EDI Adoption and Usage
in Australian Government Agencies
1989-94

Roger Clarke

Principal, Xamax Consultancy Pty Ltd, Canberra

Visiting Fellow, Department of Computer Science, Australian National University

Version of 7 May 1994

© Xamax Consultancy Pty Ltd, 1989-94

This paper was presented at the 5th World Congress of EDI Users, Brighton, United Kingdom, 14-17 June 1994

This document is at http://www.anu.edu.au/people/Roger.Clarke/EC/Brighton94.html


Abstract

This paper reports the results of a six-year project which studied EDI in the Australian federal government. Purchasing EDI has progressed less quickly in government agencies than had been anticipated during the bullish period of the late 1980s, and the reasons for the delays are analysed. Beyond purchasing, there are many aspects of public administration to which government agencies can apply EDI. In this area, experiences have been rather more encouraging, with Australian agencies being world-leaders in the application of EDI to the administration of taxation and customs.

The paper contains information gleaned from a succession of surveys regarding various aspects of EDI's implementation, use and non-use. Policy measures applied during the period are described, and policy recommendations for the coming years provided. It is concluded that a major Commonwealth Government program is critical to ensuring timely exploitation of EDI and other aspects of electronic commerce and IT-supported public administration.


Contents

1. Introduction

2. The Context

3. The Research Method

4. The Survey Results

5. Policy Measures - 1988-1994

6. Policy Recommendations - 1994-1996

7. Conclusions

References


1. INTRODUCTION

In 1988, a Research Programme in 'Supra-Organisational' Systems was formed at the Australian National University. Its purpose is to study applications of information technology which transcend the boundaries of individual organisations. Research has been undertaken into many different aspects of electronic commerce, ranging from supplier directories to EFT/POS, into a variety of relevant technologies such as corporate gateways, smart-cards and inter-networking, and into particular issues such as data security.

A great deal of the Programme's resources have been committed to EDI, because this has been an area of intense interest to business and government during recent years. Particular projects have focussed on appropriate architectures to support purchasing, EDI's impact on industry structure, EDI's inter-relationship with other components of electronic commerce, and the application of EDI in international trade and transportation, and in various aspects of public administration.

During the period from 1989 until 1994, a longitudinal survey has been conducted into Australian Government Practices and Intentions regarding EDI (GPIE). It involved four questionnaire-based surveys, supplemented by structured interviews with a sub-set of survey respondents, and case studies. A total of seventeen research reports have resulted, totalling in excess of 1,000 pages. The purpose of this paper is to express the project's key outcomes in readily digestible form.

Finance for the initial survey in 1989 was provided by the Australian National University's Bruce Fund. The federal government committee which plays a co-ordinative role in relation to IT, the Information Exchange Steering Committee (IESC) provided very important moral, and some financial, support. The arm of the federal government responsible for purchasing policy, Purchasing Australia, within the Department of Administrative Services, part-funded the second and subsequent questionnaire-based surveys, and fully funded several specific sub-projects. Three successive reports arising from the project have been published, providing rolling self-funding.

This paper firstly provides an outline of key aspects of the context within which the project was undertaken, followed by a brief description of the study method. The body of the paper reports overall progress, developments in purchasing and in public administration more generally, and other outcomes of particular significance. Policy measures undertaken during the study period are outlined, and the paper concludes with observations concerning policy measures needed now, to overcome impediments and stimulate growth.


2. THE CONTEXT

Australia is a vast nation, similar in size to mainland United States or Europe-without-Russia. There are long, empty distances between the pockets of 4.5 million people (in Newcastle-Sydney-Wollongong), 3 million (in Melbourne) and 1.5 million (in each of Brisbane, Adelaide and Perth). It has a population of over 17 million, and its economy is somewhat larger than that of the Netherlands, and somewhat smaller than that of Texas.

Australia enjoys one of the most sophisticated, reliable and cost-effective telecommunications infrastructures in the world. Six major VANs provide relevant network services, including EDI, email and bulletin board systems. Four of these are subsidiaries of, or joint ventures with, multi-national companies (AT&T, GEIS, IBM and SITA), and two are Australian (Telecom Plus and NEIS).

The first wave of EDI which occurred in the United States and the United Kingdom passed Australia by. Since the mid-1980s, however, it has been increasingly recognised in Australia that EDI has great potential to improve the effectiveness and efficiency of business processes. A specialist industry association, the EDI Council of Australia (EDICA) has been active since 1988. It represents the interests of EDI users and potential users generally, stimulates and administers working groups in particular industries (currently over 20 of them), and provides documentation and training services.

A number of purchasing-related schemes are in place, for example in the auotomotive industry, several segments of the retail sector, electrical supplies, the steel manufacturer BHP, several mining companies and electricity authorities, and pilots are running in others (EDI Research Australia 1989-92, Australasian EDI Report 1993-, EDICAST 1992-, Clarke 1993, Swatman 1993, NCG 1994 pp.15-20).

There are three tiers of government in Australia. The national level is referred to as the Commonwealth or Federal Government, and has specific heads of power defined in the Constitution of 1901. The Governments of the six States and two Territories are successors to the British colonies of the nineteenth century, and retain substantial residual powers which were not ceded to the Commonwealth. The third layer, local government, comprises hundreds of city, town and shire councils. This project has been concerned exclusively with agencies of the Commonwealth Government.

The agencies of the Commonwealth Government are highly heterogeneous, varying by function, orientation, size, and degree of dispersion. For the purposes of the project, an 'agency' was defined to be an identifiable organisational unit of the administrative arm of the Australian Commonwealth Government, including all sizable agencies, plus all smaller agencies whose practices and intentions in relation to EDI were judged by the research team to be of potential interest.

