interaction between trust and partner performance, reflected in a linear relationship, instead of curvilinear as predicted. Secondly, is how interdependence, asset specificity, behavioral uncertainty and market uncertainty contributes to this relationship. Market uncertainty is proven to decrease and even reverse the trust effect on partner performance when there is high market uncertainty. Meanwhile, there was no significant moderating influence identified from interdependence, asset specificity, or behavior uncertainty. With all the discussions and debates concerning the existence of optimal trust as a strategy or just a mere fantasy, this study provided some contribution in providing empirical evidence that trust relationship is linear. However, this study did found some signs of curvilinear relationship may exist.
5.1. Implications
This study results in several contributions. First of all, this study answered to the calling of the major need of empirical research on optimal trust literature greatly emphasized by Wicks, Berman & Jones (1999). Although this study discovered a linear relationship between trust and partner performance instead of curvilinear, this study did contribute to both perspective. It confirmed that linear relationship still exists in some cases and confirmed that there is a possible sign of curvilinear relationship between the two variables.
Secondly, the interaction model first introduced by Wicks, Berman & Jones involving only optimal trust and interdependence, was extended by this study by adding three additional moderators to be investigated in the model. This study managed to prove a negative moderating role of market uncertainty and confirms that interdependence, asset specificity and behavior uncertainty did not influence the effect of trust invested in interfirm collaboration.
Furthermore, an unpredicted outcome was discovered by this study, as a significantly negative direct relationship between behavioral uncertainty and partner performance was found. This is an insightful evidence since no research tested the direct impact of behavior uncertainty towards partner performance yet. showing that uncertainty in behavior within alliance will directly lower partner performance instead creating an indirect effect as moderator.
28
vice versa. Thus, this study highlighted the critical importance for firms to be aware and responsive towards the external environments, specifically market uncertainty.
In addition, this research also has managerial relevance that can be applicable for use in practice. Firstly, many managers may be unaware that insufficient (underinvest) or excessive trust (overinvest) does exist and leads to detrimental effect on collaboration. This research helps increase the awareness for managers to not take trust for granted and invest in building trust to the right amount. This study also suggests managers to be more sensitive and responsive towards the market conditions. When the market uncertainty is low, it is safe for managers to invest in trust in the relationship, however it is highly recommended that managers keep tracks of the net benefit that generated from the collaboration to avoid the damaging effect of blind trust. When the market uncertainty is high, it is safer for managers not to invest too much in building trust unless they aim for a long term and partner-specific collaborations.
5.2. Limitations and Suggestions for further research
Like all studies, this research also has several limitations that need to be acknowledged. The biggest and the main limitation was the small sample size which also related to time limitation where the data gathering processes were done only in approximately three months. This issue may become one of the contributing factors for insignificant result in the regression analysis in this study. Causing some models that actually statistically significantly fit to be rejected due to the insignificant variance explained. The next limitation is concerning generalizability of the result in this study. Although the sample was quite dispersed, the sample size was considered small and it is often encountered with generalizability issues (Pallant, 2010). There is also issue with single-informant bias where the data is obtained only from one perspective of informant within an alliance According to Adobor (2006) this limitation can indeed make the data less reliable, however, obtaining the data from multiple informants comes with many disadvantages that outweigh the benefit in this kind of research. Nevertheless, series of procedural and statistical remedies were applied to minimize and compromise these limitations.
29
interfirm trust and performance, it is also interesting to find out whether the same result will be discovered if this model was tested in an intra-organizational context. A longitudinal study is is also recommended to be done since doing research in this topic over a long run will provide a more representative result as the development of trust and its effect on performance can be more visible and be captured as the dynamic alliance evolve. As mentioned before, this notion of optimal trust is still in need of more empirical study, hence, there are abundance of opportunity for further research to test and empirically ground this topic.
References
Adobor, H. (2006). Optimal trust? Uncertainty as a determinant and limit to trust in inter-firm alliances. Leadership & Organization Development Journal, 27(7), 537-553.
Aiken, L. S., West, S. G., & Reno, R. R. (1991). Multiple regression: Testing and interpreting
interactions. Sage.
Arrow, K. (1974). The limits of organization. New York: Norton
Atinc, G., Simmering, M. J., & Kroll, M. J. (2012). Control variable use and reporting in macro and micro management research. Organizational Research Methods, 15(1), 57-74.
Atuahene-Gima, K., & Li, H. (2002). When does trust matter? antecedents and contingent effects of supervisee trust on performance in selling new products in china and the united states. Journal of
Marketing, 66(3), 61-81.
Aulakh, P., Kotabe, M., & Sahay, A. (1996). Trust and performance in cross-border marketing partnerships:A behavioral approach. Journal of International Business Studies, 27: 1005–1032.
Bachmann, R., & Zaheer, A. (2006). Handbook of trust research. Cheltenham: Edward Elgar.
