Full report (pdf - 613KB) - National Audit Office

The National Offender ManagementInformation SystemREPORT BY THE COMPTROLLER AND AUDITOR GENERAL | HC 292 Session 2008-2009 | 12 March 2009

The National Offender ManagementInformation SystemLONDON: The Stationery Office£14.35Ordered by theHouse of Commonsto be printed on 10 March 2009REPORT BY THE COMPTROLLER AND AUDITOR GENERAL | HC 292 Session 2008-2009 | 12 March 2009

CONTENTSSUMMARY 4PART ONEBackground to NOMIS 11PART TWOCauses of delay and cost increases 16PART THREEProgress since the moratorium in 2007 24APPENDICES1 Study scope and methodology 302 Technical Report Management 32Summary3 Offender information exchange pre 34and post NOMIS4 External review recommendations on 36information exchangeGLOSSARY 37Photograph courtesy of National Offender Management Service

SUMMARY7 This report examines the reasons for the delaysand cost increases to the original project and, since themoratorium, the aims of the revised NOMIS programmeand the progress made, the impact of the delays andrescoping on the costs and benefits achieved, and NOMS’fitness to deliver. Our methods are described in full inAppendix 1 and include:■■■a technical analysis of the project (summary atAppendix 2);assessment of the original project against the eightcommon causes of project failure; andinterviews with stakeholders including visits toprobation areas and prisons to assess the impact ofthe delays.Key findingsOn the reasons for the delay andcost increases8 There was inadequate oversight by seniormanagement. The NOMS Board relied on the C-NOMISProject Board to escalate issues to it through NOMS’routine reporting systems. Between the start of the projectand May 2007, the NOMS Board did not ask for or receiveany other more specific reports on project progress.Whilst the Project Board met at least once every twomonths, it did not actively monitor delivery of the projectand was unaware of the full extent of delays or theimplications of decisions it made upon project cost.9 NOMS did not put the appropriate resourcesand structures in place to deliver such a complexproject. The overall governance and resources appliedwere not adequate given the scale of the task. Rolesand responsibilities were blurred, in particular financialaccountability was unclear, and insufficient skilledresource was applied to the project. Operating within thenewly-created NOMS entity (a merger between HM PrisonService, the National Probation Service and associatedheadquarters and Home Office staff) contributed tothe confusion.10 Programme management was poor in key aspects,including planning, financial monitoring and changecontrol. NOMS initial planning was overly optimisticin terms of both cost and timescales. For example therewas no contingency, despite some recognition that theproject carried a high level of risk. There was no overallplan bringing together the different workstreams (up to40 at one point) and hence it was difficult for managersto assess overall progress. Budget monitoring was absent,with cost control focusing on monitoring the spendingagainst the annual budget rather than matching costagainst project deliverables. Change control was weak,and there was no process in place for assessing thecumulative impact of individual change requests on theproject budget or delivery timetable. In addition, theVAT implications of implementing C-NOMIS in privateprisons were not grasped early enough and firmly enough,resulting in these prisons being first in, then out, and nowpotentially again within the scope of the project.11 NOMS significantly underestimated the technicalcomplexity of the project. A single offender databaseis technically realisable, but NOMS did not adequatelyexplore other potential solutions and underestimated thecost of customising the software it had already selectedfor the Prison Service. This package appeared to providea reasonable fit with prison requirements but met, infull, only 29 per cent of probation service requirements.In fact, neither assessment was correct, and businessrequirements also changed as policy developed. As aresult the estimated cost of developing the applicationrose from £99 million when the full business case wasapproved in June 2005 to £254 million by July 2007,primarily because of customisation.12 NOMS underestimated the need to invest inbusiness change alongside the IT system. There are42 probation areas in England and Wales, each withtheir own ways of working. NOMS, however, made nosustained effort to simplify and standardise businessprocesses across prisons and probation areas, and didnot allocate resource for this purpose. At the outset,NOMS treated C-NOMIS as an IT project rather than amajor IT-enabled business change programme. This statusincreased the pressure on the project team to approverequests for changes and contributed to significantscope creep.6 THE NATIONAL OFFENDER MANAGEMENT INFORMATION SYSTEM

SUMMARY13 NOMS’ contractual arrangements with its keysuppliers were weak and its supplier management poor.Instead of tendering key project contracts, NOMS optedto use its current suppliers under existing frameworkagreements to develop and deliver the C-NOMISapplication. NOMS allowed these contracts to goforward on a time and materials basis for longer than itshould, which meant that there was insufficient pressureon suppliers to deliver to time and cost. In addition,NOMS’ relationship with its suppliers, particularly EDS,deteriorated during 2005 and 2006, and it did not makebest use of their expertise. Significantly, NOMS did notseek to revise its contractual arrangements with Syscon,the software developers, immediately the extent of thecustomisation became clear. Syscon is now able to marketthe improved software, but taxpayers will not benefit fromtheir investment in the product.14 In conclusion, although technically feasible,C-NOMIS was a very ambitious project thought to havethe potential to bring much closer working across thecriminal justice system. The desirability of the project’saims appears to have overly influenced decision makers,leading to the failure to evaluate other technical optionssufficiently and establish realistic budget, timescales andgovernance for the project.On the aims of the revised NOMISprogramme and the progress made so far15 NOMS followed an objective process to identifyand compare different options to rescope theprogramme in line with industry practice, and selectedthe lowest cost option available. In Autumn 2007, NOMSassessed seven different options considering strategicalignment, value and benefits, achievability and risks.In December 2007, the NOMS Board selected the lowestcost option though, by its own analysis, this option did notoffer the greatest benefit to cost ratio. Funding constraintsproved to be the main decisive factor.16 The option selected has evolved sinceDecember 2007 but has not been compared topreviously rejected options, including continuingwith the original C-NOMIS solution, on a like-for-likebasis. Since the rescoping decision was made, NOMShas undertaken further evaluation of the options for aprobation case management system and has identifiedways to make significant reductions in cost by, forexample, simplifying data migration. Although it has notconsidered formally whether applying these methodsto C-NOMIS would make it affordable and deliverable,NOMS’ present view is that the original C-NOMIS visionis both unaffordable and not required operationally todeliver end-to-end offender management. There is stillsome uncertainty over the costs of the Delius and the datashare options.17 NOMS is making steady progress on deliveringthe full C-NOMIS application to prisons for the start ofnational deployment by April 2009. Roll-out dates forthe other projects in the programme are not knownand risks remain. There are currently some delays, withsome 60 per cent of C-NOMIS deliverables two monthslate, but both the supplier and NOMS are confident ofdelivering the full application by April 2009. Successfuldelivery of the national probation case managementsystem is dependent on an existing probation serviceinfrastructure project which is itself delayed, although not,according to NOMS, critically so.On the impact of delays and rescoping oncosts and benefits18 The full financial impact of the delays andrescoping of the programme is uncertain but is likelyto be at least £41 million. We have not been able todetermine the full value of the waste and inefficienciesassociated with the failure of the C-NOMIS project withcertainty because of NOMS’ poor recording of costs.We estimate that NOMS has spent around £15 million onthe probation element of C-NOMIS which is no longerrequired. Costs to NOMS from benefits deferred due tothe delays and costs associated with inefficient processesand system failures are likely to add a further £25 million.Additional EDS costs following the moratorium to closedown work were around £1.2 million.19 Many benefits of the programme are capableof being realised but the key aim of creating a singledatabase of offenders directly accessible by prison,probation and third party intervention providersresponsible for offenders from sentencing through toresettlement will not be met. The programme will deliversignificant improvements over existing systems. One of themain aims of the programme, to replace the ailing LIDSsystem in prisons, will be met and together with a nationalprobation case management system, over 220 separatedatabases will be replaced by just three. The othermain aim of the programme, to support end-to-endoffender management, will however not be fulfilled asoriginally envisaged.THE NATIONAL OFFENDER MANAGEMENT INFORMATION SYSTEM7

SUMMARYOn NOMS’ fitness to deliver therevised project20 NOMS has worked hard to address pastweaknesses and has made significant improvementsacross the key areas including governance, leadership,project management and financial control. Since summer2007, when a new Senior Responsible Owner brought theproblems to light, NOMS has undertaken detailed reviewsof its past performance and strengthened the programmeconsiderably in all areas.21 Some weaknesses remain, particularly inresourcing, business focus and financial forecasting.Recruitment difficulties mean the programme is heavilyreliant on contracted-in staff, and not all vacancies arefilled. We believe that NOMS still needs to standardiseways of working, particularly across the NationalProbation Service, to reduce scope creep and ensure thatthe introduction of the new systems goes smoothly. WhilstNOMS has made significant improvements to financialcontrols, it still lacks the means to match spending againstdeliverables although it is pursuing this shortcoming in itscurrent contract negotiations with potential suppliers.Conclusion on value for money23 Overall the C-NOMIS project was handled badlyand the value for money achieved by the project waspoor. Many of the causes of the delays and cost overrunscould have been avoided with better management of wellknown issues. NOMS’ failure at the start to appreciate theproduct customisation and business change required, itsinadequate oversight of the project and weak relationshipswith suppliers led to a doubling in programme costs, athree year delay in programme roll-out and reductionsin scope and benefits. In particular the core aim of ashared database to provide a single offender recordaccessible by all service providers will not be met, thoughover 220 separate databases will be replaced by justthree. Although programme management has improvedconsiderably in the last 18 months, weaknesses remain.Evaluation of project weaknesses against theeight common causes of project failure22 The key factors leading to the failure of theC-NOMIS project could have been prevented.In 2002, the Office of Government Commerce and NAOagreed eight common causes of project failure whichwere disseminated to Departments in February 2003.In May 2005, as part of the project approval process,the Home Office’s Programme and Project ManagementSupport Unit certified C-NOMIS as not suffering fromthe common causes of failure. Based on our analysis ofthe underlying causes of the cost increases and delay,we found that C-NOMIS suffered from four of the eightcommon causes of project failure in full and three in part(see Figure 2).8 THE NATIONAL OFFENDER MANAGEMENT INFORMATION SYSTEM

