Government and IT - "a recipe for rip-offs": time for a new approach: Further Report, With the Government Response to the Committee's Twelfth Report of Session 2010-12 - Public Administration Committee Contents

Appendix 3: Comments by Dr Mark Thompson, University of Cambridge

Introduction and overall comments

The government is to be commended on its constructive and proactive Response to the PASC Report. The Response itself is illustrative of how far the government has moved from a situation of rounded criticism of its stovepipe, proprietary approach to technology and suppliers, of which many examples were cited in the Report, to a discernible 'open by default' position based on very different conceptual underpinnings. Positive observations in relation to the many positive initiatives set out within the Response are set out in Part 1.

And yet, whilst welcoming these initiatives, it is possible to make two key criticisms of the document. First, as might be expected there are several instances where the Government does not go far enough in explaining the detail around its proposals. In many instances this would be a relatively trivial issue, but in the case of the current ICT-related reforms within government, the detail is serious because of the extremely significant cultural barriers to be overcome in embedding the transparent behaviours throughout the public sector on which its 'open by default' ICT reforms depend. An overview of those areas where further detail would be helpful to ensure that real progress is made, together with related suggestions, is set out in Part 2.

The second key criticism lies in the relatively light engagement throughout the Response with the underlying conceptual approach that will be required to make the Government IT Strategy, the Strategic Implementation Plan, and its responses to the Report a reality. At present the Response is a well thought-out, but only loosely related, set of relatively top-down initiatives that does not explain how these initiatives will work together to embed utility economics of open standards within government. For example, whilst 'transparency' and 'cost savings' may both be desirable outcomes in themselves, each will ultimately only be successful at scale if their relationship is understood. Whilst the Government may have 'point' success with many of the initiatives outlined in the Response, it will need to pay greater attention to its leadership on the 'why', as well as the 'how', if these are truly to stick. This observation is explained in more detail in Part 3.

For each point PASC should expect the response should be more specific about a requirement for individual departments to have specific implementation plans, otherwise the traction achieved will be patchy and inconsistent, a common theme in government IT. Early indications show that the Cabinet Office level commitment appears to gaining traction in embedding a manifestly different approach to IT delivery, but at a departmental level, there is limited senior sponsorship for change to happen, or understanding about how to make it happen. PASC should ask for a commitment to seeing these initiatives to be measured and reported against at departmental level with much greater level of transparency and evidence of supporting sanctions against non-compliant departments.

There is a major gap in the plan to cascade these intentions down through into the main departments, particularly DWP, MoD and HMRC who represent the greatest challenge in changing their oligopolistic behaviour, but have the largest spend on IT in government. PASC should ask for specific detail for these departments as they represent the most intransigent groups, with the largest vested interests.

The government response includes SMEs in the debate which is a step forward, but the steps proposed are, at best, cursory. For example the appointment of just one Crown Representative for all SMEs demonstrates good intent but poor execution. Also SMEs are strongly encouraged NOT to contact their Crown Representative and he refuses to publish his contact details, conveniently reinforcing Government's traditional argument that there are just too many SMEs to deal with!

Part One: Positive Observations

The response from Government suggests a positive first step in the right direction, and the intentions at Cabinet Office level would seem to be genuine held. The response has a number of sensible first step suggestions to initiate the process,.

Part Two: Constructive suggestions

It is my view that the following areas do not yet provide sufficient detail to constitute a convincing, robust response:

  • Information, Benchmarking, and Transparency (paragraphs 15 & 20): The Government's response is principally to reference the SIP and associated metrics. However, this is not the same as an (external) NAO involvement as recommended by PASC. Although NAO is mentioned, there are no commitments, governance mechanisms, or timetable for this and will be hard to enforce. There is also little explanation of how the "contract library" will be analysed. Specifically, will contracts be separated out so as to allow aggregation of commodities across contracts and associated cost savings? A big opportunity for meaningful comparison has been missed - the Asset Register and Contracts Library should have been one and the same. Currently having separate systems means that the real understanding and insight that would have been possible with comparative analysis of asset and service provision across contracts has been lost. (The Asset Register was conceived and procured in haste without real insight, with the Cabinet Office initially asking the large Systems Integrators if they wished to bid for it!). The Government should ensure that proper benchmarking is conducted throughout the life of contracts, and in particular in respect of change controls to major contracts. The benchmarking should be independent, and not a false comparison between one expensive service and another (which has been past behaviour).
  • Available data (paragraph 25): This is a very weak and evasive response by Government. Government is making use of a tried and trusted 'get-out': "commercial confidentiality". Traditionally, CC has been invoked in order to withhold the very information that is likely to result in external questioning of its commercial judgment and competence - i.e. just the information it is usually most important to publish. This should not be allowed to stand as-is. There is anecdotal evidence reinforced by comments at the recent major supplier conference hosted by the Cabinet Office "New Ways of Working" that, particularly in the large departments, there will continue to be significant resistance to transparency through evasion and obfuscation. Additionally there is also no mention of any plans to initiate a process to disaggregate or re-compete contracts.
  • Suppliers (paragraph 30): Whilst there are a number of initiatives listed here, none of these addresses the request to "urgently commission an independent, external investigation". This has been ignored entirely and is possibly one of the most important points to come out of the PASC report. In addition, further details are required on how the Government proposes to address the oligopoly of large suppliers especially at departmental level. In respect of SMEs, the scope of the Mystery Shopper should be extended to cover the lifetime of contracts rather than just the procurement phase and the relationships between large prime contractors and their SME sub-contractors.
  • Replace legacy systems (paragraph 34): Paragraph 13 here should be challenged beyond a vague commitment to assess legacy systems near end-of-life against Government ICT Strategy principles. How would this be measured? Shouldn't there be a 'Major Legacy' list as per Major Contracts, with a more detailed migration plan for each department for replacement or disaggregation?. If these are 'assets to be sweated', then surely they are commodities with sunk costs, and should be paid for as such? Additionally, the caveat of "budgetary constraints" should not be a reason to continue with locked-in closed systems which may be cost effective initially but provide poor value for money over their lifetime.
  • Open engagement with SMEs (paragraph 42): The target of 25% of business with SMEs by the end of the Parliament seems a long way away. There should be an interim measure and tracking to ensure that this target is not missed and more pro-activity from particularly the major departments to make this a reality. In addition, the Government needs to address the Committee's concerns around Department's requiring SMEs to contract through large SIs. This is still the default position in a number of Departmental Action Plans for engagement with SMEs and is a poor substitute for direct SME engagement.
  • Breaking up large contracts (paragraph 46): The Committee noted in its report that the £100m cap was excessive, but this appears to be the main thrust of the Government response. Even contracts of £5m are beyond the reach of some SMEs. There are also indications that DWP is ignoring this cap, with possibly large contracts being awarded without proper competition on the Universal Credit programme.
  • Agile methodologies (paragraph 87): This commitment is welcome, but Agile has been mooted as being used on various current projects when it bears only a passing resemblance (some parts of Universal Credit are using a small number of Agile techniques). Also the main contractor has (anecdotally) very limited Agile experience, hence the reason for bringing in SME experts. The government's reasoning is perverse - attempting to use Agile but having to bring in experts from SMEs because their main contractor has limited competence. Surely the obvious approach would have been to have selected a competent Agile company in the first place? There also seems to be confusion within Government about when to use Agile and when to use Lean. Agile is about rapid iterative and incremental delivery offering innovation and flexibility; Lean is about standardisation. Whilst Government needs to focus on using Agile processes for technology development, it also needs to become a mature user of Lean for business processes in order to standardise (and commoditise) its IT systems. The two approaches are different halves of the same coin.
  • Security and Privacy (paragraph 99): Paragraph 44 of the Government's Response is insufficient here. 'Gold plating' security is a major, recognised problem that is primarily cultural, and there is little comfort here that this is recognised or likely to be addressed properly. Long term, the only way to drive improved behaviours here will be the introduction of transparency that reveals the true comparative cost of 'gold plating' vs utility alternatives. As such, this paragraph is very unsatisfactory.
  • Developing intelligent customer function (paragraphs 108 and 112): The Government has not taken on board the recommendation that it needs to recruit appropriate IT professionals now to act as an intelligent customer, and instead focuses on coordination and processes with the existing staff who have failed to date achieve this important role. It is difficult to see how this will achieve the same results.
  • Spread of skills (paragraph 115): It appears far from clear at this moment that the Government will succeed in commissioning such a course on technology policy, and this should be monitored closely by PASC. Educating senior public servants in ICT-enabled public service design is a critical way to move away from the practices described in the Report.
  • Accountability of SROs (paragraph 118): The Government's Response ignores the core recommendation in paragraph 118 that SROs should be held and remain accountable for decisions taken on their watch, and instead says that there is an expectation that delivery leads will be held to account. This significantly waters down the requirement for SROs to be held properly to account. This should be challenged.
  • Leadership (paragraph 124): It is difficult to stress sufficiently the importance of building greater ICT literacy and representation at senior policy and implementation levels, and the Government response seems to be non-committal on this point. Leadership in the delivery of public sector IT has been traditionally poor and it is surprising that there is such an ambivalent response. Technology-enabled, utility-based services will increasingly call into question traditional notions of government service delivery and demonstrate the progressive redundancy and waste of post-WWII public service designs. It is difficult to understand how Departmental boards will be able to take effectively informed strategic decisions on behalf of the taxpayer in ignorance of these fundamental game-changing developments. The prospect of Departmental boards without senior ICT representation is deeply concerning and PASC should push back hard.