EDI is most commonly thought of in relation to the acquisition of goods and services, arrangements for transport between the supplier and the point of use, and settlement of the debt arising from the purchase. Government purchasing practices differ significantly from approaches in the private sector, being characterised by regulatory measures, to ensure probity and minimise corruption, and centralisation for reasons of both control and bulk discounts. In recent years, the actual issuance of purchase orders has been largely devolved to the peripheries of organisations. Purchasing officers are subject to many rules, but support services are available such as access to period contracts negotiated by specialists.

A number of questions were asked in the 1993 GPIE III survey concerning the profile of purchasing within agencies. Its strategic importance is highly variable: in a little over 50% of cases it was fairly low or very low, but 25% nominated fairly high and 10% very high. Of the large majority of agencies which had a business plan, only 25% mention purchasing and only 18% give purchasing a high priority. Surprisingly, only 76% of agencies admitted to having an IT Strategic Plan, but almost half of these mention 'electronic commerce', and one-third give it high priority (1993, p.21).

The total purchasing power of respondent agencies was over $4 billion in 1992/93. This was heavily concentrated, with QANTAS representing one-quarter, and the next 4 largest a further one-third; on the other hand, there were some significant non-respondents, including a large proportion of the defence community. The majority of respondent agencies (43 of 73) had purchasing power of between $1 million and $50 million p.a. Respondents issue a total of about 1 million purchase orders each year, and 13 respondents nominated that they issue over 10,000 (1993, p.21-24).

Nearly 80% of agencies which have computer support for purchasing use standard or customised packages, with the 54 respondents nominating 31 different packages. Only 15% considered their systems to be EDI-capable. One-quarter of the respondents were aiming to replace their software within 2 years (1993, pp.25-26).


3. THE RESEARCH METHOD

The study of relatively new and dynamic technologies being applied in different ways by different organisations is not an easy undertaking. EDI has sufficient promise in terms of improvements to the effectiveness and efficiency of whole sectors of the economy that the difficulties needed to be confronted.

To cope with the complexity, a combination of research techniques was used. Mailed questionnaires have the advantage of enabling data to be collected from a relatively large number of organisations, although at a rather superficial level of detail. Additional resources were invested in structured interviews with a carefully selected sub-set of agencies, in order to provide greater depth of understanding of EDI's role and impacts in specific contexts. A small number of agencies were the subject of case studies, enabling examination in greater depth and over a longer period, and hence the development of a fuller understanding of the dynamics of EDI's adoption.

The survey population was defined as those agencies listed in the then-current Commonwealth Government Directory, which is believed to cover every existing agency. Practical difficulties arise, because committees and petty commissions (QANGOs) are listed alongside massive bureaucracies, and cohesive Bureaux adjacent to Departments comprising a dozen or more largely autonomous business units. The research team accordingly exercised judgement, by excluding many minor organisations and selecting which conglomerations were to be treated as multiple agencies and which as single agencies.

A complementary Survey of the EDI-Capability of Suppliers to Australian Government Agencies (SECSAGA) was undertaken, comprising a questionnaire and structured interviews. The population was defined to be the top 250 suppliers by value of contracts published in the Commonwealth Government Gazette, in 1992/93.

The response rates, shown in Exhibit 1, were very high, especially given the questionnaires' length (variously between 10 and 16 pages) and complexity. There were substantial response rates by agencies of all sizes and types, with the only classification identified as clearly under-represented being the defence community.

Exhibit 1: Response Rates to the Questionnaire-Based Surveys

		Year		Population	Respondents	Response Rate

	Government Agencies	
		1989/90 [GPIEEE]	158	74	47
		1991/92 [GPIE II]	167	94	56
		1993/94 [GPIE III]	151	93	62

	Suppliers to Government
		1993/94 [SECSAGA]	243	110	45

The quality of responses, as measured by the level of incompleteness and internal inconsistencies, was mostly very good. In each survey, a small number questions presented respondents with unanticipated difficulties, and the results were treated with appropriate caution, or discarded.

In early 1992, structured interviews were conducted with twenty agencies which had been respondents to the GPIE II survey. This provided valuable depth of understanding to complement the breadth achieved using the questionnaires. The GPIE III survey accordingly incorporated structured interviews with six agencies, and, in a supplementary study to SECSAGA, the SISGA sub-project involved interviews with five suppliers.

The agencies selected to be the subject of case studies were in all instances the instigators of major EDI schemes. They were the Australian Taxation Office (Ryan 1991), the Australian Customs Service in respect of three separate schemes (Goleby 1992, Cameron 1993, Jeffery 1993), and a purchasing scheme within the Royal Australian Air Force (Jeffery 1994). This paper draws on all of the questionnaire- and interview-based surveys and case studies, to compose a composite picture of the state and progress of EDI in Commonwealth agencies.

A fuller discussion of the research methods used is in Clarke (1994c).


4. THE SURVEY RESULTS
4.1 Overall Progress in the Implementation of EDI

The 1989/90 GPIEEE survey established that only 4 of the 74 respondent agencies were users of EDI systems in the conventional sense of the term. Those agencies anticipated generally higher volumes during the following year, and higher or much higher volumes in the 2-5 year time-scale. Few respondents had any doubts that EDI usage generally would grow dramatically in the following few years, and nominated 16 new applications they intended within 1 year, and a further 13 within 2 years (1990, pp.37-49).

During the following two years, there was indeed considerable progress. The 1991/92 GPIE II survey concluded that "There has been considerable growth in EDI use by Commonwealth agencies during the last two years. This has been in a variety of different classes of EDI ... Unsurprisingly, however, the growth has not been as dramatic as responses to the previous survey suggested it might be" (1992, p.79). "Of the 94 respondent agencies, 28 stated that they were currently operating an EDI system", but this included instances of electronic file transfer and even of on-line systems and email. After careful examination, it was apparent that 15 of the agencies were using EDI, and the Tax Office was running a proprietary EDI scheme (pp.32-35).

