اختصاصی بودن دارایی و هراس از بهره برداری

If asset specificity renders the investing party dependent ex post, why would the ex ante willingness to make relationship-specific investments vary? We show how specific investments generate both positive and counter-negative cooperative incentives. We also observe the influences of trust and time horizon on these incentives, which are aggregated to derive the specific investments effect (SIE). Our result suggests that while the fear of exploitation increases proportionally to the magnitude of specific investments and the attendant quasi-rents, it grows exponentially with the deterioration of inter-personal (trust) and/or inter-temporal (time horizon) contexts.

A central tenet of transaction cost analysis is that transaction-specific investments promote integration since the probability of opportunism to appropriate quasi-rents increases with investments highly specialized to an exchange (Williamson, 1985, Joskow, 1988, Anderson and Weitz, 1992 and Gulati et al., 1994). As a classic example, the Chrysler Corporation in 1986 instructed its parts suppliers to reduce prices by 2.5 percent after specialized investments had been made. Chrysler strictly enforced this policy and terminated its relationships with suppliers who did not comply (The Next Act, 1986). An interesting question arises: if asset specificity renders the investing party dependent and hence vulnerable to exploitation ex post, why would one's willingness to commit vary ex ante? Research shows that firms in Japanese automaking, Italian knitwear, and New York apparel industries demonstrate greater willingness than others to make specific investments (e.g., Sako and Helper, 1998 and Uzzi, 1997). For example, Japanese automobile suppliers develop more unique parts for their customers and make greater investments in specialized assets. In contrast, components that require considerable engineering development efforts tend to be produced in-house in the U.S. (Smitka, 1991, Dyer and Ouchi, 1993 and Monteverde and Teece, 1982). In a large-sample, cross-national empirical test, Dyer and Chu found that U.S. supplier-automaker relationships were characterized by lower asset specificity compared to Japanese relationships. The observed variation of institutional arrangement for transactions with specific investments suggests that the connection between asset specificity and governance structure may be more complex than has been explained by transaction cost economics (Robins, 1987, Chiles and McMakin, 1996 and Nooteboom et al., 1997). This paper offers an analytical approach that systematically explores factors that explain the variance of ex ante willingness to make specific investments. We argue that such variance depends critically on the extent to which specific investments create dependence upon making specific investments. Understanding the mechanisms of how specific investments create dependence is important because dependence gives rise to the fear of exploitation and diminishes the willingness to invest (Williamson, 1985, Anderson and Weitz, 1992, Morgan and Hunt, 1994 and Gulati et al., 1994). Our research suggests that asset specificity simultaneously generates two distinct incentives that create dependence by inducing the investing party to cooperate ex post: one positive, which promotes cooperation; the other counter-negative, which deters defection. The positive incentive exists because cooperation is a pre-requisite to reap the potential quasi-rent and retain the resources ( Williamson, 1985, Parkhe, 1993 and Anderson and Weitz, 1992). The counter-negative incentive exists because one would have to forego the quasi-rent and lose the committed resources when defecting. Thus, asset specificity has the effect of eschewing opportunism because opportunistic behavior risks the dissolution of a relationship, which runs counter to self-interest ( Williamson, 1985, Anderson and Weitz, 1992 and Morgan and Hunt, 1994). Moreover, relationship-specific investments cannot be made in a vacuum but must necessarily be embedded in the context of a relationship that characterized jointly along the inter-personal and the inter-temporal dimension. The inter-personal dimension refers to the strength of social ties that generate trust while the inter-temporal dimension refers to the expected time horizon of future encounters ( Granovetter, 1985, Uzzi, 1997, Sako and Helper, 1998, Heide and Miner, 1992, Axelrod, 1984 and Artz and Brush, 2000). That greater trust and/or longer time horizon is more amenable to specific investments is intuitively appealing and supported by research (e.g., Uzzi, 1997, Morgan and Hunt, 1994, Chiles and McMakin, 1996, Nooteboom et al., 1997 and Sako and Helper, 1998). There is an extensive body of literature addressing the role of the relational context in encouraging specific investments across disciplines, including marketing, strategic management, law, economics, and sociology (Macneil, 1980, Dore, 1983, Goldberg and Erickson, 1987, Noordeweir et al., 1990, Crocker and Masten, 1991, Zaheer and Venkatraman, 1995, Artz and Brush, 2000 and Dyer and Chu, 2003). Zaheer and Venkatraman maintain that the process by which the relationship is managed, in addition to the actual type of contract, is critical to determine the effectiveness of a relational contract. Promoting relational norms that facilitate information sharing, joint planning, long-term orientation, transparency, and persuasive rather than demanding negotiation strategies are considered important process elements (see Artz and Brush for a review). In this paper, we address the effect of relational context on specific investment by considering the joint impact of trust and time horizon, which captures the holistic notion of an ongoing social relationship. As such, we explore how the fear of exploitation that stems from specific investments is shaped by the joint effect of the inter-personal and the inter-temporal aspect of a relationship. Drawing on transaction cost theory within a relational context, this paper derives an indicator of specific investments effect (SIE) from a simple decision-making model. SIE captures the combined effects of the positive and counter-negative incentives as well as their interactions with the underlying relational context (inter-personal and inter-temporal). SIE measures the extent to which asset specificity compels the investing party to cooperate ex post from fear of losing the specific interments and not getting the attendant quasi-rent once resources are committed. The need to cooperate ex post thus gives rise to dependence and entails the appropriation concern of the investing party. Our research suggests that the fear of exploitation begins to grow when specific investments are made in a context deviating from full trust and assured future. It escalates at an increasing rate if a relationship deteriorates either inter-personally or inter-temporally. SIE approaches infinity when there is no trust and/or a one-off exchange, implying that no credible commitment should ever be made. Our result suggests that while the fear of exploitation increases proportionally to the magnitude of specific investments and the attendant quasi-rents, it grows exponentially with the deterioration of the inter-personal (trust) and/or the inter-temporal (time horizon) context. Because the same level of asset specificity could generate dependence to varying degrees depending upon the relational context in which it is embedded, the fear of exploitation may be higher (lower) for a lower (higher) level of specific investment. In addition, our research suggests that trust and time horizon reinforce each other in alleviating the fear of exploitation. In other words, the exploitation concern would be less sensitive to the deterioration of trust in a relationship characterized by a longer time horizon. As well, the sensitivity of such concern would be less in a more trusting environment when the time horizon shortens. This implies that engaging in “repeated transactions” with “trusted partners” has a multiplicative (rather than additive) impact on one's willingness to make specific investments. Lastly, a corollary to the question about one's willingness to invest in specific assets is: “What is the threshold level for a specialized investment that is required for cooperation to be possible?” We derive such a threshold level in this paper. In this paper, asset specificity is treated as a factor that modifies the payoff structure of a repeated exchange (see also, Gulati et al., 1994). We demonstrate how investments unique to a relationship generate both positive and counter-negative incentives that give rise to dependence. We also observe the influences of trust and time horizon on these incentives, which are aggregated to derive the specific investment effect. SIE captures the overall effect of specialized investments on inducing the investing party to cooperate ex post and thereby creating dependence. While SIE looks at the effect of specific investments from the focal party's perspective, the balance of specific investments made by both parties is considered. The implication of our research on the choice of governance structure is discussed. The model on which this research is based and the results derived therefrom are presented in the section below, followed by discussion and concluding remarks.

