Analysis and Discussion

In 1996 when the EDI Action Plan was launched, large companies had already adopted EDI (Andersen et al., 2000; H rl ck, 1996), but SMEs hesitated. One reason for the low level of adoption of EDI among SMEs could be that they lacked sufficient knowledge of the innovation. Another explanation could be that the SMEs did not recognize the business potential of EDI. A third and perhaps more plausible explanation might be rooted in issues related to critical mass, pressure, and power (Henriksen, 2002; Markus, 1983). However, it is a fact that the EDI traffic—viewed as number of messages and number of bytes exchanged—increased from 1996 and in the years ahead (Andersen et al., 2000), but it did not create a large EDI landslide. A landslide could have been expected when considering that the industry and trade associations initiated several EDI projects as a result of their involvement in formulating the EDI Action Plan. The midway evaluation in 1998 of the EDI Action Plan prepared by the Danish EDI Council documented that several efforts had been made to meet the demands, but that the goals so far had not been reached.

In the EDI Action Plan there were no economic or normative interventions (Henriksen, 2002). Contrary to the TradeNet in Hong Kong, the adoption of EDI was not enforced. One natural consequence of the Danish EDI Action Plan could have been to force private businesses to report their figures to, e.g., the Inland Revenue Service via EDI messages. This would have been highly relevant since the EDI Action Plan aimed at making the public sector a locomotive for EDI adoption. Since all Danish businesses communicate with the Inland Revenue Service, it would have created a broader scope for EDI communication. The Inland Revenue Service was among the few public sector agencies that actually utilized the opportunities created by EDI usage (Andersen et al., 2000). However, no enforcement was used to involve private businesses in the exchange of EDI messages.

Another approach to the creation of a large-scale diffusion of EDI could have been a strong encouragement of the public sector to perform its procurement electronically. Part of the reason why mandatory e-procurement was not implemented was the lack of technological readiness at both demand and supply side. The municipalities, for example, had in general a low level of adoption and diffusion of EDI. But perhaps the most likely explanation for the low diffusion rate of EDI is the fact that the public sector was not managed in a top-down style. About one-third of the employees in the Danish governmental sector are employed in central government and two-thirds at the local level in the 275 municipalities and 14 counties. Local government has a large degree of autonomy, and various management reforms in the 1980s and 1990s in the public sector have decentralized budgets and procurement to each public institution. Any ideas of making e-procurement mandatory in the public sector were therefore in direct conflict with the overall management agenda for the public sector. This is a paradox since the management reforms introduced in the 1990s had the overall aim to increase efficiency and effectiveness. However, the means to achieve the objective of decentralization and autonomy at the institutional level prevented a central, top-down-led strategy for e-procurement. Formulated more sharply: the overall management discourse prevented an effective introduction of an e-commerce agenda in the public sector (Andersen et al., 2003b).

Although financial resources were used as direct subsidies to support the diffusion of EDI in the Danish business community, the resources were somewhat limited and not allocated to individual businesses. The financial support was given to projects such as the TradeDocument Project (Henriksen & Andersen, 1999; Henriksen, 2002) initiated by the two major Danish trade and industry associations. Therefore, it can be argued that the economic incentive for adoption of EDI was absent in the EDI Action Plan with respect to individual businesses. The initiators of the EDI Action Plan relied more on a pedagogical approach (Eckhoff, 1983). Through information campaigns by the Danish EDI Council and the associations supporting the EDI Action Plan, the advantages of EDI were communicated to the potential adopters. Thus, this communication process became a central issue in the adoption and diffusion of EDI in the Danish business environment. Focusing on the communication aspect, one issue is how the information about the innovation is communicated.

Another issue is how the innovation is presented. The message can be communicated via mass-communication networks or through interpersonal relations (Rogers, 1995). In the EDI Action Plan, both types of communication channels were used. The EDI Council and the professional business associations communicated their message about EDI and its advantages at the same time as the government presented its information campaigns on the advantages of the innovation via publications, newsletters, and social arrangements. This second issue—how the innovation is presented—is however even more relevant in this context. Two different types of information are involved in the communication process: signaling and know-how (Attewell, 1992). According to Attewell, signaling refers to communication about the existence and potential gains of a new innovation, whereas know-how refers to knowledge transfer in relation to the innovation. "The technical know-how is relatively immobile, and often has to be recreated by user organizations" (Attewell, 1992, p. 7). This places a heavy burden on potential adopters, since different organizations have specific needs requiring different skills and development before adoption can take place.

The first policy statement on technology from 1995 suggested that EDI adoption could lead to "considerable rationalization gains and a closer interplay between organizations." In the EDI Action Plan, EDI was described as a tool that could give Denmark an "international lead and improve the efficiency of working procedures and development of new products and production processes." Efficiency was an attribute connected to the innovation, whereas the traditional innovation attributes (Rogers, 1995) played a secondary role. However, the statements formulated by the business community might be even more significant. They argued that EDI adoption would result in increased efficiency, cost reduction, and competitiveness leaving the impression that the investment would pay off. This suggests that the information communicated on EDI belonged to the category of signaling rather than to the category of know-how. This is problematic in a business environment like the Danish, where the majority of businesses are SMEs, which generally have limited resources available for further investigation of innovations being offered.

