installing new drivers: how to improve government's use of it

8/8/2019 Installing New Drivers: How to improve government's use of IT

1/44

INSTALLING NEW DRIVERS:How to improve governments use of IT

Michael Hallsworth, Gareth Nellis and Mike Brass


2/44

2 INSTITUTE FOR GOVERNMENT


3/44

Installing new drivers: How to improve governments use o IT 3

Contents

Acknowledgements 4

Executive Summary 5

Introduction 10

Governing IT: A very brie history 14

The current role o the centre 15

Assessing the current arrangements 16

Recommendations 27

Conclusion 39

Appendices 40-42

Glossary 40

IT ministers since 1999 41


4/44


Acknowledgements

We would like to thank Tom Gash and Patrick Dunleavy or their insightul comments onprevious versions o this report. In addition, we would like to thank all our interviewees andthose who attended our workshop on the 28th October 2009.


5/44


Executive summary

This report examines whether the centre o government (the Cabinet Oce, Treasury andNumber 10) is able to ensure that IT enables government to meet its objectives eectivelyand eciently.

As part o the Institute or Governments wider Reshaping the Centre project, the reportocuses on the current combination o organisations, processes, responsibilities and unding.It does not oer a comprehensive assessment o how government uses inormationtechnology in general.

The report suggests that there are two main reasons or the centres involvement in IT.

The centre is well placed to ensure that IT contributes to achieving governments1)overall strategic goals eectively. For example, the recent TransormationalGovernment agenda aimed to use technology to shape services around individuals,rather than departments.

Creating pan-government ambitions or IT in this way invites a central coordinatingrole to set goals, oversee progress, and resolve the incoherence between departmentsthat limits eectiveness. The centre has had undoubted successes in this coordinatingrole, and has made major progress since the Transormational Government strategywas launched. Nevertheless, many ambitions or IT-enabled public services remainunullled a decade on while the achievements to date have come at considerablecost.

The centre can make governments use o IT more ecient by identiying areas where2)components could be reused, duplication could be eliminated, and collaboration couldmake ull use o the states buying power. The Operational Eciency Programme(OEP) has estimated that 3.2 billion (20%) o public sector expenditure could becut by doing so.

Despite the savings already achieved in areas like shared services, the centre needs todo more: there is wide variation in IT expenditure between public sector bodies, whiledepartments are not taking ull advantage o the possible savings rom greater use oonline contacts with customers.

IT in government now aces considerable challenges scal pressures will demand greater

eciency (and may make IT projects appear as attractive options or spending cuts), whilescepticism relating to data security and IT-enabled business change projects needs to beovercome. Now, more than ever, government needs to demonstrate that technology canull its strategic purposes. In such a context, the centre has a crucial role to play. But does ithave the authority and capabilities needed to ull such a role?

This is not a new question - the centre has been involved in governing IT or the past ortyyears, and we briefy trace this history, concluding that we are currently in a period ocollegiality: with power dispersed between departments, the centre attempts to coordinatethrough persuasion and consensus rather than ormal mandates. Indeed, the CIO Council, themain central decision-making body, has little capacity or authority to enorce its decisions even i they could signicantly improve the eciency or eectiveness o government.

The centre does perorm many aspects o its restricted role well: the CIO Council, which drawstogether departments CIOs rom across government, has brought signicant benets in termso sharing best practice and identiying the crucial issues and agendas acing government.The diculties arise with implementation: as one interviewee put it, so everyone agrees in

Now, more than ever,government needsto demonstrate thattechnology can ulflits strategic purposes


6/44


principle but then what? In other words, the current arrangements are not sucient totackle the issues that are controversial, confict with departmental priorities, require initialunding, or whose benets do not immediately accrue to the actors involved.

Why is this? We identiy ve main barriers to greater coordination o IT:

1. Competing departmental priorities.

Individual departments have a range o competing priorities, which routinely eclipse cross-government initiatives. Political and nancial accountability resides within departments.Unortunately, the departments leadership may not ully appreciate how IT can contribute

to achieving wider strategic goals whether owing to a lack o interest or a ailure ocommunication by the CIO. Furthermore, departments oten have an inherent dislike oedicts rom the centre. Even where the benets are clear, departments may eel reluctant torelinquish control over risks in the name o intra-governmental cooperation.

2. Dependence on individual CIOs

The CIO council arrangement assigns signicant agency to CIOs, which may increase thevariability o departments engagement with cross-government agendas. This is because thepower o individual CIOs at any given moment is aected by multiple actors: the type owork carried out by the department, the structure o the departmental amily, the resourcesand ormal roles given to the CIO, and their personal relationships. Since some CIOs canstruggle or infuence in their department, the impact o the CIO Councils decisions can vary

between departments.3. The resources and reputation o the centre

The IT unction in the Cabinet Oce is directly responsible or a tiny raction o totalgovernment spending on IT. This lack o disbursement power creates a relationship odependence between the centre and departments, making it easier or departments tosimply ignore requests rom the centre. Many interviewees indicated that the centre otendoes not possess sucient inormation to analyse IT in government eectively. Furthermore,it may be that the centre could benet rom a higher concentration o IT expertise toanalyse and judge such data. Finally, observers lacked condence in the Cabinet Ocesdelivery capability to ull its role in overseeing IT initiatives. Interviewees were, however,positive about the personal abilities o the Government CIO in developing consensus andexerting sot power through personal relationships.

4. Lack o political integration

Ministers requently do not pay sucient attention to the IT dimension o policyannouncements. Coupled with the overriding importance attached to individual ministersinitiatives, this can lead to the wheel being reinvented across government. IT is oten seenas a political ater-thought because it enters the decision-making process too ar down-stream, when plans and deadlines are already airly denite. Furthermore, the e-governmentagenda has been split across departments and ministers, or combined with ministers other,unrelated duties.

5. Central expertise in procuring and managing IT projects does not always count.

Although the Oce o Government Commerce is considered to provide airly good adviceon government IT procurement, issues arise when trying to ensure departments act on itsdirection. Currently, the centre imposes relatively ew checks or a department wishing toinitiate IT procurement. In terms o contract management, there is evidence that OGCsinterventions may not have sucient impact to infuence a departments management


7/44


o a contracted IT project, no matter how fawed it may be. Despite recent attempts toaddress this issue, there is a case or increasing the level o sensible, discerning checksrom the centre.

RecommendationsWhat are the principles that should underpin an eective centre-department relationshipor IT? Government should take advantage o new technology, rather than allowing newtechnology to take advantage o it. The centre should maintain a healthy tension betweenthese two orces, whereby strategy inorms technology and technology strategy. The centre

should be able to make a strategic, evidence-based case or a particular course o action,which is crucial because departments are subject to a vast array o competing pressures.

Stronger, more strategic recommendations can still be ignored. I the centre is to act asthe guardian o wider government outcomes, then, like most guardians, it needs some realinfuence to guide and direct. We suggest our interrelated ways such infuence can beashioned and nurtured.

Discerning interventionCoordination rom the centre needs to be selective, allowing departments reedom toinnovate and use technology to achieve benets in their policy areas, while insisting onsavings where the evidence is compelling. Thus, the centre needs to understand where it canadd value by intervening. The IT stack may be a useul way o judging how to match the

actions o the centre with the IT issue at stake, although it is not intended to be prescriptive.

Specifc specialist use

(e.g. Meteorological Oce systems)

Policy-related technology

(e.g. ID cards, Connecting or Health)

Contested collaboration(e.g. reuse, website rationalisation)

Back-ofce systems

(e.g. shared services)

ICT inrastructure

(e.g. desktop systems)

Security

(e.g. codes o connection)

The technology stack Extent o Governance Possible approaches

application issues by the centre

across

government

Specialised Clear case or Allow greater variance

to department specialisation Support procurement

standards

Zone o contested Assess retrospectively

governance

Enorce greater

standardisation

Lead collaboration

Ubiquitous Clear case or procurement

in government standardisation Intervene prospectively

Figure 1: The IT Stack


8/44


Knowledge and expertise

As just noted, the centre needs robust inormation in order to intervene strategically andmake a convincing case when it does so. It needs to take innovative approaches to identiygaps in knowledge and nd ways o lling those gaps without placing excessive burdens ondepartments. The centre needs to attract a succession o skilled IT managers: working or aperiod in the centre o government should become a standard, even desirable, expectationor government IT proessionals.

Create eective relationships o respect

The way in which the centre conducts its business is crucial: departmental leadership notonly needs to trust the expertise o the centre, but also have respect or the way it conductsits business. To create this respect, the centre needs to ensure it shows sensitivity to whatthe department is trying to accomplish, thereby creating sot power based on attractionand reputation.

