Do you trust your coproducer's ability?
Do you trust your coproducer's dedication in this coproduction?
Does a partner with high ability but low dedication and a partner with both great ability and dedication mean the same thing to you?
These questions represent the focal party's concerns when selecting a coproducing partner. The rise of coproduction is derived from market uncertainty, such as heterogeneous and ambiguous tasks, infrequent and unpredictable demands, high degrees of customization, and rapidly changing customer preferences (Nightingale, Baden-Fuller, & Hopkins, Reference Nightingale, Baden-Fuller and Hopkins2011; Skilton, Reference Skilton2011). Instead of accumulating resources themselves to face market uncertainty, focal parties turn to coproduction to gain access to other parties' resources. Previous studies mainly focused on the potential partner's ability and suggested how it influences the willingness to work together (Auh, Bell, McLeod, & Shih, Reference Auh, Bell, McLeod and Shih2007; Chen, Tsou, & Ching, Reference Chen, Tsou and Ching2011; Hitt, Dacin, Levitas, Arregle, & Borza, Reference Hitt, Dacin, Levitas, Arregle and Borza2000).
Assessing the ability of a potential partner is necessary; however, the ‘coproduction’ used in this study is not merely the equivalent of collaboration but embodies cospecialization. Cospecialization makes coproducer selection a serious process because a wrong choice endangers the focal party. Cospecialization refers to the use of one party's resource depending on how the other party's resource is utilized (Chi, Reference Chi1994). Thus, cospecialization inevitably induces performance ambiguity (Alchian & Demsetz, Reference Alchian and Demsetz1972; Poppo, Zhou, & Zenger, Reference Poppo, Zhou and Zenger2008). The difficulty in dividing the contribution in coproduction forces the focal party to assess not only the potential coproducer's ability but also the extent of ability utilization (i.e., dedication). The high-dedicated coproducer exerts more ability and offers better performance than that of the low-dedicated coproducer. However, previous research has not explored the importance of dedication in the coproduction or trust-related literature. Thus, assessing and having trusting belief in the potential partner's dedication are emphasized in this study, with equal importance given to ability.
Moreover, coproduction may occur in interdisciplinary work, in which coproducers belong to heterogeneous groups, such as different industries. As the focal party does not know the coproducer before the first coproduction, the focal party relies on reputation to solve the problem of information asymmetry. Reputation involves an estimation of one's skills and other attributes necessary to the exchange (Fombrun, Reference Fombrun1996). Previous studies also assert that positive reputation induces cooperation (Connelly, Certo, Ireland, & Reutzel, Reference Connelly, Certo, Ireland and Reutzel2011; Doney & Cannon, Reference Doney and Cannon1997; Ganesan, Reference Ganesan1994; Kewell, Reference Kewell2007; Lester, Mencl, Maranto, Bourne, & Keaveny, Reference Lester, Mencl, Maranto, Bourne and Keaveny2010; Money, Hillenbrand, Day, & Magnan, Reference Money, Hillenbrand, Day and Magnan2010; Wagner, Coley, & Lindemann, Reference Wagner, Coley and Lindemann2011). By assessing reputation, the focal party may gain an understanding of the underlying quality of the potential partner. However, few works have addressed the source of reputation, while the importance of the reputation effect has been highlighted.
A comprehensive definition of reputation involves various sources of reputation, and the focal party interprets the sources differently. For example, certified partners and recommended partners may suggest different kinds of bonds between the focal party and the potential partner. Certification is a form of proof provided by an impartial party (Connelly et al., Reference Connelly, Certo, Ireland and Reutzel2011; Lester et al., Reference Lester, Mencl, Maranto, Bourne and Keaveny2010), and the bond between the certified partner and the focal party may be the trust in the third party's professionalism. However, the relationship between the recommended partner and the focal party may be emotional, such as friendship. Different kinds of bonds imply different perceptions of the information. Therefore, recognizing the source of reputation warrants our attention.
This paper focuses on the less addressed viewpoint, examining, in one study, the effect of reputation from different sources. Specifically, the study investigates the motivation to work with a potential coproducer based on the mediating effects of trust in the potential coproducer's ability and dedication, as derived from different reputational sources (i.e., certification, word-of-mouth, and recommendation). The analysis focuses on the initial coproducer selection, specifically, the period before engagement in the long-term coproducing relationship (i.e., the motivation to coproduce). The results support the view that reputation has different impacts on the focal party's trusting belief in the attributes of the potential partner.
Literature Review
Context of trust: coproduction
Coproduction is a new paradigm in the management literature, allowing participants to share resources in joint work. Participants may gain access to new resources and abilities through coproduction without having to build competency in these areas. Thus, coproduction provides not only flexibility in responding to demand uncertainty (Skilton, Reference Skilton2011) but also an efficient way to improve quality (Smith, Reference Smith1963).
The coproduction first appeared in Kogut (Reference Kogut1986)'s article. Kogut (Reference Kogut1986) used coproduction to explain the self-enforcing contract of East-West trade, and the bilateral asset-specific investments were emphasized in that case. Hennart (Reference Hennart1989) then used the term of transaction-specific investments in reference to coproduction.
Seemingly, coproduction in the sense used here is not simply a term for collaboration or team production. Although similar terms refer to coproduction, it may be best understood as a cospecialized context: the specialized cooperation with bilateral dependence. Chi (Reference Chi1994) followed the term Teece (Reference Teece1986) used and explained that cospecialization refers to the use of one party's resource that depends on how the other party's resource is utilized sequentially. Therefore, cospecialization implies that the coproducer's work must follow the states that the focal party provides. Hence, cospecialization in coproduction makes coproduction unique to other forms of collaboration.