In response to the 1993/94 GPIE III survey, 32 of the 95 respondent agencies declared their participation in EDI schemes. After these were re-coded, they disclosed 17 EDI-using agencies, and 36 instances in which it was used. There are many instances of user-agencies being involved in more than one scheme, and two agencies (Customs and QANTAS) were each involved in four (1993, p.30-34).

The concrete outcomes are that the number of agencies actively using EDI grew from 4 in 1989 to 15 in 1991 to 17 in 1993; and the total number of uses from 6 to 21 to 36. These figures must be treated with extreme caution, however, as the set of respondents differed from survey to survey. During the period 1991-1993, there was disappointingly little take-up by new agencies. Agencies generally perceived their plans to be on-time, although a few felt they were behind, and in two cases well-behind (both in relation to purchasing-related applications) (p.40).

Fully 40% of respondent agencies intended spending precisely nothing on EDI during the next few years. A total of 28 agencies nominated their intention to implement EDI schemes, 19 in the purchasing area, 6 in inter-agency schemes, and 11 in various other areas. Almost 80% of these, however, were medium-term plans (1-5 years away) rather than short-term commitments, and some were clearly presaged on the assumption that a strategic impetus would shortly be forthcoming from the Commonwealth Government.

It would be an exaggeration to say that the penetration of EDI in federal government agencies has stalled, because there are signs of progress; but the take-up rate is inadequate given the size of the benefits available. This is particularly disturbing in view of the fact that many of the most important agencies appear to have assimilated the basic arguments about EDI's benefits (by way of executive and managerial seminars, supplier briefings and visits to user sites), and to have incorporated the technology into business and IT strategic plans. In many agencies, the education and planning stages are over, but the implementation stage has not been commenced.

There is a lesson to be learned about the diffusion of EDI, from the adoption patterns of electronic mail. Exhibit 2 shows key results from the surveys, together with comparable figures for the private sector resulting from an EDICA survey in late 1993.

Exhibit 2: Late Growth in Email Usage

  	% Agency Use	% Company Use
Internal Use
Within Workgroups	50	54
Between Co-Located Workgroups	51	62
Between Dispersed Workgroups	49	56
Overall	63	74	63

External Use	22	41	41

Responses show upward trends in all categories, most markedly between workgroups and beyond organisational boundaries. From other work undertaken within the Research Programme, this is known to reflect direct, X.400-based inter-connection between pairs of agencies, subscriptions to third-party services, particularly Telecom's Telememo and Compuserve, and direct connection to the Internet.

Email is a technology which has been available for 10-15 years, yet by late 1993 only three-quarters of respondent agencies had installed it, some only very recently. Moreover, only 40% of respondent agencies had email connection to other organisations, and about half of those had commenced only during the last two years. EDI is an altogether more complex challenge than email, and its diffusion could therefore be expected to require much longer. It offers far greater potential benefits, however, and therefore justifies much more vigorous encouragement.


4.2 Purchasing Applications of E.D.I.

Although it is broadly defined, EDI is frequently thought of in relation to the procurement cycle (including quotation, purchasing, delivery, invoicing and settlement). The 1991/92 Survey concluded that, although there had been considerable growth in EDI use by Commonwealth agencies generally, "there has been only a limited growth in purchasing-related schemes" (1992, p.79). There was even less growth between 1991/92 and 1993/94. Exhibit 3 shows the organisations identified as purchasing-related EDI users in late 1993.

Exhibit 3: Agencies Using Purchasing-Related EDI

In addition, a number of government agencies are very actively involved in EDI related to international trade, and are listed in Exhibit 4. The Australian Customs Service's several schemes are world-leading, and high-volume. They have been comprehensively addressed by another, related project being conducted within the Research Programme, resulting in a series of research reports which are summarised in NCG (1994), Clarke (1994b) and Clarke & Jeffery (1994b).

Exhibit 4: Agencies Using International Trade-Related EDI

The 1993/94 questionnaire-based survey of the federal government's top 250 suppliers (SECSAGA), supplemented by structured interviews with 5 selected companies (SISGA) showed that 40 of 110 respondents were currently EDI-capable. Half used PCs, many of them without electronic integration with their internal systems. Half of them were connected to more than 1 network services provider, with the local providers Telecom Plus and NEIS dominant. The majority used ANSI/ASC X12 document standards, but EDIFACT was clearly on the rise. The primary area of application was in relation to quotations and purchase orders, followed by despatch and then invoicing. 60% handled fewer than 1,000 electronic documents per annum, and only three large-volume users were identified (37,000-200,000 p.a.). Most despatched and received documents at least twice daily, some much more frequently. About 40% had 3 or fewer EDI partners, and a further 20% had between 3 and 10.

Some companies had already experienced significant cost-savings, but a few were somewhat negative about EDI's cost-effectiveness for them. Many had already experienced other significant benefits. Nearly half of the respondents considered that entry costs had been high. The vast majority anticipate moderate volume growth. Most cited lack of EDI-capable business partners and lack of integration with internal systems as the primary factors preventing or delaying benefits. Only 2 expressed a very low overall level of satisfaction with EDI, although a further 7 expressed low satisfaction.

Among non-user suppliers, there was evidence of considerable awareness of EDI. The primary reasons they cited for their non-participation were the lack of EDI-capable business partners, inadequate benefits, and the lack of EDI-capability in their in-house applications.