Authority

I the centre can intervene judiciously, provide high-quality data and guidance, and createrelationships o respect, then it should develop enough authority to implement an eectivecomply or explain model or departments. The goal would be to create an authoritativeprocess whereby there would be reputational damage or a department that did not complyor explain in response to the centres demands.

Initial changesThese new perceptions and behaviours would take time to develop. They could be promptedby some initial specic reorms, which would also underpin the centres actions with morehard power. The ways that such changes could work are outlined below, and are exploredin depth in the report. However, we wish to stress that our main concern is with howthecentre operates, rather than where certain unctions are placed. Another reorganisation ostructures is unlikely to bring benets unless it is ounded on the vision o the role o thecentre outlined above.

The centres lead on IT would be a CIO Unit headed by the Government CIO,1.supported by CIO council members, and overseen by a Minister o State witha coherent cross-government IT brie. The Unit would be created out o thecurrent Oce o the Government CIO in the Cabinet Oce. The Unit will be

responsible, with its Minister and OGC, or drawing up a robust IT strategyor the whole o government. This IT Strategy would have Cabinet approvaland would include a committed budget to und cross-cutting programmes.The CIO Unit would have enhanced expertise, inormation-gathering andguidance capabilities, but it would not have delivery capability. Rather, deliveryo cross-cutting programmes would be carried out by individual departments.Overall, the most eective location or the Unit may be the Corporate ServicesDirectorate o the Treasury, where it would work closely with OGC.

Such a CIO Unit would have two main ways o operating: proactive2.intervention and oversight. The IT strategy would set out the criteriaor matching the approach to the issue, based on the IT stack and the

strategys priorities.Proactive intervention would attempt to enorce government-wide standards3.in those areas where there is a strong case or cross-government coordination


9/44


and less justication or departmental particularism. In case o disputesbetween the centre and line department, the issue would be escalated to astrengthened Corporate Functions Board or resolution. The political line oescalation would be up to the PS(X) Cabinet Committee. The Government CIOwould have double lock responsibility or approving major IT procurements.

The oversight unction would carry out light-touch monitoring o major4.department-led IT projects, whether they be cross-governmental or specicto a departments policy remit. OGC would be able to initiate and publishits Gateway Reviews. In terms o accountability, the Government CIO wouldbrie the Cabinet Secretary on a departments IT perormance prior to theperormance reviews o permanent secretaries that take place every six months.

In terms o transparency, the CIO Unit would publish an annual update on5.progress towards the IT Strategy, including a report on how each departmenthas contributed to the strategy and the amount they have saved throughcooperating with cross-cutting agendas (and through their own initiatives).These reports would be ormally submitted to the CSR Review process,with the expectation that they would be taken into account in subsequentnegotiations. Oversight would also come rom a Parliamentary SelectCommittee on Major Projects, which would draw on evidence rom the CIOUnit, OGCs Major Projects Review Group, and the NAO.

In the longer term, there may be a case or a new Finance Ministry that combinesbudgeting, eciency and perormance management responsibilities, without oversight omacroeconomic and tax issues. Its responsibilities could be similar to the Canadian TreasuryBoard Secretariat. This ministry might combine the responsibilities o the CIO Council andOGC to allow a more coherent overview o all the aspects o government IT. The Institutesorthcoming Reshaping the Centre report considers this idea in more detail.

This report does not recommend that the centre extends its reach into all aspects ogovernment IT. Departments are usually the best judges o their own business interests.But there are clearly some areas where greater central coordination and challenge wouldensure improved eciency and eectiveness. Where the evidence is clear and the need ispressing, the centre needs real power to act switly and decisively. Our recommendations

would ensure that the centre has such power, while retaining the advantages o the currentsetup, such as a collegiate approach to strategy and strong personal relationships betweenCIOs. Given the need to meet complex challenges with constrained resources, the need or astrong yet smart centre has never been greater.


10/44


Introduction

Put simply, there are two main justications or the centres involvement in IT: rst, to ensurethat IT contributes to achieving governments overall strategic goals eectively; second, toco-ordinate the use o IT across government so ineciency, variation and duplication areminimised.1 The ollowing sections explain these roles and judge how well the centre hasullled them.

1. Achieving outcomesOver recent years, there have been impressive claims about the crucial role that IT can playin government, such as the assertion that policy is inspired by it, business change is delivered

by it, customer and corporate services are dependent on it, and democratic engagementis exploring it.2 Such claims refect the view, increasingly prevalent rom the mid-1990s,that the main purpose o Inormation Technology in government is to enable or deliverbetter outcomes or citizens. Creating pan-government ambitions or IT invites a centralcoordinating role to set goals and oversee progress. And, indeed, in 2005 the Cabinet Oceound itsel the custodian o the wide-ranging Transormational Government agenda, whichwas based on better using technology to deliver public services and policy outcomes thathave an impact on citizens daily lives.3

One o the main goals o Transormational Government was to use IT to reconstituteservices around citizens, rather than departments. A crucial rst step towards this goal wasto improve joining up through greater interoperability o systems and common standards.The need or joining up highlights an important role or the centre: to identiy and tackle

those areas where governments use o IT l imits its eectiveness. The centre is best placedto examine and standardise the patchwork o systems that have grown up across centralgovernment, operating to dierent standards and requently unable to talk to each other.Examples abound, but small instances are oten most telling, such as the case o the DIDemployee who has to have emails rom the Ministry o Deence orwarded to him throughthe Foreign Oce because o security conficts.

Given their range o policy and delivery responsibilities, line departments oten have limitedresources, expertise, inormation and incentives to address these problems which, whentaken together, signicantly impair the eective working o government. Thereore, there is arole or the centre to take actions that benet government as a whole. For example, the issueo emailing DID rom the MoD should be addressed by the proposed Public Sector Network,but unless the centre drives orward such cross-cutting projects they are likely to be derailedby collective action problems or example, no-one wanting to be the rst to commit to thenew structure.

How well has the centre worked with departments to achieve these desired outcomes? First,its important to recognise that the centres involvement in government IT is some orty

years old; this history is assessed in the next section. Second, there are many other actorsinvolved apart rom the centre-department relationship (notably, perormance o privatesector suppliers). Nevertheless, its clear that the verdict on the recent TransormationalGovernment agenda is mixed. On the one hand, the government can point towardssuccesses like the OECDs judgment that the Greening IT strategy is world-leading and theintroduction o a national digital imaging system or x-rays that stores 500 million imagesonline.4 Improved structures or liaising with suppliers have allowed HMRC to realise annual

1 For a denition o IT, and all other initialisms and acronyms used, please see Annex A.

2 HM Government (2005) Transormational Government: enabled by technology, p.3.

3 Ibid., p.5.

4 HM Government (2009) Transormational Government Annual Report 2008, p33, p.44; HM Treasury (2009) Operational EfciencyProgramme: back ofce operations and IT, p.43.

Why shouldthe centre ogovernment beinvolved with IT?


11/44


savings o over 50 million a year, while departments have made considerable progresstowards achieving shared services.5

On the other hand, these achievements have to be seen in terms o their cost and in termso the expectations that have been raised. In terms o cost, the Treasurys OperationalEciency Programme (OEP) estimates that total UK public sector spending on IT was16 billion in 2007-8, 4.6% o total departmental expenditure.6 Perhaps the best way ointerpreting this gure is to put it in an international context. John Suolk, the GovernmentCIO, has argued that that ITs 4.6% share o public sector spending should be compared to aglobal average o 5.9% and a Europe, Middle East and Arica private sector average o 7.7%.7However, data rom Gartner eatured in the OEP indicates that Britains per capita IT spendwas $606, considerably higher than the US ($525) and other major European economies.8The UK accounts or 22% o the EU public sector IT market (with 12% o its population),ollowed by France at 18% (12.8%) and Germany and 16% (16.4%).9 Overall, it is clear thatthe UK is one o the biggest spenders on government IT in relative terms. It should alsobe noted that the UK has a history o overspending on expensive IT-led business changeprojects: most recently, the Public Accounts Committee produced a scathing report on theNational Oender Management Inormation System, which had to be halted when costs ranto three times the original estimate.10

It is dicult to link expenditure to outcomes to make international comparisons, but itis noticeable that in 2008 the UK was ranked 10th in the United Nations e-Government

Readiness Index (compared to 4th in 2005), 25th in the UNs e-Participation Index (1st in2005) and 8th in the Economist Intelligence Units e-readiness rankings (4th in 2005).11 Arecent OECD report placed the UK third in Europe or the sophistication o its e-services and11th in terms o citizen takeup o such services.12 Overall, the OEP concludes that the evidencestrongly suggests that the UK public sectors IT spend is much more than other similarcountries and that the UK does not get a proportionate return rom this much higher spend.13