Cospecialization in coproduction can be thought of as two attributes from the literature: asset specificity and performance ambiguity. First, coproduction involves the essence of asset specificity, which may not necessarily occur in a team production. Asset specificity infers that the investments in this coproduction are valueless outside the relationship, and the party that leaves the relationship may suffer from the loss of this asset-specific investment (Ganesan, Reference Ganesan1994; Kogut, Reference Kogut1986). Asset-specific investments are present in many forms, such as the accumulation of knowledge or the adaptation of a particular working process. As Kogut (Reference Kogut1986) stated, the production of firm one is valueless without the product information or nonfungible capital assets provided by firm two. Thus, coproducing parties are inclined to maintain a long-term relationship to avoid potential losses (Fenwick& McMillan, Reference Fenwick and McMillan2013; Skilton, Reference Skilton2011).
Second, coproduction may engender performance ambiguity, which derived from bilateral dependence. Performance ambiguity refers to the difficulty in attributing the contribution of each firm to output (Alchian & Demsetz, Reference Alchian and Demsetz1972; Poppo, Zhou, & Zenger, Reference Poppo, Zhou and Zenger2008), which may be similar to the problems confronted by a team production. Specifically, Alchian and Demsetz (Reference Alchian and Demsetz1972) suggested that successful exchange happens when rewards to productivity are linked effectively. That is, the productivity of each contributor can be measured and rewarded in accord with productivity. However, mutual dependence on resources prevents the focal party from being able to accurately measure and reward productivity. Coproducers thus have incentives to limit their efforts, leading to lower performance or an overcharging situation. In this case, the top priority of focal parties is to understand whether the coproducers did their best.
Moreover, poor coproducing output cannot be traced to a specific party due to the interrelated coproducing processes. Performance ambiguity in coproduction thus prevents the parties from being able to divide responsibility and thus have to solve the unexpected problem together. If the coproducers are unwilling to exert themselves to fix the problem, the irresponsible behavior may place the focal party at risk. Hence, coproducing with those characterized by high-propensity dedication may be a reasonable solution to the problems caused by performance ambiguity.
Considering the potential risk and loss, selecting a trustworthy coproducer ex-ante may diminish the potential failure costs (Larson, Reference Larson1994). Hence, it is reasonable to assume that the focal party regards coproducer selection as essential to a desirable output and they approach it with caution.
Dimensions of trust in coproduction: ability and dedication
Trust is defined as the willingness of a party to be vulnerable to the actions of another party (Mayer, Davis, & Schoorman, Reference Mayer, Davis and Schoorman1995); it has been found to be a multidimensional construct (Butler, Reference Butler1991). Thus, the trustor may have trusting belief in one aspect of the trustee but not necessarily have trusting belief in the other aspect of the trustee. For example, a competent one may be an influential mentor in the field. The ability is assured; however, it is doubtful if he is willing to deliver his knowledge to the protégé. To elaborate this argument, this study starts by introducing ability as one of the dimensions Mayer, Davis, and Schoorman (Reference Mayer, Davis and Schoorman1995) identified.
Mayer, Davis, and Schoorman (Reference Mayer, Davis and Schoorman1995) referred to ability as the group of skills and competencies that enable a party to have influence within some specific domain. Specifically, trusting belief in ability can be understood as the perception of the thoroughness with which the trustee approaches their jobs or the abundance of knowledge possessed by the trustee (McKnight, Choudhury, & Kacmar, Reference McKnight, Choudhury and Kacmar2002a). Different terms with similar meanings have also appeared in prior studies: resources (Kogut, Reference Kogut1986), competence (Butler & Cantrell, Reference Butler and Cantrell1984; Levin & Cross, Reference Levin and Cross2004; McKnight, Cummings, & Chervany, Reference McKnight, Cummings and Chervany1998), partner expertise (Auh et al., Reference Auh, Bell, McLeod and Shih2007; Chen, Tsou, & Ching, Reference Chen, Tsou and Ching2011), and task reliability (Sitkin & Roth, Reference Sitkin and Roth1993).
McKnight, Cummings, and Chervany (Reference McKnight, Cummings and Chervany1998) suggested that in the initial relationship, trust intention is formed by the trusting belief in ability. Mayer and Davis (Reference Mayer and Davis1999) proved that the perception of ability leads to trust in top management. Moreover, the vital role of a trusting belief in ability is noted by McKnight, Choudhury, and Kacmar (Reference McKnight, Choudhury and Kacmar2002a). They summarized 32 trust-related studies and found 17 articles that mentioned the trusting belief in ability. Following this argument, Kelton, Fleischmann, and Wallace (Reference Kelton, Fleischmann and Wallace2008) addressed the importance of trusting belief in ability and argued that this belief is the basis of confidence in the ability of the trustee to fulfill the trustor's needs. Further, Kim and Han (Reference Kim and Han2009) suggested that a trusting belief in ability assuages focal parties' concerns about the competence of the trustee, which facilitates continuous contact with the trustee. Song and Zahedi (Reference Song and Zahedi2007) proved that the focal party's trusting belief in ability is positively associated with the intention to use the platform provided by the trustee. Moreover, the trusting belief in ability was proved to positively affect the willingness to purchase by assuring the focal party of the trustee's competence (McKnight, Choudhury, & Kacmar, Reference McKnight, Choudhury and Kacmar2002b).
In the coproduction literature, previous studies also substantiated the positive relationship between ability and the incentive to engage in coproduction (Auh et al., Reference Auh, Bell, McLeod and Shih2007; Chen, Tsou, & Ching, Reference Chen, Tsou and Ching2011). Following Chen, Tsou, and Ching (Reference Chen, Tsou and Ching2011)'s statement that the success of coproduction is derived from the ability the coproducer brings to the partnership, this study included ability as one of the key factors. The trusting belief in ability is defined as the focal party's trusting belief in the skill and knowledge possessed by the potential coproducing partner (i.e., the coproducer) to complete the task.
Dedication is another critical factor emphasized in this study, referring to the effort that the coproducer decides to exert on tasks. The concept of dedication is complementary to the concept of ability due to its inability to describe the extent to which the coproducers use their abilities. A partner of high ability but low dedication may provide what is required by the focal party; however, a partner of great ability and dedication may possibly perform beyond the focal party's expectations. To amend the explanatory power of ability, the concept of dedication was distinguished and emphasized.