SUMMARY2Assessment of C-NOMIS against the eight common causes of project failureCommon causesof project failureLack of clear linkbetween the projectand the organisation’skey strategic priorities,including agreedmeasures of successLack of clear seniormanagement andMinisterial ownershipand leadershipLack of effectiveengagement withstakeholdersLack of skills andproven approach toproject managementand risk managementToo little attention tobreaking developmentand implementationinto manageable stepsEvaluation of proposalsdriven by initialprice rather thanlong-term value formoney (especiallysecuring delivery ofbusiness benefits)Lack of understandingof and contact withthe supply industryat senior levels of theorganisationLack of effective projectteam integrationbetween clients, thesupplier team and thesupply chainCause offailure evidentin C-NOMIS?In partYesIn partYesNoIn partYesYesCommentary on C-NOMIS project performanceWe found a clear link between the objectives of the C-NOMIS Project and the strategicpriorities of the NOMS Change Programme. Assumptions were tested during a pilot thatalso provided valuable lessons learned. However, project planning lacked both detailand robustness, and there was a failure to produce a single integrated plan representingall the planned tasks from across the individual work streams.The Senior Responsible Owner (SRO) from project inception to April 2007 hadlittle experience of major IT project delivery and insufficient time to undertake SROresponsibilities despite some reduction in her other responsibilities from September 2006.Until the moratorium, we found no evidence of Ministerial involvement beyond themreceiving standard summarised briefings.Leadership was compromised by blurred accountability between the project board andOffender Information Services (OIS) which led to reactive and indecisive decision making.Engagement with stakeholders was on an ad hoc basis. Early on, users were giventhe chance to voice their opinions of the C-NOMIS application and user groups wereinvolved in development of the system requirements. There was little communication withstakeholders on project progress. With project plans failing to schedule engagementactivities, many stakeholders were unsure how best to communicate with the project.When delays occurred, many stakeholders only found out at the last minute.The project suffered from a consistent shortage of skilled personnel, leading to a greaterreliance on contracted-in staff. A lack of financial skills led to poor financial estimatingand monitoring.Risks identified at the outset were not managed. Under the governance structure, theProgramme Manager was the only report to the Programme Director, which led to agood news culture.The project team’s decision to undertake a software trial and an initial pilotimplementation provided valuable learning. Breaking the development and deploymentinto manageable steps was sensible for a project as large and complex as C-NOMIS.The original business case, on which the decision to proceed was based, hadprojected C-NOMIS would deliver a positive net present value, but costs were seriouslyunderestimated. Despite recognition that the project was high risk and had a challengingdelivery schedule, there was no contingency within the budget, suggesting either a desireto keep costs down to achieve the go-ahead or a high degree of optimism.An existing framework contract within the Prison Service and an infrastructure renewalcontract within the Probation Service constrained the level of engagement with externalsuppliers and restricted competitive procurement opportunities. Although the maindelivery partner provided a detailed set of assumptions as part of their Best and FinalOffer, there was no evidence of an assessment to determine whether requirements couldbe met by suppliers, given competing pressures from other sectors of the economy.The project selected existing suppliers as the main delivery partners, and did not seeka wider evaluation of their requirements from the market. Although suppliers wererepresented on technical governance boards, project delays adversely impacted workingrelationships and there was a lack of communication channels at a more senior level.Source: National Audit OfficeTHE NATIONAL OFFENDER MANAGEMENT INFORMATION SYSTEM9

SUMMARYRecommendationsIn the short terma The National Probation Service covers 42individual areas with different ways of working,which inhibits the development of one nationalsystem. In order to mitigate the risk of scope creepand implementation difficulties, NOMS shouldrequire areas to develop a common approachto handling data within the new probation casemanagement system, through a business changeprogramme with a specific business change teamand budget.bcFinancial controls have improved significantlyduring the lifetime of the project, but there arestill insufficient mechanisms to assess quicklyhow events impact cost. NOMS should develop itsfinancial model to demonstrate the key cost driversand introduce a clear understandable method totrack cost against progress and completed work.Although the size of the NOMIS programme teamhas increased, difficulties in recruiting staff havemeant a greater reliance on contractors, whichcould impact the ability of the central team tomanage the programme effectively. NOMS andthe Ministry of Justice should provide the necessarysupport, training and resources to reflect the size andcomplexity of NOMIS.In the medium to long termdThrough introducing a single offender databaseaccessible by prison and probation, the originalC-NOMIS application was designed for end-to-endoffender management. The revised solution,however, does not facilitate the sharing ofinformation as initially envisaged, and has alreadyled other agencies to develop their own databases,such as the Parole Board. In the absence of a singledatabase, NOMS should:■within two years of implementation, evaluatethe effectiveness of the data sharing facilitiesit is developing for the prison and probationservices and, in the light of experience, takeearly steps to plan any improvements required;■■assess the effectiveness of the interim datasharing methods it has established with otheragencies, and investigate the potential for linksto other agency databases; andmake clear the technical standards for thedevelopment of systems to facilitate thecreation of future database links. Futuresolutions need to be built with a view to thewider picture, interoperability and the potentialneed for future enhancements.Recommendations for GovernmentChief Information OfficereThe failure of the C-NOMIS project couldhave been prevented had basic principles andexisting good practice been followed by NOMS.The Government CIO should, with OGC, seekassurance that all government departmentsmanaging large scale IT programmes are activelyavoiding these well known reasons for deliveryfailure. Such an assurance system could, forexample, require all government departments,agencies and non-departmental public bodies’IT-enabled business change projects/programmeswith a total life value exceeding £100 million tocertify that they meet minimum standards, includingstandards of financial control, governance and seniormanagement oversight, commercial arrangementsand supplier management.10 THE NATIONAL OFFENDER MANAGEMENT INFORMATION SYSTEM

PART ONEBackground to NOMISPolicy and administrative context1.1 The National Offender Management InformationSystem project (C-NOMIS) formed part of a wider changeprogramme instigated following Lord Carter’s CorrectionalServices Review in 2003. 1 The review aimed to improvethe focus of correctional services in England and Waleson the offender and reducing re-offending, rather than onmanaging the prison and probation services. 2 Lord Carterrecommended bringing together the two services andintroducing “end-to-end offender management” in termsof the supervision of individual offenders throughouttheir sentence (whether in custody or in the community)by a single offender manager. Reducing re-offendingwould contribute in large part to reducing the numbers inprison, and reduce the cost and raise the efficiency andeffectiveness of running correctional services.1.2 The Carter Review found as follows.■■■Programmes and interventions received in prisonwere often not followed up in the community.Information sharing on offenders between serviceswas often poor. Prison and probation areasoperated different systems, with different ways ofcapturing data.No single organisation was ultimately responsiblefor the offender, with the result that there was noclear ownership on the front-line for reducingre-offending.1.3 In June 2004, the Home Office created the NationalOffender Management Service (NOMS), bringingtogether Her Majesty’s Prison Service (HMPS) and theNational Probation Service (NPS) within one entity. It alsostarted work on fleshing out how end-to-end offendermanagement could be implemented and the systemsrequired to support it.The genesis and aims of theC-NOMIS project1.4 The Home Office realised very early that end-toendoffender management would require better joined upinformation systems across prison and probation servicesthan were either in existence or planned. Both serviceswere already intending to replace their existing IT systems,but data sharing across the services did not form part ofthese projects. HMPS’s PRIME project begun in 2003,aimed to replace its ageing prisoner management system(Local Inmate Database System – LIDS) by April 2007,whilst the National Probation Service aimed to introducethe key elements of a national offender managementsystem beginning in 2005. The Home Office decidedto take the PRIME project and use it as the basis for anational offender management information system acrossboth services.1.5 The scale of the challenge was considerable. Thereare 141 prisons and Young Offenders Institutions inEngland and Wales housing around 80,000 inmates atany one time, with a further 240,000 supervised in thecommunity by the National Probation Service. Together,prison and probation offender management systemscurrently hold records on around 1.5 million individuals.1 Managing Offenders, Reducing Crime: A new approach, Patrick Carter, Cabinet Office Strategy Unit, December 2003.2 ibid page 4.THE NATIONAL OFFENDER MANAGEMENT INFORMATION SYSTEM11