In terms o expectations raised, many o the ambitions or transormational governmenthave now been in place or some time, which raises the question o whether urther progresscould have been expected. Take, or example, the desire to use technology so citizens caneasily notiy government o a change o address. In 1999 there was a proposal rom theCabinet Oce, ollowing up the Modernising Government agenda, that by the end o 2001,people will be able to tell all the government departments they have dealings with o any

change o address on-line, in one easy step.14 A subsequent pilot, driven by the Oce othe e-Envoy, was discontinued ater poor take-up; tensions between departments, partnersand the Gateway process; and diculties integrating departmental databases.15 Five yearslater, Sir David Varneys rst recommendation in his 2006 Service Transormation Review

5 HM Government (2009), pp.31-2, p.43.

6 HM Treasury (2009a), p.44.

7 Gartner data, cited at http://johnsuolk.typepad.com/john-suolk---government-cio/2009/08/public-sector-productivity-vs-it-investment.html



10 Public Accounts Committee (2009) The National Oender Management Inormation System.

11 International benchmarks should be treated with caution. Their deciencies are examined at: http://www.iq.harvard.edu/blog/netgov/2009/05/one_rank_rule_benchmarking_egovernment_eu_un_brown_government20.html

12 OECD (2009) Government at a Glance, Chapter 10.

13 HM Treasury (2009a), p.60.14 At: http://archive.cabinetoce.gov.uk/servicerst/2000/panel/address_portal/index.htm. A ormal commitment to the single transaction

or change o address was made in the 1999 Modernizing Government repor t (http://www.archive.oicial-documents.co.uk/document/cm43/4310/4310-03.htm). The 2001 completion date should be taken as a representation o expectations, rather than a rm commitment.

15 SQW Consulting (2002) Evaluation o Invest to Save Budget Case Study Report: Project 2/20 - Change o Address. At: http://www.isb.gov.uk/hmt.isb.application.2/learners/case_studies/2-20%20Change%20o%20Address.doc

Achievements haveto be seen in termso their cost and theexpectations raised


12/44


was to set up a service that will allow citizens to inorm government once o their changein circumstances; initially this should cover bereavement, birth and change o address by2010.16 The Tell Us Once programme, launched in 2007, is currently piloting services ordeath and bereavement notications only (while change o address remains at businesscase stage). In April 2009, IDeA reported that it was envisaged that services or birth,bereavement and change o address would be in operation by 2011.17 Thus, as in 1999, thechange o address service remains an ambition.18

2. Ensuring efciency

The Treasurys Operational Eciency Programme (OEP) has estimated that 3.2 billion(20%) o public sector expenditure could be cut, although this should be treated as arough estimate.19 As the OEP acknowledges, the centre has a clear role in achieving sucheciencies by identiying areas where components could be reused, services could beshared, and collaboration could make ull use o the buying power o government. TheOEP points out that benchmarking IT costs between public sector bodies revealed a widevariation in IT expenditure... giving condence that signicant savings are possible. Instituteor Government analysis o central government departments resource accounts reinorcesthis point.20 In terms o procurement, the Oce or Government Commerce states that13.2bn o public sector IT is committed to external contracts, approximately 80% o totalIT expenditure. Although many contracts matched best practice in the public sector, theOEP noted that there was evidence that the public sector had been paying above the marketrate in certain areas. Again, variation between departments was seen, with the price paid orstandard laptops varying by over 300% in certain cases.21 As a result, the OEP estimates that1.6bn could be saved through extended collaborative procurement in IT by 2011-12.22

The centre has already managed to achieve eciencies through co-ordinating the use oIT. Indeed, the latest Transormational Government report points towards successes such asnearly 100m in cumulative savings by April 2009 rom the DWPs shared services work.23Nevertheless, it is clear that that more could be done. The National Audit Oce recentlypointed out that the DWP had not yet realised the substantial processing eciencies andcustomer service improvements that could be realised by exchanging inormation withcustomers online.24 There is a role or the centre to make the case or the large savingsthat could be made by improved use o IT. For example, o the 145 million contacts withthe DWP in 2008, only 340,000 (2.3%) took place online, despite the act that 51% o the

DWPs customers had broadband access by this point.25

Research rom the LSE estimatesthat annual savings o 430 million (20% o expenditure) could be made across Job CentrePlus as a whole through greater use o digitisation and online contacts. 26

16 Sir David Varney (2006) Service Transormation: a better deal or citizens and business, a better deal or the taxpayer, p.5, p.83.

17 IDeA (2009) Tell Us Once: Case Study, p.2. At: http://www.idea.gov.uk/idk/aio/10012779.

18 Our thanks to Jerry Fishenden and Public Strategist blog or inormation relating to this example.

19 HMT (2009), p.48. It should be noted that some Government CIOs have raised concerns about the quality and source o data used inthe OEP. See Minutes o Joint CIO, CTO and Local CIO council, 2-3 July 2009. At: http://www.cabinetoce.gov.uk/media/267684/cio_minutes0907.pd.

20 In order to guard against major variations owing to the IT investment cycle, we took the mean expenditure over two nancial years.

21 HM Treasury (2009a), pp.55-6.

22 HM Treasury (2009b) Operational Efciency Programme: collaborative procurement, p.26.

23 HM Government (2009) Transormational Government Annual Report 2008, p.

24 National Audit Oce (2009) Department or Work and Pensions: Communicating with customers, p.6.

25 Presentation by Proessor Patrick Dunleavy at the London School o Economics and Political Science, October 6th 2009. At: http://www.lse.ac.uk/collections/EDSInnovationResearchProgramme/pd/06-10-09_dunleavy.pd

26 Ibid.

The centre hasa clear rolein achievingefciencies byreducing variationsand encouragingreuse


13/44


The current challengeIt may be true that, as Ian Watmore has argued, IT in government is as dicult as it gets.Government does things in IT which are more complicated than anywhere in the privatesector.27 Regardless o the UKs past perormance, IT in government now aces considerablechallenges scal pressures will demand greater eciency (and may make IT projects mayappear to be attractive options or spending cuts), while scepticism relating to data securityand IT-enabled business change projects needs to be overcome. Now, more than ever,government needs to demonstrate that technology can ull its strategic purposes.

In such a context, the centre has a crucial role to play. But it aces major challenges. Asthe Digital Britain report has noted, departments o state exist largely as silos to addressa particular set o relatively stable analogue-era interests and issues. They are not welladapted to the fuid, iterative nature o the digital world where technology interacts withand re-shapes the underlying business process.28 The OEP recognises that the centre andChie Inormation Ocers should be empowered to address duplication and variationacross departments, but most o its recommendations do not tackle the underlying barriersthey ace. The Treasury Select Committee concurs, noting that the OEPs eciencies willrequire considerable cooperation between the departments, while questioning whether thenecessary structures are in place to acilitate such co-operation.29

This report examines whether the current combination o structures, processes,responsibilities and unding allow the centre to ensure that IT enables government to meet

its objectives eectively and eciently. Is a dierent role or the centre o governmentneeded to tackle the major challenges that have arisen? How can reedom to innovate bebalanced with compulsion to standardise? To address these questions, this report outlinesthe history o IT in government; examines how IT is currently managed in government;identies issues arising rom these arrangements; and oers recommendations to addressthese issues. Its ndings are based on publicly available inormation and interviews with keycivil servants, academics and consultants with experience o IT in the public sector.

27 Say, M. (2005) The Inormation Man, Government Computing, 19:2, p. 16.

28 DCMS and BIS (2009) Digital Britain: Final Report, p.208.

29 House o Commons Treasury Committee (2009) Evaluating the Efciency Programme, p.3.

IT in governmentis as difcult as itgets. Governmentdoes things in ITwhich are morecomplicated thananywhere in theprivate sector.

Ian Watmore


14/44


Governing IT: A very brie history

The history o governing IT is crucial or understanding the current situation, since it shows thatcontrol has repeatedly oscillated between poles o centralisation and decentralisation. Since the1968 Fulton Report, government attempts to manage IT can be seen to all into our periods: 30

Pre 1982: CCA directionThe Central Computer Agency (CCA) was created in 1972 within the Civil ServiceDepartment as a single, strong technically competent agency to purchase computerequipment and services and to develop and co-ordinate their use in Central Government.31The CCA brought a unied approach to procurement and had some considerable power:

at one stage, the CCA had direct ownership o 80 per cent o government computers.Nevertheless, questions remained over its success in improving departments managemento large-scale IT projects: the Public Accounts Committee ound that CCAs advice on thesematters was oten ignored.32 Opinions are divided over how eectively the CCA challengeddepartmentalism in this period.