4.3 Non-Purchasing Applications of E.D.I.

In the government sector, it is especially important to keep in view the generality with which the term 'EDI' is defined: the electronic transmission of structured business transactions. In addition to purchasing, selling and paying, government agencies exchange many structured documents with other agencies, at federal and state levels, with corporations. and even with individuals and their agents.

Exhibit 5 lists the agencies identified by the 1993 survey as users, together with two non-respondents known to be active, one from a prior survey and the other from other sources. The taxation authority's scheme has been extremely successful, and short-term growth areas are in company registration documents, and health insurance claims.

In addition to these user agencies, a government-wide gateway for EDI, email, access to external databases, etc. has been established by the IT Services Division of the Department of Adminstrative Services.

Exhibit 5: Agencies Using EDI For Other Purposes


4.4 The Reasons for Using EDI

Throughout the series of surveys, the reasons for implementation were primarily the anticipated contributions to effectiveness and efficiency of the agency's own operations. Benefits for the agency's clients are also nominated in a proportion of cases. The proportion of respondents indicating policy requirements as a reason for use increased in the most recent survey, although, to date, actual policy measures have been limited to general statements. It appears that agencies would welcome a more forceful statement of strategic direction.


4.5 The Realisability of Benefits

The original GPIEEE survey did not seek information about benefits actually gained. In response to the 1991/92 GPIE II survey, those organisations which had implemented EDI generally considered that EDI systems had produced sufficient and in some cases significant benefits (p.58). The expectations of growth within the operational schemes was generally for higher volumes during the following year, and substantial growth during the 2-year and 5-year time-frames (p.39).

In 1993/94, the GPIE III survey established that once agencies had progressed EDI applications beyond the pilot stage, they generally considered that they had gained significant benefits, in many cases in accordance with their plans and expectations (p.41). It is well-known, however, that two of the largest users, the customs and taxation authorities, have experienced difficulties in reaping the 'technology dividend' that they had used to justify the investment. The reason, however, was not inadequacies with the EDI schemes, but rather the very marked reduction in staff attrition rates which accompanied the recession. Because the Labor Government has been loathe to instigate large-scale redundancy programs, the nominally achievable staff reductions have not been feasible. The displaced staff have instead been re-trained, and resources transferred from routine paper-handling towards tasks which make greater contributions to the organisations' objectives.


4.6 The Cost of EDI

In responses to the GPIE II survey, 6 agencies, all of which implemented EDI for strategic reasons, stated that their start-up costs exceeded $100,000 (1992, pp.42-45). For a typical, more modest implementation, the setup costs were reported to be of the order of $15,000, including a PC and modem, diskette or electronic transfer to and from a midrange or mainframe host, mapping between internal and external document formats, and a modest amount of training (Clarke & Colvin 1991, p.70; Jeffery & Clarke 1994a, p.10).

Recurrent cost are highly positively correlated to the level of set-up costs; for example, 4 of the strategic users which invested a great deal in their schemes were among the 13 agencies which reported expenditure of greater than $100,000 on EDI and EDI-related services during the two years 1992 and 1993. The extent of integration of EDI with internal computer-based and manual systems has a significant effect on costs. In GPIE III, about 40% of user agencies ranked network services charges as being the largest recurrent expense (1993, pp.36, 42).


4.7 Impediments to Progress in EDI

In the GPIEEE survey in 1989/90, the primary barriers preventing or retarding the application of EDI were shown to be internal (pp.45-47). External factors (such as excessively high entry and running costs, supplier and external network inadequacies, government regulations, and legal, integrity and security constraints) were cited in only very few cases. The most common reasons for non-use were 'not relevant to my organisation', 'insufficient benefits' and 'too low a priority'. A proportion of these judgements had clearly been reached on the basis of careful assessment of the technology and the respondent organisation's needs, but some agencies appeared to have reached the conclusion hastily and might be expected to revise their opinion at a later stage. The next largest factors were 'still acquiring the necessary knowledge or expertise' and 'technology within the organisation is not EDI-capable' (reflecting the phenomenon commonly referred to as 'legacy technology').

In the 1991/92 GPIE II survey, the primary constraints on faster growth were once again internal. A moderate number of organisations did, however, consider that the lack of EDI-capable business partners was a significant constraint (p.41). Similar reasons were given by organisations who were not intending implementing EDI in the short term, but some said that they were still acquiring the necessary expertise, and others felt that entry or running costs were still too high (p.72).

In the 1993/94 GPIE III survey, there was clear evidence that, although internal factors continued to be significant, the importance of the lack of business partners had grown, as had the costs involved in establishing the necessary EDI infrastructure, and the lack of EDI-capable application software. Some smaller respondents criticised suppliers for their failure to productise EDI and make it as easy to use as the telephone. (For their part, suppliers criticised government agencies for being unprepared to fund the investment needed). Some respondents considered network services providers' charges were high. Concerns about legal problems, data security and data integrity have remained consistently low across the surveys.

There was a single, but significant, instance of a trial not proceeding to full implementation due to the perception among operational staff that jobs would be lost, and the failure of senior management to provide leadership to address the issues.

Other factors which have become increasingly apparent were the plethora of coding schemes for products, services, and suppliers, and the large number of marketplaces in which agencies make their purchases. There is a strong tendency towards industry sector-based schemes, each aligned with a particular network services provider and using a variant of the X12 or EDIFACT standards or of industry-specific standards such as VICS. Agencies are therefore facing the need to connect with multiple VANs, and handle multiple document formats. Among major users and potential users, the inadequacy of network service provider inter-connectivity and inter-operability is increasingly frequently mentioned as a serious inhibitor (1993, pp.35-38).