PART ONE1.6 The Home Office launched the C-NOMIS project inJune 2004. Its aims were to:■■■■reduce re-offending through end-to-end offendermanagement across probation and prison servicesand including providers from the private andvoluntary sectors;introduce more assertive case management andcontestability in service providers to improveoutcomes and value for money;integrate IT support across custody and communitysentences to facilitate the breaking down of barriersbetween prison and probation; andimprove means of monitoring compliance withNational Standards, including more demandingcommunity sentences and greater control andoversight of persistent offenders.A key benefit of C-NOMIS was that a single offenderdatabase would provide real time information to offendermanagers, enabling them to manage offenders effectivelyacross organisational boundaries.Progress of the C-NOMIS project1.7 The PRIME project was being carried outfor HM Prison Service by Electronic Data SystemsCorporation (EDS) under an existing agreement. HMPShad already selected a software application called TAG,developed by the Canadian firm Syscon Justice SystemsLtd, to replace LIDS in prisons. In June 2004, after analysisof the probation service’s requirements, NOMS decidedto extend the contract with EDS to carry out a design andpilot evaluation to customise TAG to fit the needs of theprobation service and NOMS and roll out one applicationacross prisons and probation.1.8 Following the pilot phase, NOMS approved a furtherextension of the EDS contract, on a time and materialsbasis, to develop, test and roll out the C-NOMIS systemincluding the migration of data from LIDS and otherlegacy prison systems such as the Inmate InformationSystem (IIS). Similarly, NOMS employed Steria, underan existing IT infrastructure contract with the NationalProbation Service, to prepare the 42 NPS legacy casemanagement systems into a common data format ready tomigrate into C-NOMIS.1.9 The Home Office Group Investment Board approvedthe full business case for the project in June 2005. At thistime, the project’s lifetime costs to 2020 were estimatedto be £234 million (including design costs of £99 million)and the new system was due to be fully implementedand operational by January 2008. In September 2006,following a critical Gateway Review 4 3 , NOMS madechanges to the management and governance of theproject, establishing a programme structure under aprogramme director, in part to provide a greater focus onthe business change aspects of implementing a new ITsystem as well as the purely technical aspects.1.10 In August 2007, following a review, the project teamrecommended to the Minister of State that he stop theproject because it had become unaffordable. By this time,C-NOMIS was two years behind schedule and estimatedlifetime costs had risen to £690 million. A version ofC-NOMIS had been implemented in three prisons onthe Isle of Wight and roll out to a further 15 prisons wasplanned. EDS was also approaching the start of full testingof the probation part of the enhanced TAG application.The Minister imposed a moratorium on all further work bysuppliers while NOMS carried out a strategic review of theprogramme. By this point, NOMS had spent £155 million,mainly on the prison element of C-NOMIS 4 , and wasincurring costs of around £350,000 per working day,which if continued would lead to a cash overspend forthe year.3 Gateway reviews are carried out at key decision points during the lifecycle of all government IT projects. They provide assurance on whether a project cansuccessfully progress to the next stage. OGC Gateway 4 assesses a project’s readiness for service.4 C-NOMIS refers to the first development of the NOMIS software for use across prisons and probation.12 THE NATIONAL OFFENDER MANAGEMENT INFORMATION SYSTEM

PART ONEThe redesigned NOMIS programme1.11 In November 2007 the NOMS Board agreedto a rescoped programme at an estimated cost of£513 million (including sunk costs) (Figure 3). The maincost savings were to be achieved by avoiding £81 millionin VAT liabilities by excluding private prisons from theprogramme and by moving from a single database solutionto multiple databases. The key decisions were to:■■■■continue to implement C-NOMIS in prisons (nowcalled Prison NOMIS);halt work on developing C-NOMIS for probation,replace their case management systems Area byArea on a needs basis and consider the optionsfor improvement;develop a mechanism to share limited offenderinformation between prisons and probation, suchas the name of responsible officer and offender’slocation; andextend the programme scope to include a project toimprove the current separate prison and probationoffender assessment systems and redevelop themas a single national system. This element, calledOASys, had previously been proceeding as astandalone project.1.12 Since the rescoped programme was approved,NOMS has continued to explore the options for replacingprobation case management systems. Current plans are torebuild an existing application called Delius, used by fourof the 42 probation areas, to increase system capabilityand roll it out to all areas by March 2011. Figure 4overleaf compares the scope of the original C-NOMISproject with the revised programme and Figure 5 onpage 15 shows the original and latest timescales forthe programme.3Estimated project costs at key stagesTotal costs:£234mSource: NOMS£217.3m£16.7mFull Business CaseJune 2005Total costs:£690mAt moratoriumAugust 2007£535m£155mRe-scopedprogrammeTotal costs:£513mSunk Costs Future Costs 1£352m 2£161mOutline Business CaseDecember 2007NOTES1 Future estimated costs exclude non-cash costs such as cost of capitaland depreciation.2 Includes costs for the OASys replacement project which were notincluded in the original estimates.THE NATIONAL OFFENDER MANAGEMENT INFORMATION SYSTEM13

PART ONE4Changes in the scope of the NOMIS programmeOriginal C-NOMIS ProjectOut of ScopePrison ServiceReplace legacy systems in public and private prisonsDevelop ManagementInformation SystemBuild interfaces to otherCriminal Justice SystemsShared DatabaseBuild interfacesfor offender riskassessment systems(OASys)Replace OASysand e-OASysProbation ServiceReplace Case Management Systems in 42 areasReplace programmeattendance system (IAPS)Revised NOMIS ProgrammeOut of ScopePrison ServiceReplace legacy systems in public prisonsBuild interfacesto other CriminalJustice SystemsDevelop ManagementInformation SystemBuild Data ShareapplicationBuild OASys ande-OASys replacementBuild interface toIAPS & OASysReplaceProbation CaseManagementSystems withC-NOMISPrison NOMIS DatabaseNational Delius DatabaseReplace legacysystems in privateprisons withC-NOMISProbation ServiceReplace Case Management Systems with a centralisedversion based on a redevelopment of DeliusReplaceProgrammeAttendance system(IAPS)Activities originally not in scope Activities in scope Activities added to scope Activities postponed or droppedfrom scopeSource: National Audit Office14 THE NATIONAL OFFENDER MANAGEMENT INFORMATION SYSTEM

PART ONE5Key programme milestonesMilestonesC-NOMIS pilotworking in smallnumber of PrisonsJul 2006Dec 2006LIDS replaced withC-NOMIS in PrisonsSep 2007 May 2010Inmate InformationSystem replacementoperationalJan 2008 2011Single CaseManagement Systemoperating in theProbation ServiceSep 2007 20112006 2007 2008 2009 2010 2011C-NOMIS planned milestone,as outlined in ProjectInitiation DocumentActual delivery dateLatest proposeddelivery dateSource: NOMSNOTENew programme elements, OASys Replacement and Data Share, are planned for delivery by March 2011.THE NATIONAL OFFENDER MANAGEMENT INFORMATION SYSTEM15

PART TWOCauses of delay andcost increases2.1 Between June 2005 when the full business case wasapproved and August 2007, lifetime costs trebled from£234 million to £690 million, and the project fell two yearsbehind schedule. This Part of the report explains the reasonsbehind the delays and cost escalation. In addition to ourown analysis, we have drawn on reviews carried out byOGC, NOMS’ internal audit and suppliers.Overview of cost increases and delays2.2 All areas of the project increased in cost, with largeincreases in software development, system support and VAT(see Figure 6).2.3 A reappraisal of costs by a new project team duringearly 2007 was the first time full forecasts were broughttogether in detail from suppliers, and concluded thatadditional funding of £456 million would be needed to seethe project through as originally envisaged. For exampleEDS, the main supplier, identified that their costs fordevelopment, deployment and support and maintenancewere likely to increase from £95 million to approximately£200 million. The major cost elements in the originalbudget were £65 million for developing the application andinfrastructure, which would be mainly spent by March 2008,and £128 million on supporting the system once it wentlive, spread between 2007 and 2020.2.4 We have identified five main areas which contributedto the cost increases and delay which are explored in thefollowing sections:■ NOMS significantly underestimated the technicalcomplexity of the project, particularly the need tocustomise TAG for both prisons and probation services;■■NOMS did not get to grips with the business changesrequired to design and implement a single offenderdatabase across both services;NOMS’ contractual arrangements with itskey suppliers were weak and its suppliermanagement poor;■■NOMS did not put the appropriate resources andgovernance in place to deliver such a complexproject; andProject management was poor in key aspects,such as planning, financial monitoring andchange control.The inter-related nature of the factors above and a lackof detailed financial information available from NOMS,means that we are not able to quantify the separatecontribution of each weakness to the overall cost increase.The complexity of the project“The complexity of the project to replace and integrate thecase management systems for offenders across the wholeof the Prison and Probation Service was significantlyunderestimated in many aspects of the original C-NOMISBusiness Case.”C-NOMIS Financial Controller 20072.5 The design and implementation of a single offenderdatabase and management system across prisons andprobation services was always going to be a majorundertaking. Both the prison and probation serviceshave complex and differing requirements (Figure 7)and there was only 23 per cent common functionalitybetween the proposed prison and probation IT solutions.In 2004, when the need for a joined up IT system acrossboth services was first articulated, HM Prison Servicewas already at an advanced stage of its planning for areplacement to its existing LIDS system and this factorheavily influenced NOMS’ approach to C-NOMIS.In particular, it contributed to the decision to continuewith the TAG software and customise it to deliver aprobation case management system, and to use itsexisting supplier of prison infrastructure, EDS, to deliverthe project.16 THE NATIONAL OFFENDER MANAGEMENT INFORMATION SYSTEM