1982 1995: DecentralisationAter the Civil Service Department was abolished in 1981, the CCTAs remaining taskswere transerred to the Treasury. From around 1982 onwards, the CCTAs power diminishedconsiderably,33 in line with the general trend o giving control over corporate services toindividual departments and agencies. By 1984, the CCTA was giving departments clearcontrol over the choice o systems.34 A year later, CCTA lost nancial control over computer

procurement to the Treasurys expenditure divisions, who lacked expertise to controlcomputer expenditure eectively.35

1995 2004: Re-centralisationIn 1995, a small Central Inormation Technology Unit (CITU) was established within theCabinet Oce to take a strategic view o the way IT is used across government.36 Thisambition was urthered in 1999 through the Modernising Government white paper andthe creation o the Oce o the e-Envoy (OeE), whose head reported directly to the PrimeMinister. The OeE grew rapidly, boasting a sta o 200 people and operating costs o 50million by 2001. Thereater, concerns about underperormance and institutional overreachled to it being scaled down and rebranded as the e-Government Unit, which retainedresponsibility or IT systems across government.

2005 to date: CollegialityThe Transormational Government agenda o 2005 signalled a shit to a more consensualand less directive model, and introduced a new ocus on developing IT proessionalism. Theoundation or this approach was the placing o Chie Inormation Ocers (CIOs) into allgovernment departments and agencies, drawn together through a CIO Council. The Council isheaded by a Government CIO in the Cabinet Oce, supported by a team o approximately50. The ambitions o Transormational Government have proved dicult to ull, owing tothe programmes lack o permanent central unding and the sheer scale o the challenge.

30 This section draws on Dunleavy, et al. (2008) Digital Era Governance: IT corporations, the state and e-government, Oxord University Press(2nd edition), pp.48-9 and Organ (2003) The Co-ordination o e-Government in Historical Context, Public Policy and Administration 18:2,21-26.

31 Subcommittee on Science and Technology (1972) The prospects or the UK computer industry in the 1970s, HC 473.

32 Public Accounts Committee (1979) Procurement o Government Computers, HC 463, Vol 1. Cited in Organ (2003), p.25.

33 The CCA was renamed the Central Computer and Telecommunications Agency (CCTA) in 1984 as it took on respons ibilityor Telecommunications.

34 CCTA (1984) IT Series Number 8, p.iii. Cited in Organ (2003), p.26.

35 Margetts (1999) Inormation technology in government: Britain and America, p.45.

36 Select Committee on Science and Technology (1996) UK Computer Industry, HC 137, Vol. 1.


15/44


The current role o the centre

Cabinet OfceDirection or government IT strategy is provided by the CIO Council, which is charged with creatingand delivering a government-wide CIO agenda to support the transormation o government, aswell as improving capability in IT-enabled business change.37 The Council brings together the CIOso central government departments, the devolved administrations and some local authorities andagencies. The Councils ederal structure means that it exists to acilitate, not mandate, changes.The Chie Technology Ocer (CTO) Council replicates the structure o the CIO Council, but ocusesmore specically on technological issues that support the achievement o ITs strategic purposes(such as interoperability and a cross-Government Enterprise Architecture).

The Cabinet Oces Oce o the Government CIO also takes the lead on specicinitiatives, such as the Public Sector Network, G-Cloud, shared services and developing the ITproession.38 The Cabinet Oces Services Transormation work also involves IT signicantly.39However, given that its resources are very limited, the Cabinet Oce depends on thoseprovided by other departments. Thereore, a large part o its role is selling the benets o

joint initiatives and coordinating the partners that sign up.

TreasuryThe Oce o Government Commerce is the main Treasury actor involved in IT. It has severalrelevant competences (not all specic to IT): Gateway Reviews or major projects, whichare mandatory in central civil government;40 a Major Projects Review Group that scrutinisesexpensive, high-risk projects;41 e-Auctions; an ICT Category Collaborative Procurementteam, which aims to nd savings within a pan-government strategy; and the ProcurementCapability Reviews.42 There are also two joint OGC-CIO council initiatives that aectprocurement. First, the Strategic Supplier Board, which brings together the leading IT industryCEOs and government IT managers. Its Tiger Teams work to improve procurement processes,quality and reliability o projects, speed o procurement and cost reduction and reuse.43Second, the Common Assessment Framework, which allows IT projects to be assessed in astandardised way, and data aggregated to provide a picture o each suppliers strengths andweaknesses. The ramework includes a 360 degree element that allows supplier accountdirectors to provide a rating o their government clients capability and perormance.44

The Prime Ministers OfceThe Prime Ministers Oce concerns itsel mainly with the high-level strategic purposes o

inormation technology, as evinced in the Transormational Government strategy. Althoughthe e-Envoy reported directly to the Prime Minister, the CIO Council and Cabinet Oce arenow the main actors in maintaining oversight o IT in government.45

37 http://www.cabinetoce.gov.uk/cio/about_the_council/the_cio_council.aspx

38 The Public Sector Network will create a Virtual Private Network or the entire public sector, and a common marketplace where governmentorganisations buy voice and data networks. It aims to both create savings and create common technical and service standards. The G-Cloudeectively aims to provide government applications online so they can be accessed easily across the public sector; this wi ll particularlyhelp the shared services agenda.

39 See, or example, the recent papers on channel strategy guidance - http://www.cabinetoce.gov.uk/public_service_reorm/contact_council/resources/channel_strategy.aspx

40 http://www.ogc.gov.uk/what_is_ogc_gateway_review.asp

41 http://www.ogc.gov.uk/programmes___projects_major_projects_review_group.asp

42 http://www.ogc.gov.uk/ogc_-_transorming_government_procurement_procurement_capability_reviews.asp. In October 2009, itwas announced that uture capability reviews will be conducted by departments themselves. This is discussed below. At: http://www.civilservicenetwork.com/latest-news/news-article/newsarticle/ogc-takes-back-seat/

43 http://www.ogc.gov.uk/gps_digest_the_supply_transormation_programme_.asp; http://www.intellectuk.org/content/view/813/47/

44 The Common Assessment Framework is not currently public. This inormation has been gathered rom a variety o sources: http://www.intellectuk.org/content/view/823/47/; HM Government (2009) Transormational Government Annual Report 2008, p.41.

45 The Prime Minister does receive quarterly updates on 42 mission critical projects rom the Oce o Government Commerce, many owhich are IT-enabled business change. Nigel Smi th, Oral Evidence to the Public Accounts Committee, 14th January 2009

The CIO Councilexists to acilitate,not mandate,changes

The Ofce oGovernmentCommerce is the

main Treasury actorinvolved in IT


16/44


Assessing the current arrangements

Extent o central coordinationBroadly speaking, the Treasury and Cabinet Oce have dierent (though complementary)approaches to co-ordinating IT. The Treasury sees departments as airly rational economicentities, which will do the things that are cheapest or them or provide best value.Economical solutions will thereore arise with little central direction because they aresensible and benet everyone. Similarly, departments will invest money where there isthe prospect o gains, so there is little need or incubation unds or inter-departmentalinitiatives. Although the Treasury may take a rm line on budget negotiations, it takes arelatively hands-o approach to achieving eciency in IT across government.

Nevertheless, the Treasury has been active in some areas. For example, the Oce oGovernment Commerce has become increasingly active in mapping IT procurement spendand achieving eciencies rom collaborative procurement (although less so in the widerpublic sector). Similarly, the Treasurys Operational Eciency Programme made somespecic proposals to strengthen oversight and coordination o IT. However, many o itsrecommendations are vague and merely refect the existing power structures saying thatCIOs and OGC Collaborative Category Boards should be empowered to address duplicationand large cost variations immediately invites the question, How?46 Without specic, bindingproposals, it is likely that the rate o progress will not match the OEPs ambitions, andentrenched problems will endure.

The Cabinet Oce model is collegial and cooperative; it depends on inormal relationships

rather than hard power. Implementation o CIO Council decisions is essentially peer-enorced and relies on CIOs securing agreement rom their departmental colleagues.47The CIO council has little capacity or authority to enorce its will, even i its decisionscould signicantly improve the eciency or eectiveness o government. One intervieweeclaimed that even letters rom the Cabinet Secretary are sometimes insucient to ensuredepartments compliance on IT issues, even in relatively trivial matters. O course, theCabinet Secretary, in his role as Head o the Civil Service, may decide to use his power overpay and promotions to bring ocials on side but this is a rather blunt instrument that canonly be applied retrospectively.

