Chapter 4: Key Findings
- Weak governance of pan-government issues related to ICT
- Agency governance mechanisms are weak in respect of focus on ICT efficiency and understanding of organisational capability to commission, manage and realise benefits from ICT-enabled projects
- The business as usual ICT funding in agencies is not subject to sufficient challenge and scrutiny
- Disconnect between the stated importance of ICT and actions in relation to ICT skills
- There is no whole-of-government strategic plan for data centres. In the absence of such a plan, the government will be forced into a series of ad hoc investments which will, in total, cost significantly more than a coordinated approach
- The Government ICT marketplace is neither efficient nor effective
- There is a significant disconnect between the Government’s overall sustainability agenda and its ability to understand and manage energy costs and the carbon footprint of its ICT estate
After detailed analysis of the evidence obtained from within and outside the Australian Government (the Government) and the survey results, the review identified the following key findings:
- There is weak governance of pan-government issues related to ICT.
- Agency governance mechanisms are weak in respect of their focus on ICT efficiency and an understanding of organisational capability to commission, manage and realise benefits from ICT-enabled projects.
- The business as usual (BAU) ICT funding in agencies is not subject to sufficient challenge and scrutiny.
- There is a disconnect between the stated importance of ICT and actions in relation to ICT skills.
- There is no whole-of-government strategic plan for data centres. In the absence of such a plan, the Government will be forced into a series of ad hoc investments which will, in total, cost in the order of $1 billion more than a coordinated approach over a 15-year period.
- The government ICT marketplace is neither efficient nor effective.
- There is a significant disconnect between the Government’s overall sustainability agenda and its ability to understand and manage energy costs and the carbon footprint of its ICT estate.
- The remainder of this chapter sets out the significant evidence that supports these findings.
The lack of a vehicle for political leadership and weak governance of pan-government issues related to ICT, coupled with agency autonomy with regard to ICT investments, has led to significant fragmentation and duplication. These issues include:
- duplication of telecommunication links, both domestically and internationally – for example, there are 37 links between Sydney and Canberra alone, five links from four different agencies into Jakarta, and five links from three different agencies into London
- duplication of gateways, with 69 gateways and a number of companies hosting the gateways that are providing much the same service, yet no aggregation is taking place – for example, one supplier provided 27 gateways for the Government
- lack of standardisation in common business processes across government
- a fragmented approach to security and authentication, and identity and access management
- lack of adoption of the Australian Government Architecture because it is not mandatory.
There are existing ICT governance bodies (such as the Secretaries’ Committee on ICT (SCICT), Chief Information Officers Committee, Business Process Transformation Committee and Chief Information Officers Forum), but their effectiveness is limited and they have little influence over agencies. Only seven agencies made any comment on the SCICT, while eight made comments on other forums. The SCICT has never had a mandate and defined role from Ministers. Overall, there is no strong line of sight between agency ICT activity and whole-of-government priorities. Furthermore, the current governance arrangements allow agencies to opt into whole-of-government ICT initiatives on a self-approved basis as though they were independent entities who did not rely on funding by the Australian taxpayer.
There is also too much variation in the degree and quality of interaction between policy formulation and implementation. There is a broad spectrum of practice, ranging from early and ongoing consultation between policy makers and policy implementers to no consultation and announcements made in the press, which come as a surprise to those responsible for delivery of the policy. This issue is exacerbated by the increasing complexity of some policies. In conjunction with the decision to adopt the Gateway Review Gate 0 (which includes the planning for benefits realisation) in particular circumstances rather than as a standard gate, there are real downstream implications and risks for policy implementation from poorly considered policy design.
The totality of these issues ultimately hinders the ability of the Government to provide efficient and effective joined-up ICT-enabled services to citizens and businesses.
AGIMO has produced various tools and standards to help agencies use ICT to improve administration and service delivery, and to achieve some level of consistency across the Government. Despite this, there are variable views on the business relevance and attractiveness of these AGIMO frameworks, and some AGIMO tools and standards are not very well known and hence not widely used. AGIMO’s remit is weak as a consequence of a lack of a clear mandate.
Finally, the mechanisms for detecting issues and emerging trends and assessing whether they should be dealt with at an agency or whole-of-government level are too weak. This leads to ad hoc, reactive and siloed responses, as demonstrated by the following examples:
- The emergence of different client computing architectures will lead to increased variability unless coordinated action is taken.
- Duplication, the lack of common standards and sharing of best practice in the use of geospatial data by the Government are issues which have recently begun to be addressed on a whole-of-government basis, but it should have been resolved much earlier before agencies began to pursue independent courses of action.
- There is an emerging issue around data centres and their long-term viability in terms of space, weight, cooling, power, the sustainability agenda, and current and anticipated availability requirements.