PART TWO6Breakdown of cost forecastsApplication development (software licences,system build and testing)Estimated costJune 2005£mEstimated costJuly 2007£mCost increases£m51 206 155System maintenance and support (to 2020) 128 243 115Infrastructure (hardware including refreshes) 14 38 24Implementation (data migration and staff training) 12 22 10Project management 13 34 21VAT 0 99 99Contingency 0 32 32Other (sunk costs, Management InformationSystems, Contracted-Out Prisons)16 16Total 234 690 456Source: National Audit Office analysis of NOMS’ financial reports7The functionality required by prisons and the probation serviceFunctionality required by both prisons and probationAdministrative details such as name, date of birth, sentence and offender managerDetail of interventions received and progressAutomated sentence calculationAccess and entry into offender’s progress recordDetail of prisoner incidents and staff observations whilst offender is in custodyRequired prison functionsDetail internal prisoner movements and prisoner visitsDetail of administration processes from sentence to prisoners’management in custodyPrison population and accommodation management functionRecord prisoner finances and help manage prisoner propertyDetail prisoner work, training and leisure activitiesProvide facility for prisoner data search and enquiryRecord all prisoner requests and complaintsAbility to share basic prisoner details with the healthcare systemRequired probation functionsProduce court reports and sentence plansElectronically refer offenders to intervention providers througha catalogueAutomatic generation of letters and notifications to the offender,solicitor and breach officerAutomatic generation of breach file with all relevant detailsAn appointment schedule and attendance recording facility foroffenders on supervision to support national standards monitoringVictim detailsImproved contacts, capturing contact information fromexternal agenciesBasic Skills Assessment templateCapability to interface with external agencies and providersSource: NOMSTHE NATIONAL OFFENDER MANAGEMENT INFORMATION SYSTEM17

PART TWOOther solutions were not explored2.6 A single offender database represented atechnically realistic solution to support end-to-endoffender management and replace ailing legacysystems. There were, however, other possible solutionsnot explored fully. For instance, the development ofa management information system linking multipledatabases might have been easier and cheaper, but thefull costs of such an approach were not estimated. There isno evidence of the team considering factors such as thenature of supporting infrastructure and the existence ofcommon levels of service, which should have informedthe selection of the technical solution in 2004.The chosen solution – TAG – appeared to bea good fit with prison requirements but notwith probation2.7 A comparison by HMPS in 2003 of the TAG systemfunctionality against business requirements found that TAGmet 73 per cent of prison service functional requirementswith the remainder met in part. Thus TAG appearedto provide a reasonable fit with prison requirements.In 2004, when C-NOMIS superseded PRIME, NOMSevaluated TAG against probation requirements. TAG metin full only 29 per cent of Probation’s 275 functionalrequirements (see Figure 8) and did not meet 31 per centat all.Customisation was high2.8 The level of customisation required to TAG is shownby the large number of changes made: around 800 changerequests in 2005 and 2006. Alongside the over-optimisticinitial assessment of TAG, high levels of customisationwere caused by the fact that, in 2004, the practicalitiesand IT system requirements for Offender Managementremained unclear. The Offender Management Model wasitself at the design stage (see box below). The difficultiesin developing a common approach were compoundedthrough prison and probation operating as separateentities, and the decentralised probation structure with42 autonomous areas each with its own way of doingthings. There is no documentary evidence of NOMSformally assessing Offender Management businessrequirements against TAG functionality, althoughthe offender management team were consulted indeveloping requirements.Offender ManagementThe Offender Management Model launched in May 2006requires a single Offender Manager based in the communityto be responsible for an offender throughout their sentence,whether the sentence is served in custody or the community.They have responsibility for assessing an offender’s risk ofharm and the factors underlying their offending behaviour,planning the sentence, supervising the offender and ensuringthe sentence is carried out. Offender Management has beenrolled out in phases:■■■Phase 1, from April 2005, focused on offendersgiven community orders and licences managed by theProbation Service.Phase 2, from November 2006, focused on offendersassessed as high or very high risk, prolific and persistentoffenders and those serving determinate sentences over12 months. This required probation and prison to deliveroffender management in partnership.Phase 3, from January 2008, applies to those imprisonedfor public protection.Offender management now covers all offenders in thecommunity (235,000 at the end of 2006) and an estimated18,000 (22% of those in custody on 26th September 2008).The scope of future roll-outs is undecided.Source: National Audit Office and NOMS’ Offender ManagementCaseload Statistics8Functional fit assessment of the TAG product at project inceptionPrison RequirementsPartially met (26%)Fully met (73%)Probation RequirementsNot met at all (31%) Partially met (40%)Fully met (29%)0 10 20 30 40 50 60 70 80 90 100Percentage of requirements metSource: NOMS18 THE NATIONAL OFFENDER MANAGEMENT INFORMATION SYSTEM

PART TWO2.9 Despite the results of its own functional fitassessment, NOMS’ formal project documentation failedto reflect the additional risks and costs associated withthe customisation of TAG. The 2005 final business casestated that with additional development and modificationTAG could meet the requirements of probation, prisonsand NOMS. The Home Office Group Investment Boardwas told ‘a requirements gap analysis has found that allneeds can be met ‘out of the box’ or with some degree ofcustomisation that have been incorporated into the costestimate.’ The judgement that TAG would require limiteddevelopment to fit the needs of NOMS was flawed, andwas a key factor in programme cost increases.The handling of business change“At the centre of the programme history is amisunderstanding of the close relationship betweenbusiness processes (and simplification of process) andtechnology. IT was seen as a tool to implement currentprocesses with all their existing flaws.”Accenture 20082.10 NOMS treated C-NOMIS as an IT project ratherthan a major IT-enabled business change programme.The initial project budget of £234 million did notinclude an allocation for business change and an earlyOGC review found that the business capability andfunctionality of C-NOMIS was not fully appreciatedacross NOMS. So, for example, the project team lackedsupport from senior people to persuade business ownersto streamline processes. Staff seconded to the project teamcomprised the main vehicle for involving users in projectdevelopment. Formal communication channels betweenthe business and project team were not established untilMarch 2006 (prisons) and February 2007 (probation).2.11 The impact of a lack of effective engagementwith users was that conflicts were generally resolvedby changing the project rather than adjusting businessprocesses. EDS, in its 2007 programme review, identifiedthat the number of functional requirements that C-NOMIShad to deliver increased by 16 per cent (see Figure 9), andconsidered that the majority of these either clarified theoriginal project or addressed omissions.2.12 In our view, the challenges were not only technical.The C-NOMIS project tried to deliver a complex ITsolution into an already complex and changing businessenvironment, recognised as a key risk in 2005. The projectcosting and approval process assumed that businessprocesses would be standardised and adapted to fit the IT,rather than vice versa, but they were not, in part becausesome processes were prescribed by legal and policyrequirements. And the risks of starting the project at suchan early stage in the evolution of NOMS and the offendermanagement concept were not recognised sufficiently.Contractual arrangements“We recommend that a fixed price agreement isconcluded with EDS at the earliest opportunity and thatEDS provides a statement of scope, assumptions andresponsibilities for the remaining project phases.”Internal Audit 20072.13 Using a contract change notice to an existingframework agreement, the Home Office contracted withEDS for the development, implementation and operationof a C-NOMIS central data service, central infrastructureand support services. By selecting the supplier without acompetitive tendering process, the project team was notrequired to develop a full clear specification which wouldhave reduced subsequent risks. NOMS selected EDSbecause of the urgent timescales, synergies and economiesof scale achievable, since they were responsible forHM Prison Service’s technical infrastructure. A formerproject director told us that, in retrospect, the decision tonot follow open tendering procedures was flawed giventhe level of uncertainty surrounding the project.2.14 Contracting arrangements between NOMS andSteria (NPS’s infrastructure provider under the OffenderManagement National Infrastructure (OMNI) contract),for cleansing and migrating probation data to C-NOMIS,were never completed because of other priorities forNOMS. Thus there was an inadequate contractual vehiclefor a project of this size and complexity. Although Steriaplanned to proceed through a formal change to the OMNIcontract, work instead commenced through a simpleInstruction to Proceed letter, extended twice up to the endof December 2007.91,000950900850800750Number of C-NOMIS functional requirementsNumber of requirements860936Source: National Audit Office analysis of EDS data998July 2006 October 2006 July 2007THE NATIONAL OFFENDER MANAGEMENT INFORMATION SYSTEM19