The current arrangements mean that the central IT unction in the Cabinet Oce has nohard power to make things happen. As one senior Cabinet Oce employee explained, Thedepartment always has more troops on the ground in the argument I the departmentwants to argue, you are lost. This is a particular diculty or the Cabinet Ocesmanagement o government-wide IT initiatives (such as the Public Service Network orthe G-Cloud), since it makes them vulnerable to collective action problems. While mostdepartments understand the public good benets that such projects bring, there isreluctance to be the rst to make a commitment and shoulder the initial risk. One potentialsolution is, as a Cabinet Oce interviewee put it, to nd a way in which departments whotake things orward are rewarded, and it doesnt become a drain on the rest o their business.An alternative is to ensure that there is sucient power and resource to drive desirable pan-government projects orward rom the centre.

There are two important things to be said in deence o the Cabinet Oce. First, there isthe wider historical context. Ater the relative ailure o the Oce o the e-Envoy, it was

understandable or the pendulum to swing back towards decentralisation and maintainingthe optimum tension between the two poles is notoriously dicult. Second, many46 Ibid., p.67.

47 It is likely that the orthcoming Government IT Strategy will introduce strengthened governance arrangements or each o itsthematic areas.

The Treasury takesa relatively hands-o approach toachieving efciencyin IT acrossgovernment

The Cabinet Ofcedepends on inormalrelationships

The central ITunction has no

hard power to makethings happen

The Cabinet Ofcedoes a good jobwithin the current

model...


17/44


interviewees stressed that the Cabinet Oce does a good job within the current model. TheCIO Council has brought signicant benets in terms o sharing best practice and developingstrategic direction. In one respondents words, that group o CIOs [has] got to the stagewhere they work together and some will take the lead on various issues that are comingthrough. Similarly, there was praise that the Cabinet Oce has identied the crucial issuesand agendas acing government: in other words, its strategy is right.

This highlights the crucial issue at stake: the Cabinet Oce model is suited to developingstrategy and improving links between CIOs, but it alters when strategy turns to action. Asone Cabinet Oce interviewee put it: the Cabinet Oce model has worked or strategicplanning; in terms o implementation, it has creaked a bit. So everyone agrees in principle -but then what? The Digital Britain report concurs:

The CIO and his Council have been signicant drivers o Digital Government PhaseTwo reorms. But there are limits to the pace at which change can be driven byguidance, exhortation and discussion. Government as a whole cannot aord anexcess o departmental particularism to rustrate the necessary drive to commonsystems and procurement.48

In other words, the current approaches, in combination, are not sucient to tackle the issuesthat are controversial, confict with departmental priorities, require initial unding, or whosebenets do not immediately accrue to the actors involved.49 These are not new problems:over the past 30 years the centre has requently ound itsel sidelined, ignored, and not seenas a signicant player in IT.50 It aces enormous diculties in creating the organisational,structural, cultural and technical changes necessary or joined-up e-government.51 Theollowing sections explore the current challenges.

1. Competing departmental prioritiesOne o the main barriers to coordinating IT across government is a lack o enthusiasm orcooperation rom individual departments. The most signicant causes are:

Other priorities

Individual departments priorities routinely eclipse cross-government initiatives. Politicaland nancial accountability resides within departments; permanent secretaries still embodypower. Introducing changes rom the centre is dicult unless a clear business case

providing clear evidence o why taking action will benet a department can be made.Even then, IT may lose out in the trade-os that are prevalent in government. A permanentsecretary may decide that, in the ace o limited resources, achieving a particular policygoal is more important than participating in a pan-government initiative with immediatecosts and delayed payos. Political priorities are perhaps the most salient actor, and areconsidered in more detail below.

Dislike o central interventions

Many interviewees suggested that line departments have an inherent dislike o edicts romthe centre.As one noted, i you start mandating, then the bright opposers think up reasonswhy the initiative doesnt apply to them and argue the toss a lot o energy is spent on thepeople who wont agree you have to drag them along. This resistance is ounded on the


49 These issues are not specic to IT and are also dealt with in the joining up section o the wider report.

50 See, or example, Public Accounts Committee (1979) Procurement o Government Computers, HC 463, Vol 1; National Audit Oce (1984)Administrative Computing in Government Departments, HC 259;National Audit Oce (1999) Government on the Web HC 87.

51 Organ (2003), p.31.

...but the modelalters when turning

strategy into action


18/44


perception that the centre does not understand the departments competing priorities (seeabove), does not suer the same pressures, and is not exposed to the same consequences orailure. There may be some truth to these perceptions, but it is perectly possible to see thecentre as an enabling orce that helps departments meet common challenges such as tightspending limits (i the centre does indeed have these capabilities).

Perceptions o increased risk

For departments, intra-governmental cooperation may seem rie with risk. They arerelinquishing control over risks to another department, which may have dierent priorities.

There may be ew or no ways o holding this department to account, unlike i the work wasconducted within their own department or the bounds o a contract. The xed cost o acontract may seem to be a less risky option, although it requently proves otherwise.52

Intra-departmental co-ordination

Problems with intra-departmental co-ordination exist, most commonly in those departmentswhose agencies and NDPBs are numerous and powerul. Some departments are highlyederated structures. The core department may thus encounter agency resistance to the ITinitiatives it attempts to implement on behal o the centre: an obvious objection is that suchinitiatives may detract rom agencies specic targets and unctions. In this sense, ederateddepartments oer a microcosm o the problems besetting inter-departmental coordination.

Appreciation o IT

The managers o departments may not ully appreciate the importance o IT in achievingwider strategic goals. As one interviewee argued: One o the main problems is thatPermanent Secretaries dont understand what the technology means or outcomes. Orrather, Permanent Secretaries might grasp the rationale but not really engage with it. Youreally need to get Permanent Secretaries to care about how IT can improve outcomes. AsDavid Clarke, President o the British Computer Society, has noted: It is unacceptable or amodern chie executive to be ignorant o the strategic use o IT but the same also goesor politicians or senior civil servants.53

Unsurprisingly, some management teams engage with IT more than others. Yet this is notwholly owing to their personal interests the competencies o Government CIOs alsomatter. Some interviewees argued that too many Whitehall CIOs retain a narrow ocus ontechnology, rather than on the business outcomes o their department. This is out o kilter

with the private sector, where the CIO is rst and oremost a business strategist who uses ITto support the corporate strategy. Some interviewees noted that this business awareness ogovernment CIOs has improved over recent years.

2. Dependence on individual CIOsThe CIO council assigns signicant agency to CIOs. The recent Operational EciencyProgramme recognised this state o aairs by addressing many o its recommendations atthe collective group o CIOs (as noted above). The diculty is that relying on individual CIOsincreases the variability o departments engagement with cross-government agendas. Thisis because the power o individual CIOs at any given moment (both within their departmentand in the CIO council) is aected by multiple actors:

52 It became expected practice [or private sector providers] to pitch prices or initially completed tranches o work relatively low, in thecondent expectation that later revisions and extensions would create negotiated contracts o between our and six times the initialcompeted contract price. Dunleavy et al. (2008), p.76.

53 David Clarke (2009) Tech policy must emerge rom the silo. At: http://resources.zdnet.co.uk/articles/comment/0,1000002985,39710789,00.htm


19/44


The type o work carried out by the department

CIOs are likely to play a signicant role in departments whose objectives and smoothunctioning are heavily dependent on inormation storage and processing (the most obviousexamples are the Department or Work and Pensions and Her Majestys Revenue and Customs).

The structure o the departmental amily

A ederated structure one with large, powerul agencies and a weak core may result inthe departmental CIO having a smaller budget and wielding less power.

Formal roles

Privileges, such as a seat on the departmental board, can give CIOs access to vitalinormation and provide an important orum to make the case or IT. It is noticeablethat in only two o the 18 main central government departments does the CIO sit on themanagement board.54

The resources the CIO can draw on

I a CIO only has only a small number o employees to create the departmental IT strategy,it can be more dicult to exert infuence in the department or over agencies. This is partly acapacity issue, but also refects the perception that power correlates with budgetary control.

Personal relationships

CIOs need to orm alliances to get the job done adequately and with proper resources. Doing

so relies on a CIOs personal characteristics, which obviously creates variability betweendepartments. Even good existing relationships are not secure, given that the internaldynamics and composition o senior management requently change.

3. Resources and reputation o the centreIn the current set-up, the centre needs to ull some crucial tasks tying together disparateactivities into a strategy, bridging dierent interests across government, providing expertadvice, and overseeing cross-government initiatives. As noted above, the centre is generallyconsidered to be strong on IT strategy; there are, however, some barriers that hinder itsullment o the other roles.

