Discussion of Results

---

Perceived Cost-Benefits of EDI

Perceived direct benefit and perceived cost are found to be important factors in determining EDI adoption. These findings concur with those found in previous studies in organizational innovation and EDI adoption (e.g., Cragg & King, 1993; Iacovou et al., 1995; O'Callaghan et al., 1992; Premkumar et al., 1994). In fact, perceived benefits and perceived cost can be considered as two different aspects - as motivators and barriers - to assess in the adoption decision.

Perceived benefits can act as motivators to encourage the adoption of an innovation because direct benefits are more visible and easier to measure (e.g., reducing paper use, improving turnaround time of transmissions). As most indirect benefits (e.g., improving the organization's image and competitive advantage) are difficult to quantify and it takes a longer period to see their impacts, their influence is not significant in EDI adoption. Another plausible explanation is that since the decision-making process in SME is always short-term and intuitive (Fink, 1998), the organization may thus pay more attention to the more obvious and explicit benefits rather than to the implicit and long-term benefits.

In addition, the benefits from EDI depend on the level of integration. Some EDI benefits, especially indirect benefits, can only be enjoyed through a full integration into an existing system (Hoogeweegen & Wagenarr, 1996; Raymond & Bergeron, 1996). However, many SME have only low levels of integration or even go without any integration planning. The SME that adopt EDI generally implement a less complicated EDI system, which simply allows them to receive EDI documents, print out hard copies of the documents and processes them manually (Tuunainen, 1998). In this scenario, only limited benefits such as faster transmission can be obtained by the SME. The obtained benefits are thus mainly direct benefits. Consequently, indirect benefits may not be considered in making the adoption decision for EDI.

Perceived cost was found to be another important factor in the EDI adoption decision. This finding not only provides empirical support of the previous literature, but also reinforces the argument about the influence of cost on EDI adoption. In traditional adoption and diffusion studies, it is suggested that the less expensive an innovation is perceived to be, the more likely it will be adopted (Rogers, 1983). In many studies, cost is a major barrier for adopting EDI (Arunachalam, 1995; Cox & Ghoneim, 1996).

SME have few resources compared with large organizations. Therefore, these organizations pay more attention to financial resources. When considering adopting new innovations, cost is a critical factor in the adoption decision (Iacovou et al., 1995; Saunders & Clark, 1992). A wide range of cost is required for EDI adoption. Most SME have simple computer infrastructures and have inadequate hardware and software (Iacovou et al., 1995). Also, many SME may not have IT budgets or plans and, therefore, the setting-up cost for adopting EDI is considered an unexpected expenditure. To buy a new computer or software for EDI purposes will become a significant barrier. In addition, work procedures may also have to be changed after adopting EDI. Training for staff is essential to catch up with the new technology. These costs create barriers for SME to invest in and adopt EDI (Kalakota & Whinston, 1997).

Organizational Context

IT knowledge is found to be a significant factor in distinguishing between early and late EDI adopters. An organization with greater IT knowledge tends to adopt EDI early. This finding parallels Attewell's (1992) theory that many organizations will postpone EDI adoption until developing the necessary skills and acquiring the required IT knowledge. This implies that less IT knowledge is a barrier to adopting innovations. Many SME, in general, lack IT knowledge and technical skills. SME tend to delay EDI adoption because they have insufficient knowledge to implement and operate EDI successfully.

Another organizational factor, the attitude of the top management towards EDI adoption, was not found to be significant. This suggests that the attitude of top managers does not have a prominent effect on EDI adoption. Two possible explanations are suggested. First, a major difference between this and earlier studies is that adoption of innovation in prior studies was made under voluntary situations. The adoption decision to implement EDI in this study is mandatory. Organizations are required to adopt EDI in order to submit trade declarations to the government. In other words, the government-mandated electronic submission is the only means to submit trade declaration applications. Top management did not have a choice regarding whether or not to adopt. Under this mandatory situation, top management may face great pressures from the external environment.

Second, the type of innovation helps to explain the insignificance of top management attitude. Unlike traditional innovations such as intra-organizational and IT-specific systems (e.g., CASE tools and Open Systems), the adoption decision to use an IOS may involve different partners like the government or various trading partners and, therefore, is not primarily based on only the organization itself. Influences from different partners have been found to be an important factor in adopting EDI (Bouchard, 1993).

External Influences

In the EDI literature, many researchers have suggested that adopting EDI may be mainly due to imposition from a trading partner and other external influences (Premkumar et al., 1994; Raymond & Bergeron, 1996). The findings of this study indicate that external influence is important in determining adoption of EDI in the SME. The government incentives and enforcement and trading partners' influences were found to be significant in EDI adoption.

Government incentives and enforcement were found to be significant in influencing EDI adoption. The greater government incentives and enforcement as perceived by an organization, the higher the likelihood of the organization to adopt EDI early. As noted above, the adoption of the EDI observed in this study was mandated by the government. The findings show that the legal concerns are among the important factors that influence adoption of EDI. In their study, Neo et al. (1994) examined TradeNet, which was an EDI system that was also mandated by a government, but they did not find significance of government influence in adopting the system. They believe, however, that government influence is vital for ensuring adoption of a nationwide innovation. The significant result in this study also suggests the importance of government incentives and enforcement.

Contrary to one of our hypotheses, trading partners' influence was found to have a negative influence on EDI adoption. This suggests that an organization that is not influenced by trading partners is more likely to be an adopter and vice versa, which differs from existing results and common sense. One possible explanation is that in this study the adopters possessed the necessary resources and perceived high levels of benefits, so they ignored the influence of their trading partners in adopting EDI. Iacovou et al. (1995) suggested that EDI initiators were those that "had recognized the need for EDI, possessed sufficient financial resources, had high levels of IT sophistication and had not been pressured by external factors into adopting EDI" (p. 476). The early EDI adopters in our study could be regarded as EDI initiators in Iacovou et al.'s terms and could be assumed to be more active in adopting EDI because they had adequate resources and knowledge. When making the adoption decision, little consideration was given to trading partners because of the existing initiatives and sufficient resources to adopt EDI.

Alternatively, when the early adopters decided to adopt the technology, there were few within the industrial community that were using it. As a result, the adoption decision of these early adopters was probably made without regard to others but with regard to costs and benefits instead. Also, the very few players in the industry meant that these firms had neither to experience any pressure to meet industry standards (Banerjee & Golhar, 1994) nor to conform to peer company pressure (Premkumar & Ramamurthy, 1995). Those late adopters, however, had pressure to see how their counterparts had done.

The third plausible explanation is that since few organizations had adopted the system, the chance of having recommendations of trading partners was relatively small. When organizations adopt an innovation successfully and gain from using the innovation, this acts as a model for other organizations to imitate. Late adopters might therefore get more information and thus be influenced by other firms in the industry as the trend of using EDI develops.

A fourth explanation suggests that since the use of EDI among organizations was not prevalent, some SME might adopt a "wait-and-see" attitude in order to have a clearer picture of what the actual benefits of the adoption of EDI were. Even when they had the necessary information, they might still prefer to defer the adoption as long as possible because, on the one hand, it might remain unclear to them that their firm would receive any of the benefits (Clemons & Row, 1993), and, on the other hand, they might believe that the technology would result in an overall weaker bargaining position with their business partners and therefore lower profits (Riggins & Mukhopadhyay, 1994).

---