PART TWO2.15 NOMS did not adequately consider the issue ofintellectual property rights in its initial procurementprocess. The proprietor of the TAG application, Syscon,engaged under a sub-contracting arrangement with EDSto develop the C-NOMIS application, has benefited fromthe value created by the project. It is now marketing animproved TAG product worldwide containing 20 per centnew or enhanced functionality 5 derived from C-NOMIS.As the extent of the customisation of TAG became clearduring 2005 and 2006, NOMS should have reopenedcommercial negotiations to achieve a discount on thecontract or a future revenue sharing arrangement. 6In 2007, the project’s new commercial director startednegotiations with Syscon to achieve a two-year holidayon future support and maintenance costs as a goodwillgesture, valued at around £2 million, but negotiationsfailed to reach a conclusion when Syscon was bought byanother company.Supplier management“The processes adopted by both the authority and EDShave been too rigid and have resulted in long delaysin decisions and implementation.” “Interpersonalrelationships have been allowed to deteriorate tounacceptable levels.”Review of EDS/NOMS relationships byNOMS Non-executive Board Member 20062.16 Initially NOMS selected a time and materialsarrangement for the development of the C-NOMISapplication but the controls and contingencies necessaryto manage such an agreement were not in place. In 2004,the Home Office Group Investment Board raised concernover the arrangement, referring to OGC’s preference for afixed cost regime. Under the arrangement, NOMS held theproject risks but did not mitigate them by either includinga contingency in the budget, as a supplier would underfixed price by adding a risk premium, or employingrobust programme and financial management processessuch as earned value tracking and change control.This shortcoming was highlighted by an internal auditreport in August 2007. The project team assumed theycould ‘concentrate on controlling the areas of high risk ofcost sensitivity’.2.17 For time and materials arrangements to work, trustedpartnership arrangements need to be established betweenclient and supplier. Up until mid-2006, difficult supplierrelations, stemming from a lack of communicationand poor supplier management, characterised projectdevelopment. An OGC report in June 2006 highlightedthe negative tensions and the strong sense of polarisationbetween EDS and NOMS. A joint EDS and NOMS reviewthen concluded that ‘interpersonal relationships hadbeen allowed to deteriorate to an unacceptable level,both management structures have been confusing andhistorically EDS staff were not of the necessary quality.’From EDS’s perspective, NOMS’ perception of EDSsimply as a source of skilled labour presented difficultiesfor them to add value and contributed to their failure toescalate issues and provide high quality staff. A lack ofsupplier representation on the Project Board exacerbatedthe issues.2.18 In the early phase of the project, NOMS had littleability to sanction EDS for late delivery because itscontractual agreement meant that it had to allow EDS anadditional two month period to carry out remedial workbefore it could take any punitive action. In summer 2006,EDS and NOMS negotiated delivery of the pilot releasefor a fixed price. This pilot, and subsequent fixed pricedeployments, were delivered on time.2.19 In our 2006 report Delivering Successful IT-enabledBusiness Change, we identified the need for organisationsto act as intelligent clients. 7 We found organisationswere more likely to succeed if they formed productiverelationships with suppliers, possessed the capacityto offer technical challenge and were clear aboutthe additional skills needed to supplement existingcapabilities. NOMS’ ability to act as an intelligent clientwas hindered through little technical understanding anda reliance on supplier knowledge, with large volumes ofwork outsourced. Poor relationships contributed to projectmilestones not being met.Governance and resource issues“Overall governance of the Programme is not sufficientlyclear”….”the Programme Director’s support organisation isinsufficient at present for such a complex programme.”Home Office Health Check 20075 As measured by the number of functional and non-functional screens, a proxy measure for a full function point count.6 Open Source Software, Guidance on implementing UK Government Policy,OGC, 2002.7 National Audit Office (2006) Delivering Successful IT-enabled Business Change, report by Comptroller and Auditor General, HC 33-I Session 2006-07.20 THE NATIONAL OFFENDER MANAGEMENT INFORMATION SYSTEM

PART TWOGovernance2.20 Operating within the newly created NOMS entitypresented difficulties for the project. NOMS was itself amerger of two bodies with contrasting cultures. The overallgovernance and resources applied to the project, however,were not adequate given the scale of the task. Rolesand responsibilities were blurred; there was inadequateoversight by senior management; and insufficient skilledresource was applied to the project. Getting theseelements right is key to the success of high risk projects. 82.21 Governance structures changed a number of timesthroughout the project. Figure 10 overleaf shows thegovernance structure for the project at its inception.There were several issues.■■■■The project was too large for the structure.The Project Board had to represent so many differentinterests that it became too large to provide effectiveoversight. For example, between October 2005and October 2006, 39 different people attended.The minutes show the Board did not actively monitordelivery and did not discuss programme financesuntil April 2007. Findings from an internal audit ofthe project in April 2006 were not discussed, andOGC reported ‘insufficient real thinking time atthe board.’Roles and responsibilities were not clearly defined,for example, between the Project Board, theBusiness/Technical Support Group and the projectteam. The C-NOMIS project team was entrustedwith the day-to-day running of the project andcoordination across the various workstreams (whichgrew to 40).Critically, the part of NOMS administering thefunding, the Offender Information Service (OIS),and the project team did not work together. Thisdisjunction contributed to the lack of budgetmonitoring: the project team thought OIS wasmonitoring expenditure and vice versa.A vacuum of leadership contributed to the confusedresponsibilities between the project team, OIS andNOMS business areas, presenting challenges forboth suppliers and the project team. The ProjectBoard had little contact with the project team andthe Senior Responsible Owner (SRO) had otherfull-time responsibilities. As a result a number ofmajor decisions on the scope and timing of systemreleases were taken without proper authority.2.22 In September 2006, steps were taken to reorganisethe SRO’s responsibilities to allow more time forC-NOMIS. The project was also partially restructured asa programme with a programme director overseeing theproject director although, in its August 2007 review ofcurrent governance structures, internal audit still reportedthat ‘there was a lack of clarity of roles and responsibilitiesof the overall programme management team.’2.23 NOMS and its parent Department should have takena more active role in monitoring C-NOMIS. The Boardrelied on the Project Board to escalate issues to it throughNOMS’ routine reporting system covering all projects inits Change Programme. This system, however, consistentlyrated the project as “green”, meaning that it appearedto be proceeding satisfactorily to time and budget. Asa result, despite the project’s strategic importance andsize, the NOMS Board did not ask for or receive specificproject progress reports from the SRO until August 2006,and so were unaware of arising issues. The significant riskof further delays highlighted by OGC in July 2006 wasalso given insufficient attention by the NOMS Board.2.24 Throughout the project there were a number ofcritical external and internal reviews, but messages did notalways get through and reviews did not have the impactthey should have had:■■■OGC recommended planning be strengthened in2005 and again in 2006 but, in July 2007, found thata C-NOMIS plan, linking resources and key deliverymilestones, remained incomplete.Despite OGC recommending that an independentfull-time time assurance resource be involved withthe project in 2004, the lack of such a resource washighlighted by internal audit three years later.Between October 2004 and July 2006, OGCperformed three Gateway reviews, two of whichreceived a red RAG 9 status. Findings from thesereviews were not always acted upon.NOMS did respond positively to other recommendations.Following OGC recommendations in July 2006 and a jointsupplier and client review in late 2006, a new integratedpartnership relationship was introduced between NOMSand EDS.8 National Audit Office, Driving the Successful Delivery of Major Defence Projects: Effective Project Control is a key Factor in Successful Projects,HC 30 Session 2005-2006. See also The Cancellation of the Benefit Card Project, National Audit Office HC 857 Session 1999-2000 and Improving ITprocurement: The impact of the Office of Government Commerce’s initiatives on departments and suppliers in the delivery of major IT-enabled projects,National Audit Office HC 361-1, Session 2003-2004.9 Gateway reviews use a red, amber, green notation to indicate whether a project is ready to move to the next stage.THE NATIONAL OFFENDER MANAGEMENT INFORMATION SYSTEM21

PART TWO10C-NOMIS governance structureNOMS BoardOverall responsibility for NationalOffender Management ServiceSRO (OASys)Overall responsibility for OASysSRO (C-NOMIS)Overall responsibility for C-NOMISC-NOMIS Project BoardMeets every two monthsStrategic decision makingAgreement of project approach and key deliverablesEnsures goal congruence with Home Office and NOMSOISTechnical arm ofNOMSHas input in allIT projectsOASys ProjectParallel project to developOASysBusiness/Technical Support GroupMeets monthlyScrutinising of key project documentationMonitoring of project activityC-NOMIS ProjectC-NOMIS TeamDay to day running of the C-NOMIS projectReports toHas input inSource: NOMSSkills and resources2.25 The role of the SRO is to take overall responsibilityfor project delivery and communicate to seniormanagement project progress and risks. 10 To date, threedifferent SROs have been responsible for C-NOMIS andNOMIS. The first SRO, in post until March 2007, had littleexperience of managing large IT projects. In 2006, OGCrecommended the SRO spend more time on the projectto instil greater support and credibility and NOMS tookaction to provide some additional support.2.26 Since inception, C-NOMIS has also faced challengesrecruiting suitable staff. There was no communicationsmanager until December 2005, no dedicated financialcontroller until February 2007 and by April 2007,almost three years into the project, key posts remainedunfilled. The challenges in recruiting staff stemmed fromprogramme headcount restrictions, lack of available skillswithin both NOMS and the Ministry of Justice and delaysin obtaining security clearance. The lack of civil servicerecruits meant greater reliance was placed on moreexpensive contractors.Project management“There is no overall NOMIS Programme plan with keydelivery milestones linked to key resources”. Home OfficeHealth Check 2007 “I knew we had problems when Iasked for the budget and was told there wasn’t one”.New SRO 200710 Managing Successful Projects with PRINCE2 OGC, The Stationary Office.22 THE NATIONAL OFFENDER MANAGEMENT INFORMATION SYSTEM