Resources

In a sign o the current systems decentralisation, the IT unction in the Cabinet Oce isdirectly responsible or 0.068 per cent o total government spending on IT.55 This lack oresources underpins all the points made in this section, but has some specic consequences:

First, it creates a relationship o dependence between the centre and departments. As notedabove, Cabinet Oce initiatives are mainly staed by volunteers rom line departments:one interviewee estimated that between 10-20 people work or the Cabinet Oce on ITinitiatives or every one Cabinet Oce employee. In the words rom one interviewee: theCIO council is run on grace and avour, and requests rom the Cabinet Oce will alwaysbe shoved down the agenda by Permanent Secretaries. The centres lack o a disbursementunction makes it easier or departments to simply ignore Cabinet Oce requests.

Second, it may mean that the centre cannot und cross-governmental projects itsel, makingthem potentially more vulnerable to changing priorities. One interview illustrated this pointby saying [take] Transormational Government as soon as the money ran out, no-one took

54 The two departments are the Department or Work and Pensions and the Foreign and Commonwealth Oce.

55 Figures stated by John Suolk. At: http://johnsuolk.typepad.com/john-suolk---government-cio/2009/08/public-sector-productivity-vs-it-investment.html


20/44


any notice. One may disagree with the specic example used (the reality clearly being morecomplex than suggested), but the principle remains true.

Third, it means the centre oten nds it easier to attract support or projects rom the privatesector than other departments. A predominance o private sector volunteers may skew theproject more towards serving the interests o industry than those o government.

Inormation and expertise

Evidence on the cost and purposes o departments IT is needed to identiy gaps andduplication; it is the oundation or eective coordination. Many interviewees claimed that

the centre does not possess sucient inormation to analyse IT in government eectively.Such a lack can mean that the centre simply does not know the scale or location o theproblems it is acing, and thus nds it dicult to ormulate a response. But inormationproblems also crop up ater a particular course o action has been identied. For example,much current Cabinet Oce work is based on selling the benets o participation inmutually benecial projects. A set o compelling evidence can greatly strengthen the caseor cooperation and the credibility o the seller: in the words o one interviewee, I youvegot the evidence base that this is the way to go, then by and large permanent secretaries arepretty sensible.56 It was argued that such an evidence base is oten lacking or incomplete.

Inadequate inormation collection is scarcely a new charge; the OEP, or example, examinedit in detail, and consequently work has started to build better data collection mechanisms

led by both the Cabinet Oce and OGC.

57

But possessing inormation is only hal thesolution: an organisation must also know what to do with it. Many interviewees expresseddoubts about the IT expertise available at the centre. O course, it would be unreasonableto expect sta at the centre to be able to cover all issues to the same level o expertise asspecialists, or to know departments as well as their CIOs. But many interviewees argued thatthat a higher level o IT expertise is needed at the centre o government; and clearly this islinked to resource constraints, as noted above.

Delivery

There were doubts over whether the Cabinet Oce had the delivery capability to ullits role in overseeing IT initiatives. These doubts are given some credence by the recentCapability Review o the Cabinet Oce, which rated developing clear delivery models as anurgent development area and suggested that the Cabinet Oces role in supporting delivery

o the Transormational Government agenda was not highly valued by other departments.58Having said that, interviewees were positive about the personal abilities o the currentGovernment CIO to develop consensus and exert sot power through relationships in orderto overcome delivery obstacles.

4. Lack o political integrationWhen considering politics and IT, its air to say that, rst, ministers should better understandthe attributes o the technology that is essential to achieving the outcomes they desire;and, second, the administration o IT should adapt better to political realities. Our researchidentied three main issues with the relationship between politics and IT.

56 For example, the latest initiative being led by the Cabinet Oce concerns a G-Cloud, as proposed in the Digital Britain report. Such workhas only started, but one o its priorities should be to oer evidence on the benets that wi ll ensue rom the initial investment. This is

particularly necessary given that there is some evidence that current cloud computing services may given only limited savings or may notactually be cost eective - or organisations with large IT inrastructures. McKinsey & Company (2009) Clearing the air on cloud computing.Drat discussion document. At:http://images.cxotoday.com/cxoimages/storyimages/matter101157.pd. Gartner (2008) DataquestInsight: A Service P rovider Road Map to the C loud Inrastructure Transormation.

57 HM Treasury (2009a), p.64; HM Treasury (2009b), p.27.

58 Cabinet Oce (2008) Capability Review o the Cabinet Ofce, p.11.


21/44


Lack o political concern or IT implications

Ministers requently do not pay sucient attention to the IT dimension o policyannouncements. As one interviewee explained, ministerial timerames and IT changeprogramme timerames are not the same. Ministers want to deliver IT-enabled businesschange very quickly. Given that, generally, ministerial priorities trump all, this need or speedis particularly likely to inhibit cross-cutting initiatives that are not associated with a particularminister even i they oer considerable savings. As one interviewee explained, i you ask toreuse the DWPs systems, they may be ne with it but may need six months to sort it out and you are under pressure because your minister has already announced a deadline. The

overriding importance attached to individual ministers initiatives, coupled with their lack oattention to IT issues, oten leads to the wheel being reinvented across government.

IT enters too ar down-stream

IT is oten seen as an ater-thought because it enters the decision-making process too ardown-stream, when plans and deadlines are already airly denite. One interviewee saidthat the levers available to the centre may only be pulled very late in the day, at which pointthere may already be substantial political commitment... the solution is to intercept theIT issues in time so they can help rather than hinder business outcomes. I this happened,the possibility o reusing components would be raised much earlier in the policy-makingprocess, meaning that lead-times are more likely to t with a ministers deadline. O course,the aleatory nature o politics means that there will not always be time to build IT in early and it is perhaps unrealistic to think that IT considerations will be at the ront o ministers

minds. Even so, more could be done to ensure that consideration o whether proposals willrequire complex IT implementation enters the policy-making process sooner, and that thebias is towards simpler IT requirements.

Complexity o political arrangements or leadership on IT issues

Over the past decade, its clear that ministerial responsibilities or IT in government havebeen mutable and overlapping. Broadly speaking, there have been three main sources oministerial input: a) in the period 1999 2004, the e-Envoy reported to the Prime Minister;b) the Cabinet Oce has had a Minister (usually a Parliamentary Secretary) responsibleor a public service reorm brie that includes some orm o electronic government; c) theDepartment or Trade and Industry and its successors have had a Minister o State with abrie that includes e-commerce.59 Responsibility has also been given to ministers or major

individual IT projects: the Minister o State (Health Services) in the Department o Healthcurrently oversees the NHSs National Programme or IT, or example.

Two relevant points arise rom this setup. First, it is clear that the e-government agenda has otenbeen split across departments: in 2003, responsibility or e-transormation o public services ellunder the Cabinet Oce, while crucial elements such as privacy, data-sharing and data protectionwere handled by the Department or Constitutional Aairs.60 O course, there are advantagesto distributing specic IT projects in this way, since it helps them to remain connected to thebusiness o their department. But cross-governmental issues may lose coherence unless thereare requent cross-departmental contacts between ministers. Interestingly, between 2008 and2009 Tom Watson had a airly coherent set o IT responsibilities as Parliamentary Secretary at

59 The Annex gives a list o the relevant ministers. The e-commerce role has been included because it was requently perceived as being thee-minister role, and so the holder requently made high-prole announcements on Government IT. See, or example, http://news.zdnet.co.uk/security/0,1000000189,2078369,00.htm. E-commerce no longer appears to be a ministerial responsibility. See Cabinet Oce(2009) Current list o ministerial responsibilities, at: http://www.cabinetoice.gov.uk/media/301888/lmr-oct09.pd

60 http://www.epractice.eu/en/news/283852


22/44


the Cabinet Oce; since June 2009, these responsibilities are more ragmented and sit with threeministers in three dierent departments.61

The second point ollows logically. Ministerial responsibility or IT has oten been combined withother, oten unrelated, duties making it harder or ministers to build expertise coherently, andraising the risk that IT may get lost amongst the rat o other roles. The table below shows thecompeting duties or the three current ministers with signicant IT responsibilities:62

61 See the ollowing table.

62 Taken rom Cabinet Oce (2009) Current list o ministerial responsibilities.


23/44


Name and role

Responsibilities (thoserelevant to cross-

government IT in bold)

Angela Smith

Minister o State, CabinetOce

Third Sector

Ofce o GovernmentChie Inormation Ofcer

Inormation Security andAssurance

Social exclusion

UK Statistics

Departmental responsibilityor Better Regulation andHuman Rights

Civil service issues

Jim Knight

Minister of State, Departmentfor Work and Pensions

Minister or the South West

Labour market and theeconomy

Labour market statistics

Welare Reorm

Jobcentre Plus

Employment programmes,including the uture o theNew Deal

European Social Fund

Ethnic minority employment(Chair o EME taskorce)

Migrants, reugees andasylum seekers

Adult Disadvantage(including ex-oenders)

City Strategy

Employers

Skills

Disadvantaged areas andregional issues

Tax Credits (where DWP hasan interest)

Habitual Residency Test

National Insurance Numbers(NINOs)

Service Transormation andChange Programme

Departmental IT and datasecurity

Directgov

Digital Inclusion

Departmental ManagementIssues

Stephen Timms

Financial Secretary to theTreasury

Parliamentary Under-Secretary o State, BIS

Strategic oversight o theUK tax system as a whole

including direct, indirect,business and personal taxation

Tax credits and integration othe tax and benet system,including welare reorm andchild poverty

Departmental Minister or HMRevenue and Customs and theValuation Oce Agency

Lead Minister on Europeanand international tax issues

and assist where necessary onbroader European issues

Overall responsibility or theFinance Bill

The voluntary sector, charities,including Corporate SocialResponsibilities

Assist the Chancellor onEuropean and Internationalissues.