Though dedication has been less frequently addressed in the coproduction or trust-related literature, it has been inferred in the psychology and managerial fields. Yeo and Neal (Reference Yeo and Neal2004) used a similar term, effort, to suggest that dedication is positively associated with performance. The positive relationship between effort spent and performance is also referred to in De Jong and Elfring (Reference De Jong and Elfring2010)'s work. Moreover, Graves, Ruderman, Ohlott, and Weber (Reference Graves, Ruderman, Ohlott and Weber2012) suggested that highly dedicated partners are prone to exhibit superior performance. Considering performance ambiguity in a coproducing context, the influence of dedication on performance, and the insufficient explanatory power of ability, the introduction of dedication to the coproduction or trust-related literature may explain the concerns of the focal party in the coproducing relationship and fill the void of existing constructs.
High dedication implies the willingness to exert effort with respect to task performance and work in a highly cooperative way. Thus, dedication may indicate strong involvement in one's work (Schaufeli, Shimazu,& Taris, Reference Schaufeli, Shimazu and Taris2009), demonstrating motivation to exert effort even in the face of setbacks (De Jong & Elfring, Reference De Jong and Elfring2010) or the willingness to make specific investments (Bensaou & Anderson, Reference Bensaou and Anderson1999).
Dedication may stem from intrinsic interest, such as the feeling of responsibility to complete one's job (Graves et al., Reference Graves, Ruderman, Ohlott and Weber2012), job recognition, work engagement (Schaufeli, Shimazu, & Taris, Reference Schaufeli, Shimazu and Taris2009), self-discipline (Van Scotter & Motowidlo, Reference Van Scotter and Motowidlo1996), or the honor of participating. Therefore, coproducers may have an egocentric profit motive to improve the performance of a joint work because a successful coproduction may bring them more job opportunities. Or they may base their work on beliefs such as job recognition and self-fulfillment. In either case, the primary intention is not to benefit the focal party, but it ultimately sustains or even improves the performance.
Hence, this study separates the definition of dedication from benevolence; Mayer, Davis, and Schoorman (Reference Mayer, Davis and Schoorman1995) suggested that benevolence is the trait of a trustee who wants to help the trustor for reasons other than the egocentric profit motive. Dedication benefits the coproduction outcome and is welcomed by the focal party, although the reason for the dedication does not fit within the bounds of benevolence. Thus, deliberation upon performance improvement may have a higher priority for the focal party than determining the intentions of coproducers. Accordingly, the core consideration of the trusting belief in dedication is the extent of the coproducer's effort regardless of altruism or egoism.
Moreover, integrity is not emphasized in this study, although Mayer, Davis, and Schoorman (Reference Mayer, Davis and Schoorman1995) mentioned it as the third dimension of trust. Following the previous literature, perceived integrity derives from the perception of the trustee's sense of justice (Mayer, Davis, & Schoorman, Reference Mayer, Davis and Schoorman1995), fairness toward others (Mayer & Davis, Reference Mayer and Davis1999), and their moral characteristics (Butler, Reference Butler1991; Butler & Cantrell, Reference Butler and Cantrell1984). However, the context of this research emphasizes selecting an initial coproducer, which implies that the coproducing relationship has not started yet. These definitions of integrity seem more applicable to an existing relationship, for example, discussing the employee's intention to quit or stay in the company. Further, Butler and Cantrell (Reference Butler and Cantrell1984)'s empirical study showed that ability is significantly stronger than integrity in determining trust. Butler (Reference Butler1991) then suggested that some trust dimensions may be more crucial than other dimensions in certain conditions. After considering the relatedness in this coproducing context and the relative importance of ability and integrity, the aspect of integrity is thus not emphasized in this study.
In sum, performance constitutes the ability that each party brings to this relationship (Chen, Tsou, & Ching, Reference Chen, Tsou and Ching2011) and how it is used. Since the indigenous essence of coproduction implies the uncountable share of the contribution from each party, it is reasonable to concentrate on the dimension of ability and an added dimension, dedication, when discussing coproducer selection.
Process of trust: signaling-based trust and trust transference
The focus of this study is investigating coproducer selection for first-time selectors who have had no previous interaction with the coproducer; therefore, the literature on initial trust is the key to our discussion.
McKnight, Cummings, and Chervany (Reference McKnight, Cummings and Chervany1998) theorized that initial trust is affected by three factors – personal-based trust, institution-based trust, and cognitive-based trust. Personal-based trust is the propensity of trustors to trust others, which is developed during their early life experience. Institution-based trust describes trust as reflecting the security the trustors feel about a situation because of formal rules and procedures defined by a legal structure (Zucker, Reference Zucker1986). Cognitive-based trust is grounded in the trustors' initial perceptions or received cognitive cues (Meyerson, Weick, & Kramer, Reference Meyerson, Weick, Kramer, Kramer and Tyler1995). These processes of trust have shown their influence in the existing literature and infer the active role of the trustor in the trustor–trustee relationship. It seems that the trustees wait passively for trustors to trust them and are unable to influence circumstances.
However, when the relationship has not started, and the information between the focal party (i.e., trustor) and the coproducer (i.e., trustee) is asymmetrical, the coproducer may want to win the trust spontaneously. Considering the active role of the coproducer, the foundation of signaling behavior lies in the unequal possession of information between two parties, and signaling theory has been used to explain decision-making in many studies (Connelly et al., Reference Connelly, Certo, Ireland and Reutzel2011; Hopp & Lukas, Reference Hopp and Lukas2014; Lester et al., Reference Lester, Mencl, Maranto, Bourne and Keaveny2010; Spence, Reference Spence1973; Wagner, Coley, & Lindemann, Reference Wagner, Coley and Lindemann2011); this study thus intends to connect signaling theory with the process of trust.