Part Three: Need for greater conceptual leadership

Whilst each of the initiatives set out in the Response is welcomed, the document reads as a relatively disconnected list rather than an integrated strategy that shows understanding of how the various parts support the whole. The relationship between the initiatives discussed in the Response is as follows. Recent developments in technology allow previously integrated systems and technology to be separated into modular 'building blocks'. In turn, this allows some 'building blocks' to be standardised (using Lean) benchmarked, and aggregated across government as a commodity, driving down costs. This requires transparency of information and data, both to benchmark/standardise - as well as to 'shame' commercial decisions that continue to ignore these opportunities, which are comparatively much more expensive (hence the need to push back on 'commercial in confidence' practices).

There is a further, vital part of the picture. By drawing together demand for standard commodities around open standards (to which SMEs need access via a direct route), government is able to use its buying power to create a 'platform' (guaranteed demand) around which suppliers (including SMEs) will invest and innovate (using Agile). In time, legacy systems will become comparatively much more expensive, and government will not specify IT, but will instead specify outcomes; it will not matter what the technology (or the supplier) is, since it is standard and works together, cheaply.

The important insight in all this is that this is a self-sustaining dynamic that needs to be 'fired up' by government through defining open standards, supported with transparency of information and data; to a large degree, the rest will take care of itself. Thus some of the 'top down', relatively disconnected initiatives listed in the Response lack punch, since they do not build understanding of how this works, and therefore the purpose of the whole exercise. Examples include paragraph 22, where a top-down requirement will "encourage a trend towards smaller projects and contracts" (because of a top-down fiat rather than organic dynamic).

Similarly, paragraph 29 seeks to clarity Government's plan to act as a "single buyer", to harmonise costs across departments so that they can pay the same price for similar goods and services. However, the important insight is missed here; namely, that standardising commoditises good and services, as well as creating a platform for innovation around these, as well as flexibility, new ways of working, etc. This, ultimately more important point is missed altogether. The issue is that there is as yet little indication of the Government's understanding of the need to deliver on these commitments as a co-ordinated, interactive whole, without which success will be difficult to achieve.

previous page contents next page

© Parliamentary copyright 2012
Prepared 26 January 2012