Among non-using agencies, the most important reason was that EDI was not relevant to the agency, e.g. most of the data moved around externally by them was in textual form. This, and the percentage of agencies stating that EDI offered insufficient benefits, has fallen over the three years, in the latter case from 20% in 1989, to 17% in 1991, to 13% in 1993. Another major reason for not using EDI has been that the organisations were still acquiring the necessary knowledge. This reason fluctuated from 11% in 1989 to 17% in 1991. This was interpreted at the time as indicating that many organisations were actively interested in EDI. The response dropped back to 9% in 1993, which suggests that EDI awareness and education may have reached a degree of maturity.

It was apparent from the questionnaires and interviews that the most critical barrier to the take-up of existing schemes and the implementation of new ones is now the lack of critical mass. It is a classic case of an initiative that is in everyone's collective interest, but in no-one's individual interest; that is to say that individual agencies cannot realise the potential gains from EDI unless a sufficient number of business partners also commit to implementation.


4.8 Implementation Experiences

The study looked at the pace of implementation of EDI systems. Between 60% and 80% of the respondents stated that implementation was running on time. The major reasons given for delays were lack of resources, and diversion of resources to projects with higher priority. Some organisations found problems with trading partners, such as a lack of knowledge, or lack of committed resources. The Australian Customs Service noted that their Sea Cargo Automation system was behind time because they decided it would be easier to expand the scope of the system during development.

Important aspects of successful implementations have been clear understanding of the purposes of the project, a plan to achieve those objectives, and commitment to the project at both management and operational levels. There have been successes with both visionary, long-term projects addressing strategic objectives, and projects of limited scope aimed at short-term cost-reduction. Agencies which are large, complex, have substantial investments in IT, or substantial backlogs of development or maintenance work tend to take a deliberate approach and implement slowly.


4.9 Migration of Pre-EDI Practices Towards EDI

Most agencies use some form of pre-EDI technology for transferring data in a standardised format to or from their business partners, especially other government agencies. In some cases this involves very large data volumes, and very significant benefits. The technologies mentioned most often were physical transport of magnetic media (about 60% in 1993), direct linkage between machines (45%), and bulk electronic file transfer (40%). Initiatives of the Department of Finance relating to the payment of salaries and suppliers' invoices have borne considerable fruit. Other transfers include archival materials, personnel data, and gazettal notices.

In 1989/90, it had been hypothesised that pre-EDI systems would migrate towards EDI in the short to medium term. Little evidence of such developments has ever been unearthed, however, and the SPREG interviews provided arguments to the contrary. The gains available from such migration generally appear in many cases to be marginal, and hence there is insufficient justification for the effort involved.


4.10 Standards

In the purchasing area, the federal government, for all its size, is not a leader in the instigation of schemes, but rather a late adopter of schemes established by particular industries. Many industry-based schemes generally use well-established standards such as X12, VICS (in the retail industry) and CARGOIMP (in the air cargo industry). EDICA has expressed an intention that gradual conversion towards EDIFACT will be undertaken. This is consistent with government policy expressed within the Australian Government Open Systems Interconnection Profile (GOSIP).

The Customs schemes use EDIFACT formats, and the air cargo imports system also uses CARGOIMP. The public administration schemes in many cases use proprietary formats, because of the absence of generic standards suitable for their particular needs, such as annual returns for taxation and company registration.


4.11 EDI Network Services Providers (NSP's)

The primary function of an EDI Network Services Provider (NSP) is to provide a buffer between the sender and the receiver of the message, by storing the message in an 'electronic mailbox' until the recipient is ready to take delivery. There are various additional session-level, message-level, and administrative services which are generally provided by NSPs in Australia.

The responses to the GPIE II survey showed that Telecom Tradelink was easily the most important provider, followed by OTC, which was in the process of being merged into Telecom (1992, pp.46-47). By the time of the GPIE III survey, however, it was apparent that new purchasing applications were tending to use NEIS, a company owned by the N.S.W. State Government and which is contracted to provide electronic commerce services to it (1993, pp.30-34, 37-38). In addition, international trade-related services are primarily provided by a joint venture between AT&T and local companies. A suitably lively competitive market exists, and government agency strategies must reflect that fact.

A number of respondents nominated difficulties encountered with EDI network services, particularly the lack of inter-connection and inter-operability. Inter-connectivity requires that a message from an agency connected to one NSP must reliably reach an addressee connected to another NSP. Unfortunately, several pairs of NSPs are not inter-connected, with NEIS being particularly unco-operative. Whereas inter-connectivity is a concern at the networking level, inter-operability builds upon it at the service level, and concerns such matters as end-to-end audit trails and integrated invoices.

The results of the survey highlight the critical strategic importance of the DAS Gateway. There are, and given the nature of a competitive market-place inevitably will continue to be, several important EDI NSPs. There is a real fear that the present poor standard of inter-connectivity, let alone inter-operability, will continue for some time yet, due to the commitment and investment required by the VANs to overcome the difficulties, the low returns being earned by them, and their natural desire to sustain differentiation from their competitors. These inadequacies can be overcome if the Gateway provides the necessary mediation between the agencies and the VANs.


4.12 The Unavailability of Services

In the GPIE II survey, respondents were asked how critical service availability was to them. The results are shown in Exhibit 7. In most cases, unavailability of EDI services for of the order of 2 hours was not regarded as problematical, but the degree of concern grew considerably if the loss of service were to last as long as half a working day. There were a couple of instances in which even a one-hour outage was regarded as critical, although these tended to be in schemes which had close associations with operational systems, in QANTAS and the Australian Customs Service.

