1989 Report of the Auditor General of Canada

Main Points

7.1 This year we examined how efficiently the federal government uses its telecommunications services. In 1962, the Royal Commission on Government Organization (Glassco Commission) emphasized the importance of co-ordinating departments' use of voice communications. We believe the same is true today for data communications. However, savings from economies of scale by combining voice communications links were not being pursued by the government in 1962 and are not being pursued for data communications 27 years later (paragraphs 7.8 to 7.16).

7.2 We estimate the short-term potential for government-wide savings in data communications to be between 20 and 30 percent of operating expenditures. This translates into annual savings of $30 million to $45 million, based on current data communication operating expenditures of $150 million. In the longer term, we estimate that additional savings of at least 15 percent of voice operating expenditures are achievable with the integration of voice and data on government-wide networks (7.17 to 7.21).

7.3 We believe that, to achieve these potential savings, a central focus is required for government-wide administration of both voice and data communications as a common service, through a responsible body, such as a common service agency, held accountable to achieve an efficient and cost-effective telecommunications service (7.22).

7.4 The Treasury Board and the Department of Communications have taken little action to enforce a common service agency role for the efficient management of telecommunication services on a government-wide basis. Departments have no reason to consider anything more than their own requirements, resulting in a proliferation of parallel networks across the country. With no clear leadership, the government is not positioned to take advantage of either the significant existing opportunities or the technological innovations that should soon be available (7.23 to 7.38).

Introduction

Audit Objective and Scope

7.5 Data and voice telecommunications are rarely given the same level of emphasis and attention as the more visible computer hardware and software products. However, the federal government spends a surprising $1 billion annually for telecommunications goods and services. These goods and services form an essential component of many of the information systems used by departments for program delivery and administration.

7.6 Our examination addressed the way departments use telecommunications services. Our audit objective was to identify opportunities for realizing significant savings in voice and data communications expenditures. We focussed on the policies and practices of five major user departments, the activities and role of the Government Telecommunications Agency (GTA), and the role of central agencies.

7.7 The five departments selected were:

• Agriculture
• Employment and Immigration (CEIC)
• Environment (DOE)
• Public Works (DPW)
• Supply and Services (DSS)

The $1 Billion Government Telecommunications Environment

Today's management of data communications mirrors the fragmented management of voice communications in 1962

7.8 We estimate, on the basis of forecasts for 1988-89, that the federal government spends in the order of $1 billion annually for telecommunication goods and services. Approximately 48 percent, or $490 million, is for operations, primarily the lease of common carrier services. The remaining 52 percent, or $540 million, is for capital expenditures. Of the $490 million for operations, almost 70 percent represents voice traffic (telephones) and 30 percent data traffic. Departments purchase $320 million in services directly from the common carriers of which $70 million is arranged through the Government Telecommunications Agency (GTA). GTA provides the remaining $170 million in services and bills departments for them.

7.9 Exhibit 7.1 shows the growth in telecommunications operating expenditures in the past 10 years. In coming years, the most significant area of growth will be data communications. The common carriers estimate that industry-wide data traffic will increase by at least 15 percent a year on a compounded basis.

(Exhibit not available)

7.10 In 1962, the Royal Commission on Government Organization (the Glassco Commission) recommended the establishment of a Government Telecommunications Agency. At that time, total government telecommunication expenditures --for voice and teletype -- were $50 million annually. The Glassco Commission expressed several major concerns of direct relevance to our audit. One of these addressed the potential savings from pooling and integrating departmental requirements:

...a single co-ordinating agency must be created with responsibility to assist departments in achieving suitable standards of communications performance, eliminating duplication of facilities and initiating such integration and pooling of services as will promote the realization of the important economies which are possible.

Significant savings are realized through government-wide management of voice communications

7.11 Voice communications conform to standard, carrier-provided services. We found that, within this environment, GTA provides an efficient inter-city voice service for departments and agencies by administering a private "Intercity Network" on circuits leased from the common carriers. This network makes possible savings in the area of 40 percent over the direct distance dialing daytime rates for message toll service. All the departments we examined use the Intercity Network to take advantage of the potential economies of scale. The network results in annual savings to the government estimated by GTA to be approximately $43 million.

7.12 In voice telecommunications, most departments demonstrate concern for minimizing costs. The use of GTA's intercity voice network creates efficient use of voice communications, where departments have access to it.