- There is no whole-of-government oversight of developments in the ICT industry, including activities such as mergers and acquisitions, and the capacity of certain suppliers to fulfil contract requirements.
There is also evidence of pockets of best practices and common projects within the Government, but there are weak mechanisms in place to share and leverage these.
4.2 Agency governance mechanisms are weak in respect of focus on ICT efficiency and understanding of organisational capability to commission, manage and realise benefits from ICT-enabled projects
The review found that there is no formal common method of assessing agency capability to commission, manage and realise benefits from ICT-enabled projects. Although explicitly asked to comment on this, most agencies answered questions around capability in terms of ICT capability, rather than organisational capability. This holistic capability requires the right leadership, skills, processes and systems within and across the business and ICT domains.
Measurement of project benefits realisation by agencies is also weak, despite evidence that many of them have adopted formalised project management methodologies. Of the 193agency projects detailed, only 5% of projects reported actual measurement of benefits and compared anticipated benefits with actual benefits realised.
Overall, there appears to be a lack of understanding of the trade-offs between commercial off-the-shelf (COTS) solutions and customisation. For example, 17 agencies reported customising their FMIS by more than 10%. Similar customisation was recorded for HR systems. This finding is consistent with views from outside the Government, where 24organisations indicated that there were no specific inhibitors to using COTS solutions without customisation. Another 10 commented that there was often unnecessary excessive customisation by agencies leading to increased costs and the erosion of the inherent benefits of using COTS products.
There are also downstream implications for machinery of government (MoG) changes. For example, a decision by one agency to go down a customised route can create a future liability for another agency when MoG changes result in whole or part of an agency being integrated with a different one. This can result in unnecessary complexity and significant costs when integrating disparate ICT systems. One agency identified that its ICT costs related to the recent MoG changes are in the order of $30 million to integrate systems and infrastructure.
4.3 The business as usual ICT funding in agencies is not subject to sufficient challenge and scrutiny
The review found that very few agencies actively and regularly measure the efficiency of their ICT spending. It is also evident that there are no common metrics for measuring ICT efficiency or the effectiveness of an agency’s ICT spend. For example, a third of agencies did not provide any information on efficiency or effectiveness, and most did not articulate plans to address or improve efficiency. This means that it is very difficult to obtain consistent and meaningful data for management purposes.
There is evidence of the under-utilisation of software and hardware assets. For example, there are 1.6 desktops (including laptops) for each APS staff member. Industry has also proposed a number of actions that could improve asset utilisation, such as aggregating bandwidth and network infrastructure, consolidating data centres, and rationalising business systems and applications.
Within BAU, there are significant variations in costs across agencies. For example:
- desktop costs ranged from $1,000 to $3,500 per desktop
- finance systems costs per transaction ranged from 10 cents to more than $30
- HR systems cost per employee supported ranged from $10 to $500
- there are 611 reported websites (excluding Defence), with costs ranging from $10,000 to $1.5 million per website.
The average split between BAU and project spend is also too heavily skewed towards the former (77:23%), with little variation in the average depending on the size of the agency. Relative to the proportion of spend BAU represents, the level of scrutiny of this spend is too low.
Agencies stated that ICT is fundamental to their business, yet there is a general lack of strategic planning for ICT workforce capability at both the agency and whole-of-government level. Only four agencies have ICT workforce plans, with a further two planning to formalise plans.
There is also a lack of professional career structures for ICT professionals in the APS, making it more difficult for the Government to recruit and retain ICT talent. Training that is provided is based only on specific agency needs.
Existing AGIMO initiatives such as the ICT Apprenticeship and Cadetship programs are useful, but not sufficient to address the skills issue.
Agency ICT activity is Canberra-centric, with 79% of ICT staff located in the ACT. During meetings with agencies, three agencies noted that they had consciously moved ICT positions to other locations throughout Australia in response to the skills shortage. One agency noted that this approach had, to some extent, buffered them from the risks inherent in the skills shortage.
Together, these factors have contributed to the ICT skills shortage in the APS, with agencies reporting total ICT staff shortages in the order of 1,000 (mainly highly skilled). There is also a heavy reliance on contractors, who represent 23% of the total current ICT workforce. Furthermore, the decentralised way in which ICT projects are managed means that agencies compete in the market for ICT staff, which affects contractor rates. On average, ICT contractors cost around $186,000 per annum, which is $94,000 per annum more than the average loaded cost for ICT staff in FMA Act agencies.
The lack of training for Senior Responsible Officials in terms of their roles and responsibilities is further evidence of the disconnect between the stated importance of ICT and actions in relation to ICT. Moreover, organisational leaders are not as well equipped as they need to be on how to harness the potential benefits of ICT.