Exhibit 7: Criticality of EDI Down-Time


4.13 EDI and Other Components of Electronic Commerce

EDI in purchasing is merely a part of an electronic commerce strategy addressing the trading cycle as a whole. Other components include pre-contractual activities such as the discovery of potential suppliers through electronic 'yellow pages'; transmission of design data using CALS (Computer Aided Logistics Support - among other interpretations); contract-related activities such as tender and purchase order transmission and on-line trading; logistics-related activities such as transport of goods and transmission of delivery advices; and post-purchasing activities such as invoice and payment instructions transmission (EDI-Electronic Funds Transfer - EDI/EFT or F-EDI), 'evaluated receipts settlement', and statistical reporting. For a discussion of the range of technologies, see Clarke (1993c) and GPIE III (1993, pp.59-61).

Other components which the GPIE III survey disclosed as being in use were on-line access to other databases (29 instances), and CD-ROM (48). Electronic Bulletin Boards have only recently begun being widely appreciated as useful for purchasing-related purposes, and so are used less (18 responses), but may grow rapidly as a further 10 agencies were considering them. Other tools which rated strong mentions were the Department of Administrative Service's own diskette-distributed Period Contract Awareness Service (PCAS) and Computer-Aided Purchase of Computers (CAPOC) schemes.


5. POLICY MEASURES - 1988-1994

During the course of the project, a considerable range of measures has been implemented by key agencies, in an endeavour to facilitate the exploitation of EDI in government. The most important and multi-faceted of these have been undertaken by Purchasing Australia, an arm of the Department of Administrative Services (DAS). Since 1988, it has funded a succession of research projects and consultancies to ensure that policy deliberations are informed, has developed and disseminated a quality, introductory audio-visual package ('EDI - A Better Way'), publishes and disseminates a bi-monthly newsletter called EDICAST, and will shortly be disseminating Facts Sheets on various aspects of EDI, and an Implementation Guide for government agencies. Associated with this has been the development by DAS IT Services of a corporate gateway, designed to support access by agencies generally to the EDI, email and database services of multiple providers.

In the Australian federal government, no central agency has responsibility for IT matters. Co-ordination among agencies is a function of the Information Exchange Steering Committee (IESC), and in particular its EDI Sub-Committee. The Commitee's Chair and Secretariat are provided by the Department of Finance, but membership and resourcing involve many different agencies. Early attempts were somewhat stunted by Committee-members' commitments to their own agencies' projects, but a greater degree of inter-agency collaboration and mutual support has recently been apparent. The IESC has provided some financial support for, and co-ordination of, several research projects and consultancies, a program of senior officer briefings has been conducted, and it is anticipated that a clearer program of government stimulation of EDI usage may emerge in the near future.

Parallel developments are occurring in some other areas. Of particular importance is the National Consultative Group on Transport EDI, a collaboration between the federal Department of Transport, the national association of EDI users (EDICA), and the international trade community company (Tradegate). The team from the Australian National University has been much involved with developments in this area also. It is documented in NCG (1994), Clarke (1994b) and Clarke & Jeffery (1994b).


6. POLICY RECOMMENDATIONS - 1994-1996

In the second survey report, published in September 1992 (GPIE II, p.80), the following measures were identified as being necessary to overcome the factors which were retarding the growth of existing schemes and the implementation of new ones:

The SPREG Report, based on structured interviews with 20 agencies, concluded that "the potential benefits of purchasing-related EDI will not be gained unless a firm and clear policy decision is taken at government level to stimulate developments" (1992, p.27).

Following the 1993/94 GPIE III, SECSAGA and SISGA surveys, further insight has been gained into these needs. Some suppliers considered that it was incumbent on the federal government to take quite general responsibility for the promotion and stimulation of EDI adoption in Australian industry, at the very least using the government's purchasing muscle to encourage adoption of EDI by its suppliers. There was a generally-felt expectation that the government should co-ordinate the activities of its many agencies in relation to electronic commerce. There was also considerable dissatisfaction expressed at the low volumes of purchase orders being transmitted using capability installed at the behest of the N.S.W. State Government.

The 1992 conclusions have recently been further refined and articulated Clarke & Jeffery 1993a, 1993b). It has been proposed that the central element in the Commonwealth's EDI implementation strategy during 1994-96 should be subsidies to a few key agencies to inter-connect via the DAS Gateway to a few key supplier-clusters. These projects would need to be observed, and their performance measured, in order to distill lessons from them, and publish that information to other agencies and suppliers. The most apparent supplier-cluster is the automotive and petroleum industries, because these comprise major corporations which are generally recognised by companies throughout Australia to be leaders in the exploitation of technology. The automotive industry presents a difficulty, in that its use of EDI is predominantly related to purchases rather than sales. Nonetheless, these companies have the infrastructure in place, and should be approached by the Commonwealth with a view to having them implement EDI transactions with one of their most important customers. Another key sector is the less glamorous, but very important stationery and printing industry, which contains a substantial number of suppliers which are already EDI-capable.

In addition to the industries discussed above, another should be considered, but using a different approach. Suppliers of information technology services, and especially products, have been remarkably slow to apply the technology they sell. Meanwhile, agencies within the Administrative Services portfolio have been pro-active in applying information technology to purchasing information, and in particular to the tendering and period contracts systems. Given the highly competitive nature of the IT marketplace, the opportunity exists for the government to mandate the use of EDI by IT products suppliers, to receive requests for tender, submit tenders, and provide updates to the period contracts database.

Serious consideration should also be given to providing direct subsidies to user-organisations in strategically important industry segments. It would be inappropriate and ineffective to provide cash. Rather:

It would be possible to significantly increase the resolve of key industry associations to encourage and support their members in relation to electronic commerce, and to enhance their standing with their members, if these measures were conceived, planned, implemented and reviewed in conjunction with those organisations.

The difficulty must also be confronted that many small companies are not yet computerised, and may remain unsophisticated IT users. It is therefore likely that intelligent fax-based communications may be important in some industry segments. This implies that fax interfacing for both outgoing and incoming messages should be a high priority for the DAS Gateway development.