Firms are increasingly focusing on core competencies while outsourcing non-core activities. Creating strategic advantages through close working relationships frequently requires sharing sensitive cost and process information and committing unique resources (Uzzi, 1997, Sako and Helper, 1998 and Dyer, 1996). This however may weaken bargaining power and increase the exposure to opportunism. Thus, along with the potential of superior performance from inter-firm asset specialization, close collaborations may bear significant risks. In an IT industry outsourcing report (KPMG, 1997), the most frequently encountered problems are related to post-contractual dependency. Forty-two percent of respondents complained of “over-dependence on suppliers,” while 39 percent felt they were “locked-in” to their suppliers. This paper asks why some firms are more willing than others to make costly-to-reverse investments. By exploring the extent to which asset specificity induces cooperation ex post, this paper suggests that specific investments and the relational context (inter-personal and inter-temporal) in which they are embedded jointly give rise to the exploitation concern. Our analysis delineates the mechanisms behind the fear of exploitation. Specifically, the immobility of dedicated assets promotes the “inclination towards cooperation” and diminishes the “attractiveness of defection” at the same time. While the intensities of both are positively related to relationship-specific investments and the attendant quasi-rents, the latter is subject to the additional influence of trust and time horizon. The aggregation of both yields the specific investment effect, which measures the overall impact of relationship-specific investments on inducing cooperation ex post. A higher value of SIE implies a higher level of dependence and hence a greater fear of being exploited. SIE equals the sum of the committed amount and the attendant quasi-rent with full trust and assured future, in which case there is no fear of exploitation. SIE attains infinity when there is no trust and/or commitment is made in a one-off encounter, in which case the fear of exploitation is infinite. In between, one would be increasingly more reluctant to make unique investments if a relationship deteriorates inter-personally and/or inter-temporally. As such, our research not only concurs with but goes beyond the proposition that the strength of social ties is a major determinant of idiosyncratic investments (e.g., Moran and Hunt). We argue that efforts made in cultivating trust are important because trust deterioration hinders specific investments at an increasing rate, and the prolongation of the time horizon has the similar effect. One important result of our analysis is the asymmetric role that relationship-specific investments and the relational context (both inter-personal and inter-temporal) play in shaping the fear of exploitation. While the magnitude of specific investments and the attendant quasi-rents have a linear, positive relationship to the fear of exploitation, the deterioration of trust and time horizon exhibit a negative, exponential relationship to such fear. Our analysis suggests that we need to go beyond specific investments per se to grasp a better understanding of the exploitation concern; a higher level of specific investment could imply a lower level of exploitation concern in a more trustworthy and/or prolonged relationship. Moreover, while an improvement in either inter-personal or inter-temporal context alleviates the fear of exploitation, their combined influences are multiplicative rather than additive. As specific investments by both parties determine dependence in a relationship, the balance of both party's specific investments is addressed. We found that one party's absolute dependence is increasingly more difficult to be offset by the other party's specific investment. The relative dependence however is restored if both parties make the same amount of investment. We also derive the critical threshold of specific investment beyond which cooperation would be enforced is derived. Our research carries implications for the choice of the appropriate governance structure for firms making specific investments. While hierarchy is considered as a more efficient way to organize when transaction-specific investments are great, our research suggests that a hybrid structure such as strategic alliances may be a good choice. At the heart of the issue is whether firms are capable of managing both the inter-personal as well as the inter-temporal aspects of a relationship successfully. Our research provides ample opportunities for empirical studies. One could test the fit of the model by operationalizing the fear of exploitation as the dependent variable and the magnitude of specific investment, quasi-rent, trust and time horizon as the independent variables. One could also test the functional forms of the coefficients as some may be linear while others are not. Understanding why one party is more willing to invest in assets that are unique to a relationship than the other party is a challenging question for researchers. More theoretical and empirical studies are needed to explore this issue further and to arrive at a satisfactory answer.