PART TWOPlanning2.27 The C-NOMIS project’s poor governancearrangements coupled with headcount constraints andskill shortages led to inadequate planning. ManagingC-NOMIS as an IT project required the Project Board tocoordinate up to 40 individual workstreams, somethingit delegated to the project team. While the workstreamshad their own plans, a report by Internal Audit found eachwas operating as a functional silo. From our analysis of theProject Board’s minutes, we found no evidence of a longterm plan drawing together these separate streams beingdiscussed until July 2006. This plan constituted a limitedschedule covering the tasks over three months to deploythe pilot software. Our analysis of external project reviewsshowed planning weaknesses were a consistent area ofconcern (see paragraph 2.24).2.28 Alongside the poor planning of deliverables, ourassessment of the plans showed poor scheduling ofkey management activities. We found no evidence thatactivities such as quality reviews and stakeholder eventswere scheduled in line with good practice guidance.When governance arrangements changed from those of aproject to a programme in 2007, NOMS did not producedocumentation showing the organisational structure’sbusiness processes, people, information systems and dataas recommended in OGC guidance. 11Financial monitoring and cost estimation2.29 The lack of a dedicated financial team before 2007led to significant cost components being underestimatedor omitted from the initial project cost estimation of£234 million, poor cost control during the project, anduncertainty over the extent of cost escalations:■ Poor initial cost estimations. Management andsupport costs were underestimated by £21 million,and the costs of live service support by £115 million.The lack of contingency funding was unrealistic,since NOMS bore all of the project risks under atime and materials-based contract. In 2004 OGChighlighted the failure to analyse cost estimates andthis warning was reinforced by the Home Office’sGroup Investment Board in 2005 which felt costshad not been fully explored.■Failure to understand VAT implications:The C-NOMIS project proceeded under an incorrectassumption that VAT was recoverable, despite aGroup Investment Board recommendation thatfurther advice be obtained. HM Revenue andCustoms ruled, in January 2006, that project VAT■was irrecoverable as offender management crosseddifferent organisations, including the private sectorand charities, and supplier contracts referred towork outside HMPS and NPS. A VAT liability of£99 million, including £17.5 million already spent,increased lifetime costs. A more recent ruling, inrelation to contracted-out prisons only, has reversedthis position (see paragraph 3.14).Lack of financial controls: Before the moratorium,forecast costs were not actively managed and costcontrol focused on monitoring spending against theannual budget rather than matching cost againstproject deliverables, which would have highlightedcost increases and delays. A failure to monitor costsagainst progress meant the cost impact of delaysidentified in June and July 2006 was not appreciated.The two year project delay added a total of£115 million. Prior to 2006-07 the ability to controlcosts was hampered by booking all costs to a singlecost centre. This situation was compounded byunclear financial accountability and the disjunctionbetween OIS and C-NOMIS (see paragraph 2.21).Change control2.30 At the start of the project, controls over changesto requirements were weak and informal. In June 2005,an OGC Gateway Review highlighted the need tostrengthen these controls, but a Change ControlBoard did not become fully operational until early2006. Although a formal change control strategy wasin place from August 2005, in line with PRINCE2methodologies, internal audit found the strategy was notroutinely followed.2.31 No process for assessing the cumulative impactof individual change requests on project budget ordelivery timetable was in place. Our review of the boardminutes found large change requests being discussedand approved collectively by board members, but noassessment of the cumulative effect. In addition, untila change database was introduced in April 2006, therewas inadequate recording of individual changes and theircosts, such that internal audit could only estimate thatthere had been between 300 and 400 change requestsin 2005. The extent to which these were properly costedbefore being actioned is not known. Since April 2006there have been a further 400 change control requests.EDS found over one-third of change requests werecomplex and had a material impact on the scope ofC-NOMIS requirements.11 Managing Successful Programmes, OGC, The Stationary Office.THE NATIONAL OFFENDER MANAGEMENT INFORMATION SYSTEM23

PART THREEProgress since themoratorium in 20073.1 In August 2007 the Minister of State put amoratorium on the C-NOMIS project as the 2007-08projected expenditure of £111 million exceeded the£69 million budget. This Part covers:■■■what the revised NOMIS programme aims to deliverand progress so far;the impact of delays and re-scoping on the costsand benefits; andNOMS’ fitness to deliver.The rescoping process3.2 The moratorium stopped further softwaredevelopment, testing and the migration of data recordsfrom legacy IT systems. NOMS then undertook a strategicreview to identify a suitable and affordable approach.The challenge was to deliver the essentials necessary tomeet the original goals of the project in an affordable wayand meet its funding constraints.3.3 We found that the strategic review followed arobust approach by objectively comparing options inline with industry practice (see Appendix 2). The reviewteam assessed seven different options using a balancedscorecard considering strategic alignment, value andbenefits, achievability and risks. The lifetime costs ofthe options, including the Do Nothing option, rangedfrom £513 million to £657 million, based on high levelestimates provided by suppliers. NOMS rejected the DoNothing option because it estimated that the inefficiency,high cost and risk of business failure associated withlegacy systems would cost £543 million over the periodto 2020.3.4 In December 2007, the NOMS Board selected thelowest cost option (Option A in Figure 11) at a cost of£513m. The Board considered the balance of costs andbenefits (Net Present Value) against the actual fundingavailable and selected what it believed to be the bestand most affordable option and which had the secondhighest net present value of the options considered. TheBoard considered but rejected the option of deliveringa slimmed down version of C-NOMIS to prisons andprobation as unaffordable, although recognising that itprovided the best strategic fit and highest net presentvalue. The C-NOMIS option (Option D in Figure 11) wasthe only option which would provide a single database ofoffender records, but some in NOMS viewed continuingwith C-NOMIS as high risk because of the potentialdemand for additional customisation and problemsadapting business processes to the new system.3.5 The technical solution that NOMS is now pursuinghas evolved since the NOMS Board approved the outlinebusiness case in December 2007. At that time the optionwhich included a national probation case managementsystem (PCMS) and data share was estimated to cost£627 million (Option G in Figure 11). During 2008, thePCMS project team evaluated in detail two packagesalready in use by some probation areas for a nationalsystem and selected the Delius system for its better fitwith probation requirements and lower technical risk.Since then, the project team has found ways to stripout some costs to keep the solution within the overallfunding envelope, principally by transferring probationarea records into a single national database without thedata cleansing anticipated in Option G. NOMS expectsthat, over time, single offender records will be createdas probation areas gradually remove duplicates andmerge records.24 THE NATIONAL OFFENDER MANAGEMENT INFORMATION SYSTEM

PART THREE3.6 The NOMIS programme has already implementedthe local version of the Delius system in probation areaswith an urgent need to replace existing systems so, byApril 2009, around one-third of probation staff willalready be familiar with it.Continuing risks and consequences ofthe rescoping exercise3.7 The re-appraisal of the approach to probation casemanagement systems has led to delays in commencingother elements of the NOMIS programme. A delivery datein summer 2008 was planned for Data Share, althoughit has not yet been delivered, and March 2010 for anOASys 12 upgrade. The NOMIS programme business caseshould have been presented to the NOMIS ProgrammeBoard in July 2008, but presentation is now likely totake place in Summer 2009, following the conclusionof contractual negotiations for all projects withinthe programme.3.8 The development of C-NOMIS for prisons hasresumed, although there are currently some delays withsome 60 per cent of deliverables up to two monthslate. NOMS has calculated the cost of slippages to be£1.18 million. Both the supplier and programme team,however, are confident of delivering the full application inApril 2009 in order to commence roll-out.3.9 The National Delius option is still at the earlyanalysis stage, and developing a new solution (even onebased on an existing application) could prove to be morecomplex than anticipated. In addition, successful deliverydepends upon progress being made by Steria, under theOMNI transformation project (itself held up by the delaysto the C-NOMIS project) to house the data and developaccess and search facilities.3.10 Internal audit has already raised concerns about theability of the programme to meet the cost models giventhe reliance on unvalidated cost figures and the low levelof contingency. At this early stage of development ofDelius and Data Share, it would be prudent for NOMSto allow 40-50 per cent contingency instead of the24 per cent in the budget.11Comparison of the estimated costs and benefits of three of the seven options considered by the NOMIS ProgrammeBoard in December 2007Option A – The least cost optionPrisons: C-NOMISProbation: replace legacy systemswith CRAMS case by caseCustom Data ShareOASysOption D – The slimmeddown C-NOMIS option 1Prisons: C-NOMISProbation: C-NOMISOASysOption G – The option most similarto current way forwardPrisons: C-NOMISProbation: National DeliusCustom Data ShareOASysStrategic alignment Good Very close GoodBenefits Good Strong StrongNet present value £27m £170m £0Achievability risk Medium risk Medium-high risk Medium-high riskLifetime costs £512.9m £563.5m £627mNOMS’ overall assessment Chosen option Best business case butunaffordable and high riskRejected as unaffordableSource: NOMSNOTE1 This version of C-NOMIS excluded interfaces with other parts of the criminal justice system, contracted-out prisons and some existing probation service functions.12 OASys improves the understanding of offenders’ risks and needs through providing a structured assessment tool. Currently, the legacy HM Prison ServiceOASys and NPS e-OASys are separate IT software applications although there is some facility to share assessments on a single individual.THE NATIONAL OFFENDER MANAGEMENT INFORMATION SYSTEM25