The complexity of data communications demands more active government-wide co-ordination

7.13 The federal government does not receive any preference in rates, which would be viewed by the Canadian Radio-television and Telecommunications Commission (CRTC) as a discriminatory practice. Regulated telecommunications rates apply to all users. However, the CRTC recently set a precedent by approving a discount based on guaranteed volumes of traffic for a packet switched service. This tariff filing was a made in response to GTA's requirement for such a service. The Government Packet Network (GPN) takes advantage of this volume discount opportunity. GTA has "bulk-purchased" capacity, enabling users of the GPN to obtain a 20 percent discount on rates. The government-wide data traffic was estimated by GTA to be sufficiently large to guarantee enough volume to secure this discount. However, by the time GTA had established the GPN many departments had already developed and implemented their own networks and chose not to switch to the GTA service. Although the GPN's current volume is sufficient to maintain the 20 percent discount, the service is not fully utilized. Therefore, the full cost savings potential of the GNP that can be realized through wider government use is not being achieved.

7.14 Most departments with significant data communication requirements have developed private networks to meet their own specific needs. These private networks were seldom developed with active GTA participation or consultation. Where departments had contacted GTA, for example to develop AgriNet, CEINet and UIT, GTA's involvement was minimal. While each department appears to meet its individual needs reasonably efficiently, together they are losing significant cost-saving opportunities by not having a government-wide efficiency perspective. With little central focus or incentive to ensure that a government-wide perspective is emphasized, departments have given little consideration to achieving government-wide efficiencies in telecommunications.

7.15 In all the departments we examined we observed cross-country data links that had developed independently from other departments with similar traffic routes. As a result, departments have created many parallel networks that are more costly and inefficient because of their smaller volume.

7.16 This multiplicity of private data communications networks is inefficient. Many of these parallel networks link regional offices in the same cities to Ottawa, with more expensive, lower speed data circuits. This results in duplication of the capital investment in modems, multiplexers and switches and in the lease of relatively high-cost, low speed circuits.

7.17 It is difficult to calculate the total potential savings because of the large number of different networks and the lack of government-wide data traffic forecasting or analysis. However, our estimates, which we believe to be conservative, show potential savings of 20 to 30 percent of data communications operating expenditures annually. On current operating expenditures of approximately $150 million, savings of between $30 million and $45 million could be achieved in the first year of full data consolidation. The potential for such annual savings will continue to increase as the total use of telecommunications by the government increases.

7.18 The potential for total telecommunications savings falls under four major headings:

• savings by multiplexing individual data networks into a few high-speed networks that enable the government to take greater advantage of existing tariff structures;
• more efficient use of the existing circuits;
• person-year savings in individual departments through organizational rationalization; and
• savings achieved by integrating voice and data services.

7.19 Within approximately five years, we estimate that further savings of 15 percent of total operating expenditures will be attainable through the integration of voice and data services. Not taking action now to improve the structure of data communications in the federal government will result in considerable additional downstream costs. We believe that the significance of future benefits derivable from new technologies will eventually force the government to address its telecommunications operations. The savings currently available in data transmission, as discussed above, will also be available in voice transmission if the government pursues the integration of voice, data and other telecommunication services. To minimize downstream expenses and to take advantage of the full potential savings, the government will have to take a consolidated approach to satisfying departmental data traffic needs.

7.20 We found that planning for data services is done by individual departments, which generally does not take advantage of the potential benefits of a common service approach. Although the planning for such services is not well documented, we found evidence that departments were striving for improved efficiency within the limited view of departmental needs.

7.21 It is disappointing to note the similarities between the deficiencies in the management of voice communications reported by Glassco in 1962, and the problems in the current data communications management environment. The message in both cases is that the government is not pursuing the benefits that could be derived from economies of scale.

7.22 To achieve the significant savings that are attainable, we believe that a central focus is required for government-wide administration of both voice and data communications as a common service, through a responsible body such as a common service agency, held accountable to achieve an efficient and cost-effective telecommunications service.

The government has failed to take advantage of significant opportunities to reduce costs by combining departmental traffic into government-wide networks

7.23 Combining departmental data traffic would allow for more efficient use of switching and transmission facilities and would enable the government to benefit from lower unit costs. Combining separate transmission needs would lead to better use of available network capacity and would allow the government to take advantage of a sliding price scale for data speed costs (costs per bit per second decrease as circuit capacity increases). Larger capacity, and therefore fewer, lines could be leased from carriers to meet the planned needs of departments. Technology developments have made it possible to increase efficiency in telecommunications through the use of broadband capacity for voice and data traffic. A single, high-capacity "pipe" can be leased on which many voice and data circuits can be multiplexed. Such an approach would also take advantage of variations in demand and lower unit costs. These are options that the federal government telecommunication managers could be pursuing.

7.24 Savings can also be achieved with better planning and co-ordinating of networks within the government. Activities like data traffic forecasting, which common carriers and major users of telecommunications services regularly undertake to support their planning and operations, are not done on a government-wide basis. This restricts GTA's ability to identify overall government volume and take advantage of potential cost saving opportunities. With better knowledge of operations and government-wide potential, a common service agency could plan for and negotiate more efficient communications services. As volumes of traffic become large enough, other options such as high speed data services become viable alternatives where the greatest potential exists for cost savings.