An issue that played a minor role in the policy statements was the cost of EDI. One of the seven initiatives in the EDI Action Plan explicitly mentioned development of EDI software as an objective, but the cost of hardware, integration of EDI software, and implementation of EDI in organizations was not mentioned in the statements. It is generally acknowledged that to gain full advantage of EDI, organizations must integrate their internal systems with the external IOS (Riggins & Mukhopadhyay, 1999). In a survey from 1996 on EDI adoption among members of the Confederation of Danish Industries, it was found that the two highest scoring items in terms of barriers related to EDI implementation were cost of EDI and the uncertainty of realizing actual advantages from EDI (H rl ck, 1996). The third highest scoring item in the survey was related to the lack of standardization of trade data standards. Apart from serving as a tool for creating awareness of EDI, the EDI Action Plan aimed at introducing the EDIFACT standard in the Danish business community and the public sector, but the cost of EDI was not a major issue in the EDI Action Plan.

As mentioned earlier, the aim of this chapter is to focus on the institutional initiatives deployed to promote the diffusion of EDI in the Danish business community. One outcome of this analysis is a mapping of the degree of institutional involvement in the formulation of the EDI Action Plan versus the rate of EDI adoption. Appendix A provides a list of institutional involvement at different governing levels. Based on insights gained from the research projects reported in Appendix B and the data summarized in Appendix A, Table 3 presents a qualitative mapping of the variables, institutional involvement, and EDI adoption in selected businesses and public sector agencies.

Table 3: Institutional Involvement and Adoption of EDI for Different Industry Segments

Institutional Involvement	Degree of EDI Adoption
	Low	High
Low	Municipalities	The financial sector The MedCom project The agricultural sector (DLF) Transportation and logistics
High	Business and trade associations	The Inland Revenue Service

The Inland Revenue Service was involved in formulating the EDI Action Plan and was eager to gain the benefits that most likely would result from an increased EDI usage (Andersen et al., 2000). Based on information from the Inland Revenue Service, its EDI usage is assessed to be high in both dimensions in Table 3. On the other hand, representatives from the municipalities did not play an active role in formulation of the EDI Action Plan, nor did they pursue EDI adoption and implementation at any significant level with respect to EDI usage in their daily transactions both in relation to the public and private sector.

The professional business and trade associations played an active role in formulating and operationalizing the EDI Action Plan (Henriksen, 2002; Henriksen & Andersen, 1999). However, they did not succeed in influencing adoption and diffusion among their members. Our assessment is therefore that the trade and business associations' institutional involvement was high, but the degree of adoption among their members was low.

A number of associations did not take an active part in formulation of the EDI Action Plan. These associations are clustered in the upper right corner of the two-by-two matrix in Table 3. Nonetheless, these associations represent some of the business sectors that have been most successful in utilizing EDI in Denmark (Andersen et al., 2000). This leaves the impression that critical mass and power, rather than institutional initiatives and institutional discourse, determined the rate of adoption in the Danish business community.

One of the fascinating issues concerning the Danish scene is the public sector's role as a player in the EDI Action Plan. The Danish EDI Action Plan was not solely a governmental plan. The business associations were deeply involved in the formulation of the EDI Action Plan, and a great responsibility was laid on the business associations in order to fulfill the plan. However, the institutions in the public sector acted as major players in fulfillment of the plan. That is especially the case with respect to single units such as the Inland Revenue Service. This important insight into the role of government might however not include the traditional top-down command and the concept of "government" versus "market." The successful EDI initiatives originate at lower levels of government and in the market where they have managed to spiral and gain success despite a lack of policy saliency. One explanation for this outcome of the EDI Action Plan could be the consensus in the formulation of the action plan between the public sector and the private sector. However, the consensus among the business associations and the governmental players did not eliminate the obstacles to the fulfillment of the EDI Action Plan. The outcome of the plan raises the question whether the resources should have been prioritized differently. On the one hand, major Danish companies such as Maersk, Danfoss, Grundfos, and FDB have not benefited from the EDI Agenda. They have all managed well without the EDI Action Plan and have been able to influence their suppliers and shipping agents to adopt EDI—often in a proprietary format—regardless of the political agenda's content. On the other hand, the EDI Agenda has helped to stimulate a range of EDI projects that otherwise might not have been implemented. These projects were especially targeted at the majority of Danish businesses—SMEs, which through this exposure were provided with the necessary information to consider the relevance of EDI in their businesses. If we subscribe to this last point of view, then it might only be a matter of time before a B2B e-commerce landslide takes place.