Digital Britain

Communications andcontent industries, includingcreative industries; and ITand electronics sector


24/44


5. Central expertise in procuring and managing IT projects doesnot always count

The impact o the OGC

Government procurement o IT is a large and complex eld; this study cannot consider allthe issues it raises. Broadly speaking, the OGC is considered to provide airly good advice ongovernment IT procurement. Its deals with Microsot, or example, have been credited withsecuring a 50 per cent markdown in the bulk purchase o government sotware. It has a goodstrategic overview o procurement and its Procurement Capability Reviews have generated

some impressive results.63 Despite this progress, questions remain over whether morecould be achieved i OGC had greater authority to ensure departments act on its strategicdirection. A recent statement by the Chie Executive o OGC is interesting in its equivocationon the OGCs clout:

I have got to say I do not have one instance where I have recommended that somethingshould be done when it was not done. I do believe I have clout; the issue or me is notwhether I personally have clout because, with the best will in the world, I am going to getinvolved in only a small number o maybe highly sensitive contracts; the issue is whetheror not we institutionalise the process whereby there is clout in the process so that when Idisappear at the end o my term the clout does not disappear with me as an individual. 64

This report concurs with the need to give processes clout. For example, the centre imposes

relatively ew ormal checks or a department wishing to initiate IT procurement.65 In termso overseeing project management, a recent NAO report on commercial skills or complexgovernment projects concluded that there has been a lack o departmental engagementwith some OGC initiatives... and some departments continue to run initiatives whichduplicate those o the OGC.66 As the Digital Britain report noted, currently, nal sign-oor all new internal system procurements rests with individual departmental AccountingOcers; the Government CIO is consulted but it is not his decision. 67 The rest o thissection considers the centres oversight o IT-based project and contract management, sincethe NAO recently concluded that OGC could do more to improve departments contractmanagement throughout the lie o the project.68

Oversight o project management

OGCs Gateway Reviews oer an established means o overseeing the progress o majorIT projects and highlighting potential problems beore they escalate. Ian Watmore, givingevidence to the Public Accounts Committee, recently claimed that Gateway is one o thebest examples o quality assurance reviews I have seen in project management in 25 or 26

years... they do eed in lessons at the right time.69 Since Gateway reviews are not madepublic, it is dicult to assess their impact. Those in the public domain have been releasedthrough Freedom o Inormation requests, and thus usually deal with controversial or ailedprojects. In other words, the 57 reviews that have been released may oer a biased sample o

63 Oce o Government Commerce (2009) Procurement Capability Reviews End o Wave 1 Overview Report.

64 Nigel Smith, Oral Evidence to the Public Accounts Committee, 14th January 2009

65 Initially, the only contact a department needs to have with the centre is to apply the OGCs Risk Potential Assessment tool. Under anEU Directive, the department must liaise with HM Treasury to have projects worth over 97,000 publi shed in the Ocial Journal o theEuropean Union.

66 National Audit Oce (2009) Commercial skills or complex government projects, p.8.

67 DCMS and BIS (2009) Digital Britain: Final Report, pp.213-4.

68 Nigel Smith, Oral Evidence to the Public Accounts Committee, 14th January 2009; National Audit Oce (2009) Central governmentsmanagement o service contracts, p.6.

69 Public Accounts Committee (2009) Learning and Innovation in Government, Ev.4.


25/44


the total 2,500 conducted.70 Nevertheless, the 31 Gateway Reviews on the NHSs NationalProgramme or IT (NPIT) alone make a case or the centre to exert more infuence overdepartments major IT projects. It is clear that work on the various strands o the NPIT didnot suciently heed warnings and negative reports rom Gateway reviews.71 Indeed, theOGC was only asked to perorm two OGC reviews o the programme as a whole, both atvery early stages. This was despite its recommendation that more reviews should be carriedout.72 The second, nal, review was published in November 2004 and gave the overall statusas Red, pointing out the lack o a coherent benets realisation strategy and the absence oclarity regarding the organisational structure.73

This case highlights two points about the role o the centre: rst, that a department was ableto continue a Red rated project without signicant adjustments;74 second, that whateverOGCs concerns, in practice it only carries out Gateway Reviews when requested. It appearsthat the departments position o power meant that it could respond to unwelcome news byailing to request any more news at all.75 Similar ailings are evident in the Public AccountsCommittees recent, damning, report on the National Oender Management System.76With this in mind, there is reason to be cautious about the news that the well-receivedProcurement Capability Reviews are now to be carried out by departments themselves.77

In sum, although Gateway reviews can provide an eective challenge process... there isevidence they are not always taken seriously.78 In other words, Gateway reviews maynot have sucient impact to infuence a departments management o a contracted IT

project, no matter how fawed it may be. There have been some attempts to address thisproblem: now, a single Red rating means the Chie Executive o the OGC writes a letterto the permanent secretary, it is disseminated to the centre o excellence on PPM, it isdisseminated right across the department, it goes to the NAO, the NAO notiy... the PAC.79The OEP also makes some welcome recommendations to strengthen the governance o IT-enabled projects, but these are mostly based on the premise that CIOs can exert signicantinfuence rom their place in the departmental hierarchy. For all the reasons outlined above,such infuence may not be possible; there is a case or increasing the level o sensible,discerning checks rom the centre.

The length o procurement

Having said that, there is a need to be cautious about extending oversight rom the centretoo much: it can become cumbersome. There is a danger that applying a standard central

oversight procedure to the wide range o government IT procurement may require signicantbureaucratic eort. Indeed, there are some concerns that current government procurement70 The gure o 57 is compiled rom: http://www.ogc.gov.uk/ogc_and_the_reedom_o_inormation_act_blank_page.asp; http://www.

computerweekly.com/blogs/tony_collins/2009/06/order-to-publish-more-gateway.html.

71 http://www.connectingorhealth.nhs.uk/about/oi; http://www.computing.co.uk/computing/comment/2244750/unheeded-warnings-highlight-nhs-4731810

72 Similarly, the NAO has noted that Gate 5 Reviews, which assess the benets o a programme, are only applied by 20% o departments despite the act they are mandatory. NAO (2009) Helping Government Learn.

73 Oce o Government Commerce (2004) NHS National Programme or IT: Gateway Number 214. At: http://www.connectingorhealth.nhs.uk/about/oi/g0npt0217.pd

74 This is not an isolated instance. In 2006, the NAO recommended a review o how red reviews, and multiple red reviews, are dealt within uture in terms o guidance to the senior responsible owner and bringing them to the attention o the Permanent Secretary. NAO(2006) The delays in administering the 2005 Single Payment Scheme in England.

75 Ipsos Mori conducted two annual surveys on the National Programme or IT that produced unwelcome results. No more reviews werecommissioned, although there may be no causal link. http://www.computerweekly.com/blogs/tony_collins/2009/06/16-key-points-in-gateway-revie.html

76 Public Accounts Committee (2009) The National Oender Management Inormation System.

77 http://www.civilservicenetwork.com/latest-news/news-article/newsarticle/ogc-takes-back-seat/. In deence o the move, OGCargues that the PCR process is based on rigorous quantitative analysis that is not amenable to manipulation.