7.25 Exhibit 7.2 shows that as data traffic is combined, other options become available that offer more cost-effective service to the government as a whole, at reduced tariff rates. We found that the departmental data communication links operate at low data speeds (9.6 to 56 Kbps). Combining data traffic to obtain larger single volumes would permit the government to move down the cost curve by using higher capacity, less expensive per bit lines (1.544 Mbps).

(Exhibit not available)

7.26 The penalty for failing to take action soon is likely to be a further balkanization of telecommunications facilities. Departments may continue to develop independent networks less able to benefit from future advances in technology and, therefore, further distance themselves from opportunities to gain from government-wide efficiencies.

7.27 Although the government obtains savings in voice communication expenditures through GTA, gradual reductions in the tariff rates for message toll and WATS services are eroding them. As this trend is likely to continue, and rates continue to decline for message toll services, GTA may be unable to maintain its overall economic viability.

7.28 The six departments included in this audit have all agreed in general with the conclusions reached, and with the potential for significant cost savings. They gave support to the concept of a government-wide focus to telecommunications and to the notion of a common service network. However, a caution was issued that any common service approach must ensure that it can maintain an adequate level of service to all users and satisfy the needs of departments requiring special consideration. We support these concerns.

GTA's mandate is unclear and has resulted in significant missed opportunities for efficiency gains

7.29 The reasons for GTA's unclear mandate became an important focus of the audit. We began by examining the roles and responsibilities of the organizations involved with telecommunications in the federal government.

7.30 The Treasury Board's Common Service policy designates GTA as the mandatory common service agency for the provision of telecommunications within the federal government. However, this is in conflict with the Telecommunications policy, under which departments can act on their own to obtain telecommunications services.

7.31 The Department of Communications, under Section 5(d) of the DOC Act, shall "...plan and co-ordinate telecommunications services for departments, boards and agencies of the Government of Canada." Within this legislation, GTA's mandate is " . . . to plan, establish and manage telecommunications facilities and services that will satisfy the requested needs of federal departments and agencies on an economic basis." This objective lacks the strong mandatory wording of the Treasury Board's Common Service policy. In addition, the Treasury Board does not enforce its mandatory common service policy for GTA. When reviewing submissions for telecommunication projects, we found examples of Treasury Board giving approval to projects before requiring that departments confer with GTA. For example, approved Treasury Board submissions from DPW and CEIC both stated that the departments should seek GTA's assistance, rather than enforcing the GTA role before approving the submission.

7.32 GTA has not established its administrative role for government data communication services. Inactivity on the part of the Treasury Board and DOC in establishing a co-ordinating body for data services has contributed to the lack of focus on administering government-wide data communications. The Treasury Board's current management philosophy, embodied by IMAA (Increased Ministerial Authority and Accountability), emphasizes that departments have increased responsibility to manage within their own scope of operations, with reduced intervention from central agencies. However, if the full potential of government-wide opportunities to improve efficiency is to be realized, it will require a central, co-ordinated focus. We believe that departments are not well placed to take full advantage of these opportunities without involvement from the central agencies. We also noted a perception that IMAA discourages the government-wide co-ordination necessary to take advantage of telecommunications efficiency opportunities.

7.33 We found that conflicting authorities in the Treasury Board Common Service policy and Telecommunications policy have resulted in many departments planning and operating their own private data networks. Departments interpret the Treasury Board policy as giving them the option to lease common carrier packet switched data services rather than using the less expensive, GTA administered, GPN. Unless there are overriding security issues, the use of any service other than common service leads to inefficient telecommunications usage.

7.34 The Treasury Board established the Telecommunications Advisory Committee (TAC) to advise GTA on long-range planning and co-ordination of the use of telecommunications on a government-wide basis. The Advisory Committee on Information Management (ACIM), an interdepartmental advisory committee to the Treasury Board Secretariat, deals with the broader area of information technology. Neither these committees nor the Treasury Board appear to have a clear-cut position on whether a mandatory common service organization is required. This has contributed to the segmenting of government data traffic by departments and the development of networks which, collectively, result in higher costs. ACIM has recognized that the current data communications environment should be addressed and has tasked a sub-committee to report on telecommunications in the government.

7.35 The separation and lack of active co-ordination between TAC and ACIM also furthers the fragmented approach to telecommunications across the government. In most departments it is policy to regard telecommunications as two distinct services -- voice and data. Advances in technology are making this division less appropriate for the future.

7.36 The Treasury Board has no apparent mechanism for ensuring that departments treat telecommunications issues with a government-wide perspective. None of the parties -- Treasury Board, ACIM, TAC, or GTA -- has shown any effective leadership in this area. As a result, the federal government is losing an opportunity for significant savings in the order of $30 million to $45 million a year at the present level of data communications expenditures.