Spence (Reference Spence1973) used ‘buying a lottery ticket’ to describe a situation where the focal party does not know the partner's capability until the partner is hired. To solve the problem of information asymmetry, Spence (Reference Spence1973) suggested that the existence of signals and defined signals is observable and alterable. The signal receivers (i.e., the trustors or the focal party in this study) can receive the signal, and the signalers (i.e., the trustees or the potential partner in this study) can show their quality by sending signals with a cost. Moreover, the signaling cost is assumed to be negatively correlated to productivity in the discrimination of good ones from bad ones. That is, the low-quality ones have a higher cost of signaling quality. The negative correlation between signaling cost and productivity leaves only the low-quality ones, which infers that signaling behavior can only be achieved by the good ones. On the one hand, high-quality parties use signaling to show underlying quality and to separate themselves from low-quality parties. On the other hand, the less informed party obtains the information they do not know by interpreting the signal.
Using trust terminology, a trustee may actively attract the trustor's attention through signaling. Signaling preparation infers not only the determination of the trustees to win the other party's trust but also confidence in their underlying qualities. As the less informed party, the trustors understand that the cost of signaling is negatively related to the underlying qualities. Low-quality ones are unable to signal; thus, trustors may have a positive opinion or feeling about the qualities emphasized by the trustee after the signal is sent. Moreover, costly signaling guarantees the underlying quality of the trustee and instills further trust in the trustee. In the end, signaling-based trust is produced through the interpretation of signals.
Furthermore, the transference of trust is also discussed in this paper, as the focus of this study is a new coproduction between newly established partners. Doney, Cannon, and Mullen (Reference Doney, Cannon and Mullen1998) suggested that trust may develop from a known entity to an unknown one. They reasoned that the trustor (i.e., the focal party) might base its outlook on ‘proof sources’ that transfer trust to the trustee, even if the trustor has little or no direct experience with the trustee. Moreover, trust may transfer from different kinds of sources. A prior study argues that transfer may be made from an industry association to an individual (Milliman & Fugate, Reference Milliman and Fugate1988). In such situations, trust transference rests on the assumption that institutions act as trustworthy proof sources. To illustrate, the industry association reflects the actions of the people involved. Thus, beliefs about the institutions help to form beliefs about the people who are involved in the institutions. That is, if the focal party trusts the institution (i.e., the reputational information provider) to have strict examining standards or procedures, the trusting belief in the professionalism of the institution will transfer to the trust in the potential partner's professionalism.
Other studies focus on individuals (i.e., known targets) who serve as the source of trust transfer to unknown targets as represented by social networks (Doney, Cannon, & Mullen, Reference Doney, Cannon and Mullen1998). A social network is constructed of nodes (actors) and paths (relationships), in which paths link pairs of nodes and then string together a network. A social network may be formed from a group of people with similar attributes (e.g., personal values, interests, or occupations). Within the social network, network members share opinions about objects and sources of information they interact with (Zucker, Reference Zucker and Bacharach1983). Although each member has very limited information, the linkage allows actors to connect with others or exchange resources (e.g., information and affect) (Tichy, Tushman, & Fombrun, Reference Tichy, Tushman and Fombrun1979). Based on the assumption of trust transference, the bestowment of trust on a trustee rests on the trustworthiness of individuals closely associated with the trustee (Stewart, Reference Stewart2003). Thus, not only the information will transfer in the network but also trust. The focal party may make decisions by collecting and combining information from other members within a social network. The interactions between network members trigger an aligned network of trust in the end.
Hypotheses development
The focus of this study is to examine partner selection for newly begun coproductions that tend not to have established partners. However, there are plenty of potential candidates that have never worked together, and it is impossible to review all the candidates. Thus, focal parties may simplify the selection process by making decisions based on reputation.
Reputation is ‘a representation of a company's past action and prospects that describe the firms' overall appeal to all of its key constituents when compared with other leading reveals’ (Fombrun, Reference Fombrun1996). The favorability of reputation has been proven to affect the focal party's trusting belief in a potential partner (Doney & Cannon, Reference Doney and Cannon1997; Ganesan, Reference Ganesan1994; McKnight, Cummings, & Chervany, Reference McKnight, Cummings and Chervany1998; Wagner, Coley, & Lindemann, Reference Wagner, Coley and Lindemann2011). Although published studies have discussed the reputation effect, how the reputation was assessed has not been discussed. Considering the potential influence of reputation sources, this study divides reputation into three categories: certification, word-of-mouth, and recommendation. Figure 1 illustrates the conceptual model with hypotheses regarding the relationships between reputation sources, trusting belief, and motivation to coproduce with a potential partner.
Figure 1. Conceptual model and theoretical reasoning
Certification
Certification from a well-known nonstakeholder has been discussed as a strong signal of high quality (Connelly et al., Reference Connelly, Certo, Ireland and Reutzel2011; Lester et al., Reference Lester, Mencl, Maranto, Bourne and Keaveny2010; Spence, Reference Spence1973). Certification pre-examines the candidate's required capability; therefore, the potential partner may through professional credentials serve as a clear signal of role preparedness (Lester et al., Reference Lester, Mencl, Maranto, Bourne and Keaveny2010; Spence, Reference Spence1973). Moreover, passing the certification gives the focal parties greater confidence in potential partners' skills or knowledge in a professional standard. Hence, the one path of trust production is signaling-based trust. As Kim (Reference Kim2008) suggested, one develops trust when others impress that person with their professional training and experience; it is reasonable to indicate that the signaling from certification produces trusting beliefs in their ability.
The other path of trust production is trust transference. The focal parties may trust the certification provider due to professionalism or perceived information reliability. The focal party does not need to test the potential partner's ability because professional associations and credentialing agencies provide guarantees through certification (i.e., acceptance in their professional community). Hence, the trusting belief in ability will transfer from the trusted professional third party to the potential certified partner.
As Fishbein and Ajzen (Reference Fishbein and Ajzen1975) suggested, the belief directly affects the intention, and McKnight, Cummings, and Chervany (Reference McKnight, Cummings and Chervany1998) linked trusting belief to trusting intention; this study suggests trusting beliefs in ability derived from signaling and transference will lead to the motivation to coproduce with the candidate.
Therefore,
Hypothesis 1
Third-party certification leads to the motivation to coproduce with a potential partner through its effect on trusting beliefs in ability.