78 Public Accounts Committee (2009) Learning and Innovation in Government, p.10.

79 Nigel Smith, Oral Evidence to the Public Accounts Committee, 14th January 2009.


26/44


is currently too unwieldy. Intellect, the trade association or the technology industry, hasclaimed that enormous savings can be made by urther reducing the cost and time opublic sector procurement. In particular, it notes that the average time to let a major ICTprocurement in the UK is 15 months, compared to nine months in Europe and as little asthree or our months in a crisis.80 Indeed, procurement can take three years rom start tonish,81 oten leaving government with an end product that is ten years old in technologyterms. O course, lethargic procurement is not a good reason in itsel or limiting oversightas such, since the delays could be addressed by tackling other causes. But the level o suchoversight needs to be based on eective risk assessment82 and interventions rom the centre

need to be targeted rather than excessive.

The overall pictureThe major issues aecting IT across government identied by this report are summarised below.

The Treasury and Cabinet Oces approaches to coordinating IT in government are1.collegial and based on persuading not mandating. This model has advantages - theCIO council works well together and has developed a good strategy but it is notsucient to tackle thorny problems o cross-department cooperation that hinder therealisation o better outcomes or citizens.

Departments are reluctant to cooperate with the centres agenda because they have2.other priorities; dislike central interventions; perceive it to be risky; may not appreciatethe importance o IT; and have complex internal power structures.

The CIO Council is based on departmental CIOs having power in their own3.departments. However, the basis or such power is variable and fuctuating: the typeo work carried out by the department; the CIOs ormal role and resources; andpersonal relationships.

There is a mismatch between the centres role and its resources. The Cabinet4.Oce has a dependent relationship with other departments and cannot commandattention through disbursement. The centre does not possess sucient inormationto analyse IT in government eectively, and may require a higher level o IT expertise.The Cabinet Oce may also lack the delivery capability to ull its role in overseeingIT initiatives.

There is a lack o political concern or the IT implications o policies. Given that5.

ministerial priorities trump all, cross-cutting initiatives are likely to lose out. IT entersthe decision-making process too ar down-stream, oten when rm commitmentshave already been made. Political leadership or IT has been ragmented betweendepartments and buried inside ministerial bries.

While the centre has expertise in procurement, departments have considerable reedom6.to initiate and manage IT projects regardless o its advice. However, it is important toprevent the centres helping hand rom becoming an oppressive dead hand.

The nal section oers recommendations to address these issues.

80 Memorandum rom Intellect, in House o Commons Treasury Committee (2009) Evaluating the Efciency Programme, Ev 46.


82 The OGCs Risk Potential Assessment is currently being redeveloped. http://www.ogc.gov.uk/programme_and_project_news_the_sutherland_inquiry.asp


27/44


Recommendations

Our recommendations outline the principles that should underpin an eective centre-department relationship or IT, and suggest how these principles could be realisedthrough actions. Finally, we speciy the initial structural changes that will help such arelationship develop.

The principles o an eective relationshipI IT is intended to help achieve strategic outcomes or government, it seems obviousthat the role o the centre needs to be considered in terms o how it contributes to theseoutcomes. Looking rom the outside in allows us to identiy what we needthe centre to do,

versus the line departments, rather than simply trying to divide up responsibility or pre-existing roles.

In crude terms, it is possible to identiy our roles or IT in government:83

Improving services. In other words, allowing government to do new things,both within and between departments. This role is strongly represented inthe Transormational Government agenda and the Digital Britain reportsGovernment o the Web concept.84

Improving efciency. Allowing government to do the same things more ecientlyand productively. For example, this is the thrust o the attempts to share servicesbetween departments.85

Enabling itsel. Supporting the preceding two roles by controlling any issuescaused by the specic use o IT. For example, ensuring common standards,inormation security, and eective project management o IT-enabled change.

Involving citizens. Allowing citizens to access, consume and recombine inormationheld by government, and contribute towards this store o inormation. The Power oInormation Task Force outlines this role in more depth.86

In sum, IT must be considered in tandem with the purpose and design o governmentas a whole; this means that it should neither be seen as a stand-alone panacea, norsomething or specialists only: its ull benets will be only realised when it is tightly woveninto the work o government. This means that although new technologies open up newpossibilities or government, the end must be kept in sight improved outcomes or citizens.Government should take advantage o new technology, rather than allowing new technologyto take advantage o it. This requires a healthy dialogue between these two orces, wherebystrategy inorms technology and technology strategy. I the overall purpose o IT is orgotten,then it is very possible, or instance, to gold-plate technology and provide more unctionsand repower than are needed.

The centre should act as the guardian o the wider government outcomes and ensure thatthere is this healthy tension between technology and strategy.87 Accordingly, permanentsecretaries and senior managers need to have a rm grasp on the relationship between IT83 For a uller discussion o the role o IT in government, see Dunleavy, et al. (2008) Digital Era Governance: IT corporations, the state and

e-government, Oxord University Press (2nd edition).

84 http://www.cabinetoice.gov.uk/cio/transormational_government.aspx; DCMS and BIS (2009) Digital Britain: Final Report, pp.208-224.

85 The Operational Eciency programme is clearly related to this role, but it ocuses on improving the eciency o IT itsel - rather than onimproving the eciency o government through IT.

86 http://poit.cabinetoce.gov.uk/poit/

87 This accords with the recommendation o the Chakrabati Review that the Cabinet Oces oer be based on an overarching view onGovernment priorities. Chakrabati (2007) Role o the Cabinet Ofce: Leadership through eective collaboration, p.2. At: http://www.cabinetoice.gov.uk/media/cabinetoice/corp/assets/publications/reports/chakrabarti_review/chakrabarti_review_co.pd

The centre shouldact as a guardian orwider governmentoutcomes andmaintain a tension

between technologyand strategy


28/44


and strategic outcomes, and attempting to achieve this should be one o the main tasks othe departments CIO. CIOs, in turn, should be able to draw on the centre as a respectedresource that provides high-quality, relevant inormation. The ability to make a strategic,evidence-based case or a particular course o action is crucial because departments aresubject to a vast array o competing pressures. Making trade-os amongst these pressuresmay be the permanent secretary or boards responsibility, or it may be a political decision.Either way, the centre and CIOs should be able to make a convincing strategic case or the ITdecisions being recommended or enable the permanent secretary to do so. I ministers areinvolved, then the strategic case should be presented early in the decision-making process.

The orthcoming Government IT strategy should help to link individual cases into thebroader cross-government context.

Stronger, more strategic recommendations are all very well, but they all victim to whatDavid Henderson called the unimportance o being right.88 In other words, i the centregives recommendations alone they are liable to be dismissed, no matter their strength, orthe reasons given above. Indeed, past experience shows that coordinating IT below the levelo permanent secretaries lacks impact in the nal analysis. More is needed to tackle thecontested areas where the current setup ails. I the centre is to act as the guardian o widergovernment outcomes, then, like most guardians, it needs some real infuence to guide anddirect. We suggest our interrelated ways such infuence can be ashioned and nurtured:discerning intervention, knowledge and expertise, relationships o respect, and authority.

Discerning interventionCoordination rom the centre needs to be selective, allowing departments reedom toinnovate and use technology to achieve societal benets in their areas, while insisting onsavings where the need and evidence is compelling. Thus, thecentre needs to understandwhere it can add value by intervening - when it is best to standardise and when it isappropriate to allow departments reedom to make judgements themselves, since this canlead to innovation and positive deviance89 with net benets or all. There are three mainways o building such an understanding.

First, the centre needs to assess how important the issue is to the governments wider ITstrategy, and to achieving the departments objectives.90 To what extent will implementingthe technology as proposed jeopardise success in either or both domains? In other words,the centre must know what departments have to accomplish and oer ways or them to

do it better.

Second, the centre needs to understand, based on evidence, whether the current proposalswill have any plainlyundesirable eects, be they intra- or inter-departmental (e.g. excessiveor unnecessary costs, lack o interoperability).

Finally, the centre should have a conceptual ramework or understanding its interventions.The task o overseeing the relationship between technology and strategy changes accordingto the issue at stake. There are some issues where it makes sense or departments to havea great deal o agency: perhaps speciying and procuring customer-acing systems, sincethe department knows its customers and has an overall strategy or dealing with them. Butor more basic, underlying IT unctions, the centre should have the authority and strategicinsight to decide the best course o action. Put simply, the CIO should be aware when to

88 P. D. Henderson (1977) Two British errors: their probable size and some possible lessons, Oxord Economic Papers, 1977, pp.159205, p.190.

89 Spreitzer and Sonenshein (2004) Toward the construct denition o positive deviance,American Behavioral Scientist, 77(6): 828-847.

90 A government-wide IT strategy is currently being developed.

More is needed totackle the contestedareas where thecurrent setup ails

The centresinterventions needto be selective


29/44

Installing

installing new drivers: how to improve government's use of it

Documents