Baum, J. A., Calabrese, T., & Silverman, B. S. (2000). Don't go it alone: Alliance network composition and startups' performance in Canadian biotechnology. Strategic management journal, 267-294.
Beccerra, M., & Gupta, A. (1999). Trust Within the Organization: Integrating the Trust Literature with Agency Theory and Transaction Cost Economics. Public Administration Quarterly, 23(2), 177-203.
Bland, J. M., & Altman, D. G. (1997). Cronbach's alpha. BMJ, 314, 572.
30
Cattell, R. B. (1966). The scree test for the number of factors. Multivariate Behavioral Research, 1, 245-276.
Calton, J., & Lad, L. (1995). Social contracting as a trust building process of network governance.
Business Ethics Quarterly, 5: 271-295.
Campion, M. A., Medsker, G. J., & Higgs, A. C. (1993). Relations between work group characteristics and effectiveness: Implications for designing effective work groups. Personnel psychology, 46(4), 823-847.
Chiles, T. H., & McMackin, J. F. (1996). Integrating variable risk preferences. trust. and transaction cost economics. Academy of Management Review, 21(1), 73-99.
Colquitt, J. A., & Zapata-Phelan, C. P. (2007). Trends in theory building and theory testing: A five-decade study of the Academy of Management Journal. Academy of Management Journal, 50(6), 1281-1303.
Cook, R. D., & Weisberg, S. (1982). Residuals and influence in regression. New York, NY: Chapman & Hall.
Cook, T. D., & Campbell, D. T. (1979). Quasi-experimentation: Design and analysis issues for field
settings. Boston, MA: Houghton Mifflin.
Cullen, J. B., Johnson, J. L., & Sakano, T. (2000). Success through commitment and trust: The soft side of strategic alliance management. Journal of World Business, 35(3), 223-240.
Currall, S. C., & Epstein, M. J. (2003). The Fragility of Organizational Trust: Lessons From the Rise and Fall of Enron. Organizational Dynamics, 32(2), 193-206.
Currall, S. C., & Inkpen, A. C. (2006). On the complexity of organizational trust: A multi-level co-evolutionary perspective and guidelines for future research. Handbook of trust research, 235-246.
Das, T. K., Teng, B. S., 2003. Partner analysis and alliance performance. Scandinavian Journal of
Management 19,279–308.
Dillman, D. A., Sinclair, M. D., & Clark, J. R. (1993). Effects of questionnaire length, respondent-friendly design, and a difficult question on response rates for occupant-addressed census mail surveys. Public Opinion Quarterly, 57(3), 289-304.
Dyer, J. H. (1996). Does governance matter? Keiretsu alliances and asset specificity as sources of Japanese competitive advantage. Organization Science, 7(6), 649-666.
Enzle, M, E., & Anderson, S. C. (1993). Surveillant intentions and intrinsic motivation, Journal of
31
Ferrin, D. L., & Dirks, K. T. (2003). The use of rewards to increase and decrease trust: Mediating processes and differential effects. Organization Science, 14: 18–31.
Field, A. (2013). Discovering statistics using IBM SPSS statistics. Sage.
Flores, F., & Solomon, R. C. (1998). Creating trust. Business Ethics Quarterly, 8(2), 205-232.
Galea, S., & Tracy, M. (2007). Participation rates in epidemiologic studies. Annals of
epidemiology, 17(9), 643-653.
Ganzach, Y. (1997). Misleading interaction and curvilinear terms. Psychological Methods, 2(3), 235.
Gargiulo, M., & Ertug, G. (2006). 9 The dark side of trust. Handbook of trust research, 165.
Gelman, A., & Hill, J. (2007). Data analysis using regression and multilevel/hierarchical models. New York, NY: Cambridge University Press.
Ghasemi A, & Zahediasl S. (2012). Normality Tests for Statistical Analysis: A Guide for Non-Statisticians. Int J Endocrinol Metab, 10(2):486-9.
Ghoshal, S., & Moran, P. (1996). Bad for practice: A critique of the transaction cost theory. Academy
of Management Review, 21(1), 13-47.
Gulati, R. (1995). Does familiarity breed trust? the implications of repeated ties for contractual choice in alliances. The Academy of Management Journal, 38(1), 85-112.
Gulati, R., & Singh, H. (1998). The Architecture of Cooperation: Managing Coordination Costs and Appropriation Concerns in Strategic Alliances. Administrative Science Quarterly, 43(4), 781-814.
Gujarati, D. N., & Porter, D. C. (1999). Essentials of econometrics.
Grindley, P., Mowery, D. C., & Silverman, B. S. (1994). SEMATECH and collaborative research: Lessons in the design of high-technology consortia. Journal of Policy Analysis and Management, 13: 723–758.
Harrigan, K. R. (1986). Managing for joint venture success. Lexington, MA: Lexington Books.
Harrigan. K. R. (1988). Joint ventures and competitive strategy. Strategic Management Journal, 9: 141– 158.