Coproduction is known as interrelated usage of resources; therefore, coproduction requires participating parties to solve problems together. For instance, when the output does not comply with the customer's need, it requires high dedication for modifications. If the coproducer refuses to make the modifications, the focal party may be unable to shirk the responsibility and inevitably receives a negative evaluation from the customer. Accordingly, the focal party may want to distinguish highly dedicated candidates from those trying to shirk.
Thus, the trusting belief in dedication is mainly produced from signaling-based trust. Connelly et al. (Reference Connelly, Certo, Ireland and Reutzel2011) noted that the process of obtaining a certification (e.g., ISO 9000) is time consuming. Supporting this notion, this study suggests certification may also signal a guarantee of high dedication. Requiring certification costs time and money; therefore, only those who take their job seriously may regard acquiring certification as important. On the one hand, potential partners show the determination that they want to be set apart from other candidates in the market by taking the certification examination. On the other hand, passing the certification becomes an important signal to focal parties regarding the candidate's value and respect for this job.
In sum, the certified potential partner signals as a high-dedicated candidate and implies more willingness to complete the job. Based on the signaling statement, hypothesis 2 is thus derived.
Hypothesis 2
Third-party certification leads to the motivation to coproduce with a potential partner through its effect on trusting beliefs in dedication.
Word-of-mouth
Word-of-mouth is the perceptual representation of a party's past actions that describes its overall appeal to its key constituents and is central to human decision-making (Kewell, Reference Kewell2007). Wagner, Coley, and Lindemann (Reference Wagner, Coley and Lindemann2011) suggested that the received signal regarding a potential partner's word-of-mouth at the very beginning of the selection process positively influences the focal party's willingness to collaborate in future. The academic literature also confirms that positive word-of-mouth about a potential partner at the start of a project is, thereby, linked directly to trust (Doney & Cannon, Reference Doney and Cannon1997; Ganesan, Reference Ganesan1994; McKnight, Cummings, & Chervany, Reference McKnight, Cummings and Chervany1998; Money et al., Reference Money, Hillenbrand, Day and Magnan2010; Wagner, Coley, & Lindemann, Reference Wagner, Coley and Lindemann2011).
The main process of trust from word-of-mouth here is signaling-based trust. Word-of-mouth is regarded as a signal because it requires time and effort to build up. Moreover, word-of-mouth signals adequate ability and dedication to delivering products or services that meet the promised terms. Although word-of-mouth builds over a long period of time, it ruins quickly. Thus, maintaining word-of-mouth signals recognition by and dedication to one's job. Compared to the high-dedicated candidates, the low-dedicated candidates evaluate the costs and benefits of signaling and are unwilling to sustain the payment in the long term. Following the signaling argument, word-of-mouth may segregate low-quality candidates, leaving potential partners with adequate ability and dedication in the review process, which leads to the motivation for coproduction. Therefore,
Hypothesis 3
Word-of-mouth leads to the motivation to coproduce with a potential partner through its effect on trusting belief in ability.
Hypothesis 4
Word-of-mouth leads to the motivation to coproduce with a potential partner through its effect on trusting belief in dedication.
Recommendation
To distinguish the concept of recommendation from word-of-mouth in this study, the term recommendation is defined as a reference or suggestion from friends, colleagues, and acquaintances. Thus, the influence of newspapers and printed advertisements is not included in this category. Brown and Reingen (Reference Brown and Reingen1987) used a similar term, referral behavior, to describe how recommendations travel around a network through the linkages between network members.
Hence, the process of trust from recommendation here is the trust transference in a social network. A typical example of trust transference in a network is a focal party that trusts someone who is recommended by respected colleagues. From the perspective of the focal parties, they choose the trustworthy reputation source when information is not easy to verify. If the trustee is perceived to be related to the trusted source, trust is then transferred to another unknown trustee (Doney & Cannon, Reference Doney and Cannon1997; Doney, Cannon, & Mullen, Reference Doney, Cannon and Mullen1998; Kim, Reference Kim2008; Lim, Sia, Lee, & Benbasat, Reference Lim, Sia, Lee and Benbasat2006; Stewart, Reference Stewart2003). Accordingly, the process of transference can be used to predict a positive relationship between a recommendation from a close entity and trust in the recommended candidate.
Social network members tend to share common goals and values (Zucker, Reference Zucker and Bacharach1983); therefore, members within the same network may understand the practical requirements of a coproducing relationship. The focal party may believe that the potential partner had adequate ability to accomplish the task, given they are recommended from a reliable network member. Therefore,
Hypothesis 5
Third-party recommendations lead to the motivation to coproduce with a potential partner through its effect on trusting beliefs in ability.
In addition to similar values and goals, social network members hold common beliefs about appropriate behavior in a coproducing relationship. Members within the same network are assumed to provide information in line with the focal party's needs and preferences. The focal party regards the social network member as a reliable third party who understands the practical requirements of the relationship. Therefore, the trusting belief in the social network member will transfer to the recommended potential partner's dedication. The trusting belief in the dedication of the recommended potential partner then further enhances the motivation to participate in coproduction. Therefore,
Hypothesis 6
Third-party recommendations lead to the motivation to coproduce with a potential partner through its effect on trusting beliefs in dedication.
Methods
Sample and data
We chose dental prosthesis fabrication as the empirical setting for two reasons. First, this is a typical coproduction scenario (Davenport, Basker, Heath, Ralph, Glantz, & Hammond, Reference Davenport, Basker, Heath, Ralph, Glantz and Hammond2000), starting with the dentist impressing a patient's oral cavity onto a mold, while the dental technician completes the denture according to the model. However, the success of the output relies on an accurate impression (Lynch & Allen, Reference Lynch and Allen2005; Radhi, Lynch, & Hannigan, Reference Radhi, Lynch and Hannigan2007; Smith, Reference Smith1963), and the dentist's particular habits (i.e., the gesture of an impression) may lead to inaccuracies. The dental technician may need to use their knowledge and understanding about the dentist to correct minor inaccuracies (Haj-Ali, Al Quran, & Adel, Reference Haj-Ali, Al Quran and Adel2012; Radhi, Lynch, & Hannigan, Reference Radhi, Lynch and Hannigan2007). Understanding a particular partner takes time and effort to accumulate specific knowledge, which is regarded as an asset-specific investment. Second, the dentist has a responsibility to the public and the profession in safeguarding quality (Henderson, Reference Henderson1966), even though dental prosthesis fabrication is a joint process. The responsibility, asset-specific investment, and performance ambiguity problem make partner selection crucial for the dentist.