Finally, in relation to security clearances there is evidence that best practice in transferring security clearances (below the ‘Secret’ or ‘Highly Protected’ level) between agencies is not widely adopted, leading to unnecessary delays and costs. In addition, it impacts on the responsiveness and flexibility of agencies. For example, agencies need to wait to utilise skills and this can have subsequent effects on project delivery. I also noticed that the verification of my UK security clearance by Finance happened a lot faster than the reported 6 weeks of one agency to accept the security clearance granted by another agency. It seems strange to me that an international clearance would take less time than an agency-to-agency clearance.
These points raise some questions about clearance processes:
- Is the process consistent and well-designed? The variation in performance raises doubt on that score.
- Is the process supported by standard, well-designed, interoperable tools?
- Could some form of shared, central process – drawing on current facilities and systems where they exist – deliver a more satisfactory outcome without compromising agency responsibilities and needs?
4.5 There is no whole-of-government strategic plan for data centres. In the absence of such a plan, the government will be forced into a series of ad hoc investments which will, in total, cost significantly more than a coordinated approach
From the agency bilateral meetings, subsequent data gathering and industry responses, it is clear that the Government is facing significant issues with respect to agency data centre arrangements.
The key issues are:
- Power. The ACT has only a single power grid29 which poses a risk given the ever increasing demand in power.
- Capacity. Eighteen data centres of the largest agencies based in the ACT have identified that they have 10 m2 or less available capacity for growth. With industry benchmarks indicating that growth in data centre floor space is 5–10% a year, there is emerging pressure for additional data centre capacity.
- Inadequate facilities. Lack of investment has contributed to existing facilities not meeting the demands of modern technology in the key areas of floor loading, heating, ventilation and air-conditioning, and availability requirements. A significant number of agencies indicated that they would need data centres over the next 10–15 years with higher levels of availability than their current facilities.
The lack of a whole-of-government data centre strategy has led to agencies independently establishing or sourcing data centre facilities without considering the economies of scale that can be achieved from having a whole-of-government perspective. A few agencies are now actively planning the requirements for replacement of some of their existing facilities. Based on high-level modelling work, the review considers that there is a prima facie case that a whole-of-government approach will cost in the order of $1 billion less over a 15-year period than allowing each agency to handle the replacement of their existing data centres on an autonomous basis.
Additional benefits of taking a whole-of-government approach to data centres are:
- Geographic location benefits can be realised. Cooler climates locations have the potential to reduce cooling costs.
- Reduction in communications infrastructure. Fewer physical data centres require less physical communications infrastructure.
- Disaster recovery. It will be easier to accommodate disaster recovery at a whole-of-government level rather than each agency dealing with the issue autonomously.
The review found evidence to support the finding that the government ICT marketplace is neither efficient nor effective. This evidence included:
- high transaction costs due to overly onerous procurement processes and slow decision making
- barriers to entry, particularly for small and medium enterprises
- too much use of prescriptive statements of requirement, which limits the range of solutions proposed by vendors, including commercial/government off-the-shelf (COTS/GOTS) solutions
- failure to engage sufficiently with industry before the start of the formal competitive process
- the position taken by agencies on intellectual property is seen as a deterrent by suppliers
- the Government’s liability policy is not consistently implemented, which results in unlimited liability or caps on liability that are set at amounts that are so high they are essentially the same as unlimited liability
- the Government is not making use of its collective buying power for commodity products, leading to multiple tenders for the same product and agencies buying these products at different prices
- the Government’s $80,000 threshold for open tender
- insufficient use of selective sourcing to reduce dependency on large prime contractors and increase the likelihood of ‘best-in-class’ solutions.
4.7 There is a significant disconnect between the Government’s overall sustainability agenda and its ability to understand and manage energy costs and the carbon footprint of its ICT estate
There is a significant disconnect between the Government’s overall sustainability agenda and its ability to understand and manage its energy costs and the footprint of its ICT estate. This is evidenced by the lack of data agencies were able to provide in relation to energy usage and costs. For example, of the 72 agencies that were requested to answer the energy questions, 35 provided no data on energy, and a further seven provided inconsistent data. Only 23 agencies indicated that they had an energy plan, while 49 agencies advised that they did not have a plan in place. Of the 23 agencies that reported having an energy plan in place, 13 indicated that their energy plan included an ICT energy component; however, only six of those agencies were able to provide meaningful data in response to the questions asked in the survey.
- There are two feeds from New South Wales to the single ACT electricity grid. One feed provides 85% of electricity to the ACT and the other 15% . The smaller 15% feed is only to the immediate Fyshwick area and is insufficient to support agency needs. Agencies with data centres in Canberra are therefore susceptible to the single ACT power feed.
Contact for information on this page: email@example.com