PART THREEAssessment against NOMS’original aims3.11 The changes to the NOMIS programme since 2007means that the original aims of the project are unlikely tobe met in full without future enhancements (Figure 12).Under the revised programme, the scope of the systemhas changed from a single database to separate casemanagement systems for prison and probation. As aresult, some of the benefits initially envisaged, such asthe support for end-to-end offender management andimproved information sharing, have been compromised.Appendix 3 shows the information flows between prisons,probation services and other agencies at present, asplanned under C-NOMIS and under current plans.3.12 Published in 2006, the Offender Management Modelidentified the facility for offender managers to read andwrite to a single case record in real time as an ‘absolutepriority’. Implementing the Model without the supportingIT has led to interim solutions and adaptation. Through ourdiscussions with stakeholders we identified the followingchallenges to offender management which the currentsolution presents.■■■Where the offender is in custody, it is moredifficult for offender managers to retain controland responsibility over offenders without sight ofthe interventions available or undertaken, sentenceplans and the factors impacting an offender’srisk assessment.The risk that offender managers do not have fullinformation about those they are managing remains.Although it will be much easier for offendermanagers to find out where and when particularoffenders have been dealt with before and theircurrent risk profile, obtaining detailed informationheld by the prison service will still require telephoneand email enquiry and manual amalgamation ofrecords. Obtaining information from other probationareas should improve particularly if, over time,probation areas remove duplicates and mergerecords in the National Delius Database to createsingle offender records.Some administrative features supporting offendermanagement are not available, such as theability to have real time access to offenderdischarge information.3.13 At the outset the Criminal Justice IT fund providedfunding to the C-NOMIS project to support informationexchange and allow courts, Police and Inspectoratesaccess to data. It was estimated that such access wouldachieve efficiency savings of £7 million per annumand improve offender assessments. Under the revisedprogramme, this information exchange will no longerbe fully supported. Offender information will not beshared more widely across the criminal justice system, forexample with drug and education providers and, voluntaryorganisations and the Parole Board. The importanceof efficient and effective information sharing has beenoutlined in a number of reviews (see Appendix 4) andraised during our stakeholder discussions.3.14 The exchange of information between contracted-outprisons and state prisons is important to the streamlinedfunctioning of the prison system as a whole. NOMSexcluded contracted-out prisons from the revisedprogramme because it did not have provision within itsbudget to pay the VAT which providing services to theprivate sector attracts. NOMS has now resolved this issuewith HM Revenue and Customs who have agreed that theuse of Prison NOMIS by the 11 contracted-out prisonswould not constitute business use on their part; henceNOMS will be able to continue to recover VAT incurred ifit implements Prison NOMIS in contracted-out prisons.3.15 Partners, such as the Parole Board, the Youth JusticeBoard and the Learning and Skills Council either havealready or are in the process of developing their owndatabases at their own expense. Separate systems increaseduplication, the risk of error and may present difficultiescreating future interfaces. We have been told there areMinistry of Justice standards to influence technical systemdevelopment but these are not well-communicated acrossthe Department.NOMS is currently considering the roll-out of offendermanagement to a wider range of offenders which mayaffect how the Model is applied in future.26 THE NATIONAL OFFENDER MANAGEMENT INFORMATION SYSTEM

PART THREE3.16 Despite the reduction in scope of NOMIS, therevised programme should deliver a number of benefits toprisons and probation:■■Alongside replacing the ailing prison casemanagement system with a more useable andmodern database, NOMIS provides additionalfunctionality and brings together previously disparateprison systems, such as visitor booking and prisonerexpenses. Most importantly, it introduces a singlecase management database across all prisons tosupport the more efficient and secure exchange ofinformation between institutions.All probation areas will receive a commoncentralised version of a redeveloped casemanagement system, National Delius, to replacethe seven systems currently in place. A commonsystem will allow probation areas to exchangeinformation more efficiently through the ability toperform national offender searches and a commondata format.■■Going forward, the new national system forprobation provides the potential for data sharingwith prisons.The development of a single, centralised offenderrisk assessment system across both prisonand probation will allow the more efficientexchange of information relating to the risksposed by the offender, supporting end-to-endoffender management.12Comparison of original project aims against revised programme scopeObjectives of the C-NOMIS project (2004)Objective 1: improve positive offender outcomes(i.e. reduce re-offending) through end-to-endoffender management across probation and prisonservices and including providers from the privateand voluntary sectors.Whether objectives likely to be met under revised programmeLikely to be met in part. Shared risk assessments (through OASys) will provideup-to-date information on offender risks, but there are significant barriers toend-to-end offender management:■ for offenders in custody, communication barriers are likely to significantlyimpede offender management: sentence planning must take place by phone,email or post; and■ private and voluntary sector providers have no access to offender systems.Objective 2: introduce more assertive casemanagement and contestability in service providersto improve outcomes and value for money.Objective 3: integrate IT support across custodyand community sentences to facilitate the breakingdown of barriers between prison and probation.Objective 4: improve means of monitoringcompliance with National Standards: includingmore demanding community sentences and greatercontrol and oversight of persistent offenders.Uncertain. Commissioning of services in the community by outside providers isstill at an early stage. It is not clear how C-NOMIS was intended to be used tomeet the objective.Uncertain. NOMS did not explicitly state how integrated IT systems wouldfacilitate the breaking down of barriers between prison and probation but,with the development of separate offender management systems for eachservice, it is unlikely that the revised programme will, on its own, affect therelationship between the two services.Likely to be met in part. National case management systems for probationand prisons should assist monitoring once planned management informationsuites are built.Source: National Audit OfficeTHE NATIONAL OFFENDER MANAGEMENT INFORMATION SYSTEM27

PART THREEImpact of the delays and rescopingon costs3.17 At the end of October 2007, £161 million had beenspent on the project overall. We have not been able toascertain precisely what this money was spent on becauseNOMS did not record expenditure against workstreambefore July 2007, but we believe most was spent ondeveloping the C-NOMIS application and testing it inprisons. When rescoping the programme, NOMS soughtto use what it had already paid for, where possible, tominimise wastage. Nevertheless, the overall impact ofthe project delay and change in project scope has ledto additional costs or wasted expenditure in a numberof areas.■■■■■Capital expenditure on software development andhardware which is no longer required under therescoped programme. For example, money paid tocontractors to develop C-NOMIS for probation isprobably in the region of £13.5 million, and NOMShas written off £150,000 of expenditure on trainingsuite equipment it is unable to reuse.Additional costs incurred by EDS after themoratorium, amounting to £1.2 million, to closedown work in hand.NOMS’ own costs in managing the developmentof C-NOMIS for probation and in rescopingthe programme, which NOMS estimates as£1.45 million.Additional expenditure incurred by the business asa result of the delay in replacing worn out legacysystems, for example the cost of additional stafftime in carrying out procedures when IT systems aredown. Based on information supplied by NOMS, weestimate the total cost of delay in implementing newsystems to be around £17 million.Efficiencies foregone as a result of the reducedprogramme scope. For example the InterimAccredited Programme System (IAPS) used acrossprobation to monitor programme attendance andcompletion will no longer be replaced. NOMSestimated that incorporating IAPS functionalityinto the single database would achieve efficiencysavings of £3.4 million a year or £10.2 million overthree years. 133.18 We estimate the financial impact of NOMS failureto deliver the original C-NOMIS programme on timeto be at least £41 million (see Figure 13). This figure is,however, likely to be an underestimate because it doesnot include expenditure on the programme incurredthrough inefficiency and mismanagement, for examplethrough poor supplier management. Nor does it includethe wasted effort and value of benefits foregone across thecriminal justice system because interfaces with C-NOMISwill no longer be built (see paragraph 3.15). For example,the Learning and Skills Council estimates that it wasteda year pursuing solutions for an offender learning recordsystem capable of joining up with C-NOMIS.NOMS’ fitness to deliver3.19 Following the moratorium, NOMS and, in particular,the NOMIS programme team reviewed what went wrongwith the original project and worked hard to address pastweaknesses and put in effective structures going forward,although further work is needed in some areas.■Changes to the governance and leadership of theprogramme have led to a greater focus on deliveryand improved supplier relations and reporting lines.The current Senior Responsible Owner, appointedin April 2008, has previous experience of deliveringlarge, time and materials-based IT contracts. Thedirect relationship between client and suppliercreated through EDS’s presence on the NOMISprogramme board has led to both parties beingbetter informed and a better relationship.■ Commercial arrangements introduced with the newcontract change notice for Prison NOMIS are animprovement. For the first time there are penaltiesfor late delivery, albeit only up to five per cent ofamounts invoiced. There is also a prescribed changecontrol process. At EDS’s request, NOMS agreeda further time and materials contract because ofaccounting implications for the supplier. Both EDSand senior NOMIS staff say they are confident thepartnership needed to deliver this type of contractis now in place. In August 2008 the first monthlycontract management meeting was held to discussissues arising from weekly checkpoint meetings at amore senior level and measure supplier performanceagainst joint plans.13 The National Delius Database, when implemented in 2011, will go some way towards improving efficiency by reducing the rekeying of data.The £10.2 million has been reduced to £8.2 million in Figure 13.28 THE NATIONAL OFFENDER MANAGEMENT INFORMATION SYSTEM