We conducted qualitative and quantitative analyses to gain a deeper understanding of how reputation works within coproducing behavior. In the first qualitative stage, a semistructured interview was conducted with two experienced dentists and two coproducers. They were asked to describe the working process and the interactions with their coproducers. In the second qualitative stage, four dentists and two researchers specializing in interorganization relationships helped us correct our conceptual model for further empirical testing. They were also asked to identify any ambiguities or misunderstandings in the questionnaire. Based on their feedback, some minor changes of wording were made to help the respondent understand the academic terms in their language. For instance, it may be difficult for the respondent to understand the academic term ‘coproducer,’ that is, the partner in the coproducing relationship. Thus, instead of providing a long description, replacing the term ‘coproducer’ with the term ‘partner’ may be easier for respondents to answer.
In the quantitative stage, we excluded public-owned clinics because the selection of dental technicians from such clinics is restricted by law. However, the flexibility to select a coproducer freely was particularly crucial in this research context. Thus, only self-owned clinics were included in the sample. Moreover, to ensure respondents had an adequate ability to make selections, only the dentists in charge were chosen. That is, when more than one dentist was working in the same dental clinic, only the one in charge received our questionnaire.
To minimize selection bias and to be more representative, the sample was systematically selected from registered self-owned dental clinics in Taiwan. A total of 1,000 questionnaires were sent out; 13 (1.3%) were undeliverable because of incorrect addresses. Of the remaining 987 delivered questionnaires, 109 completed ones were received. This response rate is considered to be acceptable in Taiwan, where the average response rate for similar survey studies is around 10% (Harzing, Reference Harzing2000). The majority of respondents were male (95%) with an average tenure in the dental industry of 25.92 years (SD = 8.10). According to the statistics, the respondents had worked with their current technicians for an average of 15 years (SD = 8.57). They reported that, including themselves, an average of 1.72 dentists (SD = 1.68) served in their dental clinics, which may imply the respondent's ability to make decisions without interference from others. Interestingly, they had worked with an average of 2.09 technicians (SD = 2.12) over their service years. To be more specific, in a dentist's 26 years of experience, only three dental technicians (including the current one) had coproduced with them. This partially supports our assertion in the research context that initial coproducer selection is important, and asset-specific investment means that a dentist rarely replaces their coproducer.
To ensure the analyses were not biased by nonresponses and that the sample was representative of the industry, early and late respondents were compared using an independent samples t-test. The results indicated no significant differences between the attributes of two response groups, including firm size (p = .74; t = 0.33) and years with current coproducer (p = .71; t = 0.37). In addition to procedural controls, such as anonymous submission, a short questionnaire, and minimization of the ambiguity of the measurement items, Harman's single-factor test was performed to check for common method bias. The result shows the first factor accounted for <50% of the total variance, which suggests that common method bias is an unlikely concern in the data.
Measures
We adapted our measures from previous studies based on the relevant literature and the interview results. Some modifications were necessary to accommodate the research context, and the respondents were asked to answer questions based on their selection experiences, rating on a 5-point Likert scale. All multi-item constructs showed good discriminant validity based on factor analysis, using scree plots of eigenvalues, and all expected factor loadings are above 0.50. With all Cronbach's α above 0.70, constructs exhibit good convergent validity as well. Multi-item variables were based on an unweighted average of relevant items.
For the source of reputation, certification was assessed by the respondent's perception of the certification the potential partner possessed. The first two items were adapted from Lim et al. (Reference Lim, Sia, Lee and Benbasat2006) for the linkage with the certification, and the last two items were based on the interview result and the concept from the signaling literature (Connelly et al., Reference Connelly, Certo, Ireland and Reutzel2011; Lester et al., Reference Lester, Mencl, Maranto, Bourne and Keaveny2010). Word-of-mouth was gauged by the extent to which the respondent knew about the potential partner, and four items were from Pirson and Malhotra (Reference Pirson and Malhotra2011) and Ganesan (Reference Ganesan1994). Recommendation measured the perception of the potential partner's recommendation by a social network member, and three items were adapted from Lim et al. (Reference Lim, Sia, Lee and Benbasat2006).
Consistent with previous studies (Auh et al., Reference Auh, Bell, McLeod and Shih2007; Mayer and Davis, Reference Mayer and Davis1999), trust in ability was measured by the extent of the trusting belief in the potential partner's ability after receiving reputational information, and six items were used. As no prior studies have examined trusting belief in dedication, we select items most related to the interview results. Six items were selected from Bensaou and Anderson (Reference Bensaou and Anderson1999) and De Jong and Elfring (Reference De Jong and Elfring2010): two items related to the concept of dedication, two items associated with the coproducer's attitude toward the coproduction, and two items related to the degree of specific investment in the coproduction. The motivation for coproduction was measured by asking the respondent the extent to which they considered cooperating with the candidate (Doney & Cannon, Reference Doney and Cannon1997; Lim et al., Reference Lim, Sia, Lee and Benbasat2006).
Two business attributes, such as tenure and firm size, were included as control variables to ensure their effects did not confound the measures of the proposed relationships. The more experienced a focal party is, the more thoughts it has regarding coproducer selection. Accordingly, tenure was used as a control variable owing to its reflection of the focal party's accumulated experience. Consistent with prior studies (Finkelstein & Hambrick, Reference Finkelstein and Hambrick1990; Zenger & Lawrence, Reference Zenger and Lawrence1989), tenure was measured by the number of years the respondent had served in the industry. Regarding the measurement of firm size, the literature suggests that larger firms may have more resources and bargaining power (Jarvenpaa, Tractinsky, & Vitale, Reference Jarvenpaa, Tractinsky and Vitale2000), which could potentially influence the logic of decision-making. Thus, firm size was included and determined by the number of dentists serving in the same clinic.