The studies conducted during the last six years have made very clear that, for many organisations, significant and/or more immediate benefits can be gained from elements of electronic commerce other than EDI. Of particular importance are 'white-page' and 'yellow-page' directories and bulletin boards. The N.S.W. State Government expressly adopted the approach of investing in directories first, and seeking growth in EDI volumes in a second wave. Extension of existing and new electronic information services, and mandating by the Commonwealth of their use by agencies and suppliers, would provide an important impetus to electronic applications in purchasing. Important examples include tender advertisements, notices of sales of used and surplus goods, supplier contact points, purchasing officer contact points, agency delivery points, period contracts and supplier product catalogues. The Australian Government Publishing Service (AGPS), a division of the Department of Administrative Services, runs the Government On-Line Directory (GOLD), which provides an appropriate basis for some of these vital initiatives.

It is in the Commonwealth's interest for industry support to extend beyond user-organisations to encompass key product and services suppliers. Suppliers of purchasing-related packages should be directly encouraged to implement EDI interfaces within their next releases. This could be achieved through direct subsidies to key suppliers, development grants awarded on the basis of competitive bidding, and/or period contracts for which EDI-capability is a condition of tender. A special category of packages which requires attention is inexpensive interfacing software, to provide an entry point for small companies into EDI. Similar approaches should be adopted in relation to EDI network services suppliers, in order to overcome the inadequacies in inter-connectivity and inter-operability.


7. CONCLUSIONS

The first GPIE report, published in September 1990, concluded that "there are many existing document flows which could be far more effectively and efficiently handled by electronic means. This is because EDI is not restricted only to trading-related matters, but is applicable to all structured communications between organisations which have formal relationships, particularly where the volumes and the frequency are significant. Moreover, there are areas in which rationalisation of existing procedures and organisation structures [these days called 'business process re-engineering' or 're-design'] could reap even greater benefits, for agencies and for their clients" (GPIEEE, p.48).

Those observations still hold, but it is all too apparent that market failure is occurring. A serious and concerted program by government policy agencies is essential, in order to convert the potential for national competitive advantage into a reality. The European Community recognised that problem many years ago, and acted. The longer that Australia waits, the more the national competitive opportunity is turning into a national competitive necessity.

As early as May 1992, it had been possible to determine that "the greatest benefits can be gained by envisioning, and managing the transition towards, an infrastructure which supports and integrates the various purchasing-related electronic services" (SPREG 1992, p.27). The focus has moved beyond EDI alone, to encompass complete electronic business support systems for the particular needs of specific industry segments and sectors.

This is most easily described in the purchasing area, where the focus is increasingly on electronic commerce as a whole, including wide-area access to directories of suppliers and supplier catalogues, tendering for custom goods and services, and on-line spot-markets for commoditised goods and services, as well as EDI and email support for purchasing, logistics, settlement and reporting. Beyond purchasing, there is a need to re-conceive many different aspects of public administration in the context of the information society and the networked world, using such notions as electronic delivery systems, IT-supported government business and information-age public administration.


Acknowledgements

This research was conducted as part of the Australian National University's Research Programme in Supra-Organisational Systems. It enjoyed support from the University's Bruce Fund, the University's Department of Commerce, Purchasing Australia within the Commonwealth Department of Administrative Services, and its staff-members, and the Information Exchange Steering Committee (the primary vehicle for coordination among federal government agencies' IT activities). The Programme is largely self-funding, and copies of the full survey reports are accordingly not distributed gratis, but offered for sale. Copies of short articles are available on request from the author.

My personal thanks to the research assistants, students and colleagues who have assisted me in this project, and whose names appear below as authors and co-authors. I record my particular gratitude to Chris Akeroyd, Director in the Policy Branch of Purchasing Australia, partly for his efforts in arranging financial support for many of the sub-projects, but especially for his substantial intellectual contributions to the undertaking as a whole.


Survey Reports (in chronological order)

Clarke R.A., Pedler M., Swatman P.M.C. & Campbell P.C. (1990) [GPIEEE] 'Commonwealth Government Practices and Intentions Relating to EFTS, EFT/POS and EDI: Survey Report' 100pp., Department of Commerce, Australian National University September 1990

Clarke R.A. & Colvin P. (1991) 'EDI & Commonwealth Purchasing: Concepts Paper' Consulting Report for the Department of Administrative Services, Xamax Consultancy Pty Ltd, Canberra (May 1991), 70 pp.

Ryan J. (1991) 'Electronic Lodgement System' Proc. Conf. SOST'91, Adelaide, October 1991. Republished in Clarke R.A. & Cameron J. (Eds.) 'Managing Information Technology's Organisational Impact, II' Elsevier, Amsterdam, 1992

Clarke R.A., Campbell P.C. & Maguire D. (1992) [SPREG] 'Survey of Purchasing-Related EDI in Government' Report for the Information Exchange Steering Committee, and the Departments of Finance and Administrative Services, Canberra (May 1992)

Clarke R.A., Campbell P.J. & Telfer S.G. (1992) [GPIE II] 'EDI: The Practices and Intentions of Agencies of the Commonwealth Government: 2nd Survey Report, 1991/92' 140pp., Department of Commerce, Australian National University September 1992

Goleby M. (1992) 'Case Study of EXIT: The Australian Customs Service's Export System' Research Report, Department of Commerce, Australian National University, November 1992

Teitler U. (1992) 'The Department of Finance's Electronic Data Transfer Facilities' Research Report, Department of Commerce, Australian National University December 1992

Cameron J. (1993) 'EDI and the Air Cargo Industry' Research Report, Department of Commerce, Australian National University, June 1993