Departments lack confidence in GTA's abilities

7.37 We noticed that user departments view GTA as simply a negotiator of bulk deals with the telephone companies. We believe that GTA's working relationship with departments has contributed to the perception that it cannot provide sophisticated data communications assistance and is mainly there for voice communication services. GTA is not positioned to exploit a dynamic technology for the greater efficiency of the government as a whole. For example, DPW has been unable to wait for GTA to identify and develop necessary services. DPW took action to provide a more efficient cabling system for its government buildings, to reduce costs.

7.38 There are several areas where co-ordination is needed to examine and evaluate the opportunities for government-wide savings. We believe action needs to be taken to enable departments, and the government as a whole, to plan for changes in technology and potential efficiencies in the following areas:

• new building cabling systems;
• assistance in understanding and remaining current on regulatory changes;
• in-depth data traffic analysis for the government as a whole;
• advice and leadership on the directions departments should take to be prepared for new service offerings such as Integrated Services Digital Network and voice-data integration planning; and
• telecommunications technology advances offered by the common carriers, and the application of these advances to government services.

7.39 The Treasury Board should:

• ensure that an administrative infrastructure exists for all government telecommunications services, supported by a clear policy framework;
• ensure that this infrastructure supports a co-ordinating body, such as a common service agency, that has a government-wide perspective, responsibility, authority and capability to obtain or provide all telecommunications services for the departments and agencies of the federal government in an economical, efficient and effective manner; and
• ensure that the savings attainable with a government-wide approach to telecommunications administration are achieved through a co-ordinating body, and hold it accountable for realizing those savings.

Management's response: The Treasury Board Secretariat (TBS), and the members of the Advisory Committee on Information Management (ACIM), are well aware of these issues and have for some time been addressing them.

In 1987, the Treasury Board approved and promulgated an Information Management Policy Overview entitled "Strategic Direction in Information Technology Management in the Government of Canada - 1987" which identifies these issues and other critical issues for the efficient use of Information Technology (IT) in the government. In 1988, the TBS promulgated, after endorsement by ACIM, a proposal for a new strategic planning process for the management and co-ordination of IT, with special emphasis on common service organizations and their planning role.

A proper planning base for IT across government is required if the potential dramatic savings and efficiencies are to be achieved. In close cooperation with ACIM, the TBS has proposed drastically new policy approaches to IT management, and incorporated the work of the various ACIM Working Groups. The proposals will be submitted shortly to Treasury Board, and implementation will start immediately after approval.

The Working Group of ACIM mentioned in the AG study, has been examining the situation for several months in some depth, and will soon table its report. Action will then be taken to put in place the necessary changes within the above mentioned new overall policy framework.

The TBS, as well as ACIM, see the problem as more complex than indicated by the AG study. The merging that is occurring is not only "voice" and "data", but also "telecommunications" and "computer systems". The government should therefore not only look to telecommunications carriers for solutions, but also to other suppliers. In this context, having a common service agency offering a "mandatory" telecommunications service, based almost solely on carrier offerings is not an adequate long-term solution. The Common Services Policy will be revised accordingly, in the light of developments in the use of IT and in the industry.

7.40 The Department of Communications should:

• review the past performance and current capabilities of GTA to determine whether it is the appropriate vehicle for a common service agency for all government telecommunications;
• take action to clarify the mandate of GTA with respect to the intent of the Treasury Board's administrative policy framework; and
• re-examine the provision of common services for telecommunications, with a view to studying the role of a common service agency as re-seller of carrier services and negotiator of agreements similar to "master standing-offers" for services.

Department's response: The Department accepts the conclusions and recommendations of the audit and makes the following additional comments. Despite the very limited resources available, GTA has managed to implement common services for data communication in the forms of a VSAT network, a packet switched network and conversion of the circuit switched network of digital facilities, all of which have brought and will bring further economies to the government in the data communications area.

It should be recognized that DOC does not have any power to enforce a common service agency role (implied mandatory) and, in the IMAA environment, such an action would be questionable. Short of the mandatory inference, the Department and GTA have vigorously pursued a common service role even to the extent of creating a marketing group to ensure the benefits of the common service approach are widely publicized throughout government.

The Agency has developed a plan (GTN 2000) which clearly shows the direction to achieve the greater savings which will become available in the future.

GTN 2000 examines the major changes in technology and regulation now unfolding and provides the blue print by which GTA will evolve a single universally accessible network to carry voice, data, image and video in the future. The baseline for this network is the existing circuit switched network which is being converted to a universal digital technology network as permitted by economy, technology and regulation. A Request for Information issued to industry in 1988 indicates that it will be feasible to establish a pilot digital channel network service by 1990.