Results
Table 1 shows the means, standard deviations, internal reliability, correlations, and the significant bivariate relationships between our independent and dependent variables. The positive correlation between three sources of reputation (word-of-mouth, certification, and recommendation) and the motivation for coproduction confirms the face validity that reputation is essential in the coproducing context.
Table 1. Descriptive statistics for study variablesa
a Internal reliabilities are presented along the diagonal in parentheses.
Two-tailed tests; *p < 0.05; **p < 0.01; ***p < 0.001.
We tested our hypotheses using the Hayes (Reference Hayes2013) PROCESS macro, which provided the coefficients from ordinary least squares (OLS) analysis and an inferential test. The former helped us to follow Baron and Kenny (Reference Baron and Kenny1986) criteria, and the latter was used to reduce the probability that type II error occurs when mediation is unable to be claimed once the parameters fail to meet the criteria (Preacher and Hayes, Reference Preacher and Hayes2004). We also followed the suggestion of Hayes (Reference Hayes2013) to report the coefficients in unstandardized form.
As predicted by Hypothesis 1 and 2, Table 2 shows certification had a positive effect (c = 0.39; Equation [1]) on the motivation to coproduce with the potential partner. The focal party had trusting belief in a certified potential partner's ability (a = 0.23; Equation [10]), and trust in their ability increases the motivation to coproduce with them (b = 0.51; Equation [2]). A bias-corrected bootstrap 95% confidence interval for the indirect effect (ab = 0.11), based on 10,000 bootstrap samples, was entirely above zero (0.030–0.288). This result supports Hypothesis 1; certification indirectly influenced the motivation for coproduction with the potential partner through its effect on trust in the potential partner's ability. As for Hypothesis 2, the significant statistic result (see Equation [13]) shows the focal party also had trusting belief in a certified potential partner's dedication (a = 0.20), and trust in their dedication increases the motivation to coproduce with them (b = 0.38; Equation [3]). Moreover, based on 10,000 bootstrap samples, the bias-corrected bootstrap 95% confidence interval for the indirect effect (ab = 0.07) was entirely above zero (0.009–0.241). Hypothesis 2 also passed the two tests for mediation; certification indirectly influenced the motivation for coproduction with the potential partner through its effect on trust in the potential partner's dedication. The result suggested the linkage between signaling and trust, in which certification was signaling well as a candidate's capacity to create a level of trusting belief in their ability and dedication.
Table 2. OLS regression resultsa
a Unstandardized coefficients shown with standard errors in parentheses.
*p < .05; **p < .01; ***p < .001.
As predicted by Hypothesis 3 and 4, word-of-mouth had a positive effect (c = 0.35; Equation [4]) on the motivation to coproduce with the potential partner. The focal party had trusting belief in a positive word-of-mouth potential partner's ability (a = 0.13; Equation [11]), and trust in their ability increases the motivation to coproduce with them (b = 0.67; Equation [5]). A bias-corrected bootstrap 95% confidence interval for the indirect effect (ab = 0.09), based on 10,000 bootstrap samples, was entirely above zero (0.007–0.226). These results support Hypothesis 3; word-of-mouth leads to the motivation to coproduce with a potential partner through its effect on trusting beliefs about their ability. The insignificant effect (see Equation [14]) suggests that the focal party did not have trusting belief in the dedication (a = 0.11), although the effect of trust in their dedication to the coproducing motivation (b = 0.51; Equation [6]) is significant. The 10,000 bootstrap samples with a bias-corrected method also showed the 95% confidence interval for the indirect effect (ab = 0.06) contains zero (−0.002 to 0.179). There is no evidence to support Hypothesis 4; although word-of-mouth may be a good signal to represent the candidate's ability, the statistical result is unable to state that positive word-of-mouth leads to the motivation to coproduce with a potential partner through its effect on trusting beliefs about their dedication.
With respect to Hypothesis 5 and 6, recommendation directly influences motivation for coproduction with the potential partner (c = 0.44; Equation [7]). The focal party had trusting belief in a recommended potential partner's ability (a = 0.80; Equation [12]) and dedication (a = 0.67; Equation [15]); trust in their ability therefore increased motivation to coproduce with them (b = 1.45; Equation [8]); while trust in their dedication did not increase motivation for coproduction (b = 0.37; Equation [9]). The indirect effect with a bias-corrected bootstrap 95% confidence interval for Hypothesis 5 (ab = 1.17) based on 10,000 bootstrap samples was entirely above zero (0.453–1.942). Thus, Hypothesis 5 is supported. However, the 10,000 bootstrap samples with a bias-corrected method for Hypothesis 6 showed the 95% confidence interval for the indirect effect (ab = 0.25) contains zero (−0.101 to 0.689). There is no sufficient evidence to support Hypothesis 6. In sum, recommendation increases motivation to engage in coproduction with the potential partner through a level of trusting belief in ability but not in dedication.
Discussion and Conclusion
This paper attempts to show the importance of ability and dedication in a coproducing context by examining how reputation works in the search for a coproducing partner. The findings are consistent with the hypotheses that certification, word-of-mouth, and recommendation are, in turn, positively related to trust in a potential partner's ability. The trusting belief in ability mediates all the relationship between three reputation sources and motivation for coproduction. However, only the certification links to motivation to coproduce with the potential partner through trust in dedication. These empirical results suggest different sources of reputation influence the focal party's selection choice variously.
First, it is worth noting that focal parties address one's attributes separately, and the inconsistent result of trusting belief in ability dedication from word-of-mouth is one example. In the word-of-mouth case, trusting belief in ability mediates the relationship between word-of-mouth and the motivation to coproduce. However, trusting belief in dedication was unable to claim a mediating effect between word-of-mouth and the motivation to coproduce. The segmentation between trust in the ability and dedication is reasonable since the focal party expects to obtain required resources through coproduction and to utilize these completely. Assessing the ability of the coproducer is equally important as evaluating the possibility to utilize this ability fully. Certain types of reputational information (e.g., word-of-mouth) may be unable to provide sufficient information for the focal party to assess the two objects simultaneously.