Jeffery K. (1993) 'SISGA: Structured Interviews with Suppliers to Government Agencies: Survey Report, November-December 1993' Research Report, Department of Commerce, Australian National University December 1993

Jeffery K. (1994) 'The Sea Cargo Automation Pilot Project' Research Report, Department of Commerce, Australian National University, January 1994

Clarke R.A. & Jeffery K. (1993a) 'GPIE III: Government Practices and Intentions in Relation to EDI: 3rd Survey Report, September-December 1993' 90pp., Department of Commerce, Australian National University December 1993

Clarke R.A. & Jeffery K. (1993b) 'SECSAGA: Survey of the EDI Capability of Suppliers to Australian Government Agencies: July-November 1993' 70pp., Department of Commerce, Australian National University December 1993, rev. May 1994

Jeffery K. & Clarke R.A. (1994a) 'GPIE III: Structured Interviews with Government Agencies: Survey Report, February-March 1994' 90pp., Department of Commerce, Australian National University March 1994

NCG (1994) 'EDI Implementation in the Transport Sector: An Assessment / Stocktake: Preliminary Report for Review' National Consultative Group in Transport EDI, Department of Transport, Canberra, March 1994

Clarke R.A. & Jeffery K. (1994a) 'EDI in Australia - 1993/94' 100pp., Department of Commerce, Australian National University Forthcoming, June 1994

Jeffery K. & Clarke R.A. (1994b) [EPU] 'Case Study: Electronic Purchasing by Units' Research Report, Department of Commerce, Australian National University Forthcoming, June 1994

Clarke R.A. and Jeffery K. (1994b) 'International Trade EDI in Australia' Department of Commerce, Australian National University Forthcoming, July 1994


Short Articles

Clarke R.A. (1992a) 'EDI und die öffentliche Verwaltung' ('EDI and government administration', in German) Verwaltung/Organisation/Personal - VOP 14,2 (März/Apr 1992)

Clarke R.A. (1993a) 'EDI in Australian Government Agencies: A Research Report' Proc. 3rd World Congress of EDI Users, Gold Coast, Australia, March 1993

Clarke R.A. (1994a) 'Progress in the Use of EDI in the Australian Federal Government, As Revealed by Longitudinal Research' EDI Window 2 (April 1994)

Clarke R.A. (1994b) 'EDI in Australian International Trade and Transportation' Proc. Conf. 7th Int'l EDI-IOS Conference, Bled, Slovenia, June 1994

Clarke R.A. (1994c) 'The GPIE Saga: A 6-Year Longitudinal Research Project Into Australian Government Practices and Intentions Regarding EDI' Paper presented at the EDI Research Symposium, Bled, Slovenia, June 1994

Clarke R.A. (1994d) 'The Path of EDI Application in Australian Government Agencies: 1989-94' Proc. 5th World Congress of EDI Users, Brighton, UK. Forthcoming, June 1994 [this paper]


References

Australasian EDI Report Monthly newsletter (successor to EDI Research Australia), 1,1 in Feb/Mar 1993, Technosocial Research Services, P.O. Box 21, Brunswick VIC

Clarke R.A. (1992b) 'A Contingency Model of EDI's Impact on Industry Sectors' J. Strat. Inf. Sys. 1,3 (June 1992)

Clarke R.A. (1992c) 'Data Interchange and International Competitiveness' Business Council of Australia Bulletin (September 1992)

Clarke R.A. (1993b) 'EDI in Australia: A Status Report' Dept of Commerce, Australian National University, June 1993

Clarke R.A. (1993c) 'EDI Is But One Element of Electronic Commerce' Proc. 6th Int'l EDI Conf., Bled Slovenia, 7-9 June 1993

Clarke R.A., Griçar J., de Luca P., Imai T., McCubbrey D. & Swatman P.M.C.(1991) 'The International Significance of Electronic Data Interchange' in Palvia S. Palvia P. & Zigli R.M. (Eds.) 'Global Issues in Information Technology' Ideas Press, 1991

Clarke R.A. & Hudson P. (1992) 'Corporate EDI Gateways: Their Rationale, Functions and Architecture' Proc. 5th Int'l Conf. EDI, Bled, Slovenia (September 1992)

EDICAST Periodical of the Department of Administrative Services, Purchasing Australia, 1992-. Available from the Editor, Eva Burns & Associates, P.O. Box 87, Waverley NSW 2024, Fax: (02) 369 4245

EDI Council of Australia Newsletters, particularly those reporting on progress in EDICA's Government Working Party. Available from EDICA, 854 Glenferrie Rd, Hawthorn VIC 3122, Fax: (03) 818 3129

EDI Research Australia Monthly newsletter, Feb/Mar 1989 - December 1992, Technosocial Research Services, P.O. Box 21, Brunswick VIC

Swatman P.M.C. (1993) '' Unpublished PhD Thesis, Curtin University of Technology, W.A., 1993

Swatman P.M.C. & Clarke R.A. (1990) 'EDI's Organisational, Sectoral and International Implications' in Berleur J. & Sizer R. (Eds.) 'Information Technology Assessment', Elsevier / North-Holland, 1991


Navigation

Go to Roger's Home Page.

Go to the contents-page for this segment.

Send an email to Roger

Created: 16 January 1998

Last Amended: 16 January 1998


These community service pages are a joint offering of the Australian National University (which provides the infrastructure), and Roger Clarke (who provides the content).
The Australian National University
Visiting Fellow, Faculty of
Engineering and Information Technology,
Information Sciences Building Room 211
Xamax Consultancy Pty Ltd, ACN: 002 360 456
78 Sidaway St
Chapman ACT 2611 AUSTRALIA
Tel: +61 6 288 6916 Fax: +61 6 288 1472