Heide, J. B. and R, L. Stump (1995). Performance Implications of Buyer-Supplier Relationships in Industrial Markets: A Transaction Cost Explanation, Journal of Business Research, 32, 57-66.
32
Janssen, O. (2001). Fairness perceptions as a moderator in the curvilinear relationships between job demands, and job performance and job satisfaction. Academy of management journal, 44(5), 1039-1050.
Jeng, D. J. F., & Mortel, C. H. A. (2010, July). Impact of trust, asset specificity, and environmental uncertainty on manufacturing collaborative processes. In Computers and Industrial Engineering (CIE),
2010 40th International Conference on (pp. 1-7). IEEE.
Jones, T. M. (1995). Instrumental stakeholder theory: A synthesis of ethics and economics. Academy of
Management Review, 20:404-437.
Kaiser, H. F. (1974). An index of factorial simplicity. Psychometrika, 39, 32-36.
Kaiser, H. F. (1960). The application of electronic computers to factor analysis. Educational and Psychology Measurement, 20, 141-151.
Kim, H., Sutton, K., & Gong, Y. (2013). Group-based pay-for-performance plans and firm performance: The moderating role of empowerment practices. Asia Pacific Journal of Management, 30(1), 31-52.
Krishnan, R., Martin, X., & Noorderhaven, N. G. (2006). When does trust matter to alliance performance? Academy of Management Journal, 49(5), 894-917.
Kwon, I. W., Suh, T., (2005). Trust, commitment and relationships in supply chain management: a path analysis. Supply Chain Management: An International Journal 10(1),26–33.
Lepper, M. R., & Greene, D. (1975). Turning play into work: Effects of adult surveillance and extrinsic rewards on children's intrinsic motivation. Journal of Personality and Social Psychology, 31, 479-486.
Lin, X., & Germain, R. (1998). Sustaining satisfactory joint venture relationships: The role of conflict resolution strategy. Journal of International Business Studies, 179-196.
Lunnan, R., & Haugland, S. A. (2008). Predicting and measuring alliance performance: A multidimensional analysis. Strategic Management Journal, 29(5), 545-556.
MacCullum, R.C., & Marr, C.M. (1995). Distinguishing between moderator and quadratic effects in multiple regression. Psychological Bulletin, 118, 405-421.
Madhok, A. (2006). Revisiting multinational firms' tolerance for joint ventures: A trust-based approach. Journal of international Business studies, 37(1), 30-43.
Mayer, R., Davis, J., & Schoorman, F. (1995). An Integrative Model of Organizational Trust. The
33
Miller, J. W., Stromeyer, W. R., & Schwieterman, M. A. (2013). Extensions of the Johnson-Neyman technique to linear models with curvilinear effects: Derivations and analytical tools. Multivariate Behavioral Research, 48(2), 267-300.
Min, S., Mentzer, J. T., & Ladd, R. T. (2007). A market orientation in supply chain management.
Journal of the Academy of Marketing Science, 35(4), 507–522.
Mintzberg, H. (1978). Patterns in strategy formation. Management Science, 24: 934–948.
Molina‐Morales, F. X., & Martínez‐Fernández, M. T. (2009). Too much love in the neighborhood can hurt: how an excess of intensity and trust in relationships may produce negative effects on firms. Strategic Management Journal, 30(9), 1013-1023.
Murray, J. Y., Kotabe, M., (2005). Performance implications of strategic fit between alliance attributes and alliance forms. Journal of Business Research 58 (11), 525–1533.
Muthusamy, S. K., & White, M. A. (2006). Does power sharing matter? The role of power and influence in alliance performance. Journal of Business Research, 59(7), 811-819.
Nohria, Nitin,, Eccles,Robert G.,. (1992). Networks and organizations : Structure, form, and action. Boston: Harvard Business School Press.
Nunally, J. C. & Bernstein, I. H. (1978). Psychometric Theory.
O'Driscoll, M. P., & Beehr, T. A. (1994). Supervisor behaviors, role stressors and uncertainty as predictors of personal outcomes for subordinates. Journal of organizational Behavior, 15(2), 141-155.
Pallant, J. (2013). SPSS survival manual. McGraw-Hill Education (UK).
Park, C. W., & Sheth, J. N. (1975). Impact of prior familiarity and cognitive complexity on information processing rules. Communication Research, 2(3), 260-266.
Park, S. H., & Russo, M. (1996). When competition eclipses cooperation: An event history analysis of alliance failure. Management Science, 42: 875–890.
Pfeffer, J., & Salancick, G. R. (1978). The external control of organizations. New York: Harper & Row.
Poppo, L., & Zenger, T. (2002). Do formal contracts and relational governance function as substitutes or complements?. Strategic management journal, 23(8), 707-725.
Poppo, L., Zhou, K. Z., & Li, J. J. (2015). When can you trust “trust”? Calculative trust, relational trust, and supplier performance. Strategic Management Journal.
Pouder R, St. John C. (1996). Hot spots and blind spots: geographic clusters of firms and innovation.