Second, a unitary acceptance of ability and a divergent acceptance of dedication may be attributed to the inconsistent results of recommendation. The significant mediating effect of trust in ability supports the fact that recommendation represents sufficient ability to meet an accepted requirement. However, the respondents showed that their trusting belief in one's dedication did not significantly mediate the relationship between recommendation and motivation to coproduce. The result may derive from different levels of acceptance in defining adequate dedication; that is, two parties may have different evaluations of one coproducer's dedication. Focal parties may have a trusting belief in the dedication of a recommended coproducer; however, focal parties turn to their assessment when deciding whether to coproduce with this recommended coproducer. Namely, from their social experience, focal parties realized the recommendation may be distorted by individual differences. Although recommenders believe they provide the recommendation that suits the focal party's need, the focal party may slightly adjust the evaluation, which influences the mediating effect of trust in dedication.
Implications
This study makes several key contributions and implications. First, we mainly distinguish dedication from ability. Previous studies focus on assessing the ability of the partner (Auh et al., Reference Auh, Bell, McLeod and Shih2007; Chen, Tsou, & Ching, Reference Chen, Tsou and Ching2011; Connelly et al., Reference Connelly, Certo, Ireland and Reutzel2011; Hitt et al., Reference Hitt, Dacin, Levitas, Arregle and Borza2000); however, dedication is a major factor worth noting in coproduction. The essence of coproduction is resource obtainment and performance ambiguity; the former implies the importance of ability, and the latter addresses the importance of dedication. Thus, a coproducer with both ability and dedication and one with ability but little dedication may be regarded as two types of potential partner. In addition to showing the ability, empirical practitioners may need to present their dedication to the work to gain more coproducing opportunities.
Second, we suggest that distinctions between reputation sources are salient. Previous studies include all reputational information under one category: reputation (Doney & Cannon, Reference Doney and Cannon1997; Fombrun, Reference Fombrun1996; Ganesan, Reference Ganesan1994; Wagner, Coley, & Lindemann, Reference Wagner, Coley and Lindemann2011); however, the focal party may obtain information from different channels and therefore respond differently. For example, certification from a professional institution may be regarded as factual, without personal distortion. Recommendations from a known friend may represent an assessment of a similar view worth considering. Moreover, first-hand and second-hand information have different meanings for the information receiver. For example, word-of-mouth may be regarded as second-hand information. The focal party may obtain word-of-mouth from a secondary source and is unable to trace the original information source. Regarding attributes that can converge to generally accepted criteria (e.g., ability), the focal party may be able to judge whether individual preferences distort word-of-mouth. However, for an attribute that is unable to quantify (e.g., dedication), it is hard for the focal party to assess the extent of distortion in word-of-mouth. The essence of reputation suggests the differences between reputation sources may induce different reactions; therefore, empirical practitioners may need to gather reputation from various channels, thereby selecting partners in a more comprehensive way. Empirical practitioners may also try to establish their reputation in different channels – and may, therefore, receive more job opportunities.
Third, we extend the application of signaling theory by linking it to the trust-related literature. Spence (Reference Spence1973) suggested that actors will respond to the signaler (e.g., hire them) after receiving the signal. This logic is originally derived from an economic perspective; however, little is addressed regarding individual differences. Combining signaling theory with trust-related concepts allows researchers to explain why particular signals may be more appealing to the receiver, thereby inducing future actions.
Fourth, we differentiate this study by treating the selection process in detail. Li, Eden, Hitt, and Ireland (Reference Li, Eden, Hitt and Ireland2008) assume that partner selection is derived from exchange attributes; however, the selecting process is more complicated in reality. Selection is regarded as a way to protect the focal party from being vulnerable in the relationship; therefore, the focal party may start by gathering information and forming beliefs regarding their intention to work with a partner. Theoretical developers may benefit from finding out how decisions change as priorities vary in different selection processes; empirical practitioners may attract focal parties by meeting their expectations.
Last but not least, empirical practitioners who benefit from this study are not limited to this research context; any coproducing context is applicable. The context may be the selection of leading roles in a stage play, the member of an orchestra, or any context involving performance ambiguity. Empirical practitioners may find direction regarding what to consider and how to reduce problems ex-ante from this study.
Limitations and directions for future research
Notwithstanding the aforementioned implications, our study has limitations. First, we were unable to examine the strength of, or compare, reputation sources in the current study; this remains for future studies. Second, the setting of our research context may provide a partial explanation for the small sample size; it focused on dental prosthesis fabrication and excluded tooth implants. These two treatments serve as different research contexts, and we were unable to ascertain those clinics specializing in tooth implants from the brochure before sending the questionnaires. Accordingly, specialized clinics may be unable to answer our questionnaire. Moreover, the sensitivity of potential respondents' concerns on revealing the quantity of dental prosthesis fabrication may have created an unwillingness to reply. Although not a perfectly large sample, this study reveals the practice of coproduction. Third, we sent questionnaires only to the focal party – the dentists. This study thus simply represents the focal party's perspective. Significant work remains to be done from the perspectives of coproducers. By comprehending the perspectives of both sides, we may have a better understanding of coproduction.
Future research may further examine the role of third parties, which deserves more attention. The third party in this study is a nonstakeholder in the coproducing relationship, which ensures information authenticity; therefore, it may be interesting to see future researchers examine the context in which the third party is a stakeholder. Moreover, future research may further examine the relationship between the motivation to coproduce with the potential partner and the actual coproducing behavior. In the relationship, it may be an interesting topic to include factors such as affective and calculative commitment (Nyhan, Reference Nyhan1999). Finally, future research may examine this framework in a context without bilateral dependence and specificity or a craft-based research context, which may be another interesting issue worthy of attention.