PART THREE■ Financial controls have improved significantlyfollowing the appointment of a programmefinancial controller, a dedicated finance managerresponsible for Prison NOMIS, regular reportingto the programme board, and clearer budgetaryauthority lying with the programme. Our reviewof the current financial tool, however, showed thatit was not sufficiently sophisticated to show thecost implications of potential events and decisions;key cost drivers were not properly reflected; andno sensitivity analysis had been undertaken sinceSummer 2007. There is a gap in financial controlsbetween the Project manager and supplier asinformation such as expected cost and time tocompletion is missing. Without this information,the Project manager cannot identify where moneyhas been spent or what has, or is forecast, to costmore than planned. A way of matching spendingagainst deliverables, for example, earned valueanalysis, is not yet established although NOMS ispursuing this in its current contract negotiations withpotential suppliers.■The programme now has in place key projectmanagement staff, although challenges in recruitingstaff to some posts (as described in paragraph 2.26)have meant that some project components are notbeing provided the necessary support, contractingcosts have increased, there have been problemsdrawing together NOMIS programme componentsand delays in completing benefit realisation workand communication plans. As at 1 July 2008,there were 137 posts, of which 20 (15 per cent)were vacant and 48 (35 per cent) were filledby contractors.■ Business change: The lack of a common definedway of working across 42 locally managed probationareas still presents challenges for the introductionof a single case management system. AlthoughNational Standards 14 provide a high level standardfor managing offenders, there is no common nationalway of working. Currently, there is no budgetavailable for probation business change. Managingthe introduction of a national probation casemanagement system as a technical solution ratherthan a business change programme risks repeatingproblems previously experienced.13Cost implications for NOMS of delays andchanges in scopeExpenditure on probation elements of C-NOMISno longer required 1Development of probation code for combinedC-NOMIS system by SysconEDS costs on probation work no longerto be implementedDesign of data migration tools for probationservice by SteriaCost impact(£m)4.14.64.8Investment in training suites and hardware written off 0.15Sub-total 13.65Costs relating to the moratorium 2EDS demobilisation costs 1.2Sub-total 1.2NOMS’ internal costs 2NOMS’ management of probation elementsno longer required0.7Local probation area preparation for C-NOMIS 0.5Rescoping exercise 0.25Sub-total 1.45Costs stemming from delays and benefits foregone 3Impact of delays on the OMNI-T programmecovering probation infrastructure6.5Cost to the business from failure of existing systems 10.6Cost of reduced efficiency from not replacing IAPS 8.2Sub-total 25.3Total 41.6Source: National Audit Office analysis of NOMS’ and supplier dataNOTES1 Actual expenditure incurred by NOMS.2 Estimated expenditure based on NOMS’ and supplier estimates.3 Forecast cost implications estimated from NOMS’ benefit realisationwork after allowing a 20 per cent discount for optimism bias.14 National Standards, set by NOMS, dictate how offenders should be managed on a daily basis, for example, by detailing the timescales for completing risk ofharm assessments.THE NATIONAL OFFENDER MANAGEMENT INFORMATION SYSTEM29

APPENDIX ONEStudy scope andmethodology1 This study set out to examine the value for money ofNOMS’ management of the NOMIS programme, focusingon the events that led to the demise of the originalNOMIS programme and how far the revised programmeannounced in early 2008 meets the objectives of theinitial programme.2 This section outlines the research methods used overthe course of our examination, which took place betweenJuly and September 2008.Review of programme documentation3 We reviewed programme documents from theinception of NOMIS to the latest documentation outliningthe way forward including business cases, contracts andlegal documents, NOMS and NOMIS board minutes,planning documentation such as project plans, benefitsrealisation strategies and governance charts, andrequirements gap analyses. We used the documentationto assess: the soundness of the initial decision makingprocess; the adequacy of cost and programme monitoringarrangements; and progress against key milestonesand level of senior management involvement. Whereappropriate we used OGC and NAO guidance to providea framework for our analysis.Programme technical analysis4 We contracted Accenture, a global managementconsulting and technology services company, to providetechnical IT expertise throughout our fieldwork. Accentureassessed the original 2004 and 2007 rescoped NOMISprogramme from a technical perspective and reviewedthe accompanying cost models. In particular we askedAccenture to conclude on the technical feasibility ofthe initial and revised NOMIS solution and whetherthe revised programme made best use of technicalachievements to date.Semi-structured interviews withprogramme staff5 We interviewed some 45 key individualsresponsible for programme delivery, including the seniorbusiness owners and project managers for each of thefour elements of the programme, technical specialistsand finance representatives. We also spoke to formerprogramme team members, to understand the reasonsfor past decisions, and suppliers, to obtain a balancedview of the issues faced by the programme. Through theseinterviews, we aimed to develop a clear understanding ofthe issues encountered by the programme and how theywere addressed.Stakeholder consultation6 To gauge the level of consultation carried out by theNOMIS team and assess how NOMIS fitted stakeholderneeds we visited three probation areas and two prisons.The probation areas visited (West Mercia, Surrey andDevon and Cornwall) were selected based on their currentcase management system. Our prison visits (HMP Fordand HMP Winchester) covered an open and a Category Bprison. Towards the end of fieldwork, we held a numberof telephone interviews with staff at HMP Camphill onthe Isle of Wight, which had been part of the C-NOMISpilot. These interviews improved our understanding ofC-NOMIS functionality, how well it fits business needsand whether the roll-out was well managed.7 We held two focus groups with representatives fromthe Learning and Skills Council, HM Court Service, theYouth Justice Board, the Howard League and NACRO.Through these sessions we assessed information needsacross criminal justice bodies, the value of the NOMISproduct to these stakeholders, and how well-informedthey were of the progress and changing outputs ofthe programme.30 THE NATIONAL OFFENDER MANAGEMENT INFORMATION SYSTEM

APPENDIX ONEQuantitative analysis8 We collected and analysed a range of financialdata from NOMS including cost models, budgetarydocumentation and internal cost increase analyses. Weanalysed in detail the financial tool behind the costs forthe revised NOMIS programme, actual costs incurred andfinancial reports presented to the NOMIS Board showingextent of cost increases. This analysis helped us drawconclusions on the appropriateness of financial control inplace for a project of the size and scale of NOMIS.9 We analysed non-financial data collected from theprogramme team relating to staff to understand the level ofstaff vacancies and the split between consultancy and civilservice staff.Review of secondary data10 The NOMIS programme has been subject to anumber of independent reviews by internal audit, theOffice for Government Commerce and, at the request ofthe programme team, external consultants. We undertooka detailed review of their findings to gain a third partyperspective on the NOMIS programme and assess howwell NOMS responded to recommendations made byindependent reviewers.THE NATIONAL OFFENDER MANAGEMENT INFORMATION SYSTEM31

APPENDIX TWOTechnical ReportManagement SummaryThe programme to deliver end-to-end offendermanagement across Prison and Probation has sufferedfrom cost overruns, scope changes, delays, restructuringand insufficient governance. Our task in this report,focusing on certain technical aspects highlighted by theNAO, is to identify causes and illustrate any risks for thefuture success of the NOMIS programme. Overall, therehave been weaknesses in technical decision making, andmanagement and commercial process.iiiiiA single database is not required to provideend-to-end offender management, but it does givethe best real-time view of offender information.Separate technical solutions with a managementinformation system reporting solution is a feasiblealternative to a single database solution andcould have been easier, cheaper and deliveredbenefits sooner. This alternative was notsufficiently considered.Technical review of the2004 programmeAt the centre of the programme history is amisunderstanding of the close relationship betweenbusiness process (and simplification of process) andtechnology. IT was seen as a tool to implement currentprocesses with all their existing flaws. By not changingbusiness requirements to fit the solution where possible,the package became unaffordable to implement andeventually required a radical re-scoping.1 The assumption that TAG 15 would be an off-the-shelfsolution with minimal customisation was invalid forProbation and to a lesser extent for the prison service, andit had not been proven on the scale required for the UK.This was the main reason for delay and cost increases.2 In assessing the technical solution of a singledatabase for end-to-end offender management weconclude that:iA single database solution is realistic and doable,and similar successful solutions exist in the publicsector (e.g. the National Insurance RecordingSystem), but the programme did not appreciate,and therefore plan for, the complexity of doing thisfor separately governed organisations with separatetechnical infrastructures.3 Extending an existing contract saved time andmoney because it allowed a quick start with a supplierthat understood the current situation in prisons, butthe time and materials agreement meant there was nofinancial incentive for the supplier to deliver withinbudget, and the programme did not have controls orcontingency to manage this. As a result, managementlacked early warning of programme failure, and we didnot find evidence of the supplier alerting NOMS that therewere problems.4 High level technical requirements werecommunicated at the outset to suppliers, but it isnormal for requirements to get more detailed duringdetailed analysis. The programme should have budgetedcontingency to account for the normal process ofrequirements clarification, and re-estimated costs for adecision to proceed after detailed analysis. This couldhave identified the financial and time delay impact ofproduct customisation by late 2005 instead of 2007 whenthe financial problems were first escalated.15 TAG is the name of the chosen commercial off the shelf package provided by Syscon.32 THE NATIONAL OFFENDER MANAGEMENT INFORMATION SYSTEM

APPENDIX TWOTechnical review of the 2007NOMIS ProgrammeThe revised 2007 programme is broadly on track, butthe scope of information sharing between Prison andProbation has been reduced to the minimum data set(e.g. offender identifiers, location, offender manager,sentence details and dates).5 The analysis to select the revised programme wasa good approach, providing an objective comparisonof seven options aligned with industry practice. Nowthat the programme has spoken to stakeholders at theworking level, the Probation and Data Share componentsof the approved solution have evolved since the outlinebusiness case. We suggest that the National Deliuswith Data Share option now being assessed (previouslyconsidered unaffordable by the analysis) should beanalysed with the same degree of formality together withstakeholders, which means comparing with all previouslyrejected options.6 Past technical achievements for Prison NOMIS arebeing used in the revised programme, but the Probationcomponents which have been built for C-NOMIS will notbe used, so the investment in this to date will be lost.7 The solution being assessed is technically realisticand achievable, but there is an assumption that the 42Probation areas will accept a common national solution,which will require a common way of working. Since the42 areas work in different ways today, it is important todefine and agree a common national way of workingfor Probation before building the technical solution.A technical solution to meet all of the different Probationarea requirements is unrealistic, and would result insignificant delay and cost overrun.A review of the cost estimatesThe cost estimates in 2005 were fundamentally flawed.The approach for 2007 is stronger, but risks still remainsuch as uncertainty in the estimates for rebuilding Deliusand developing a bespoke Data Share given the earlystages of planning.9 The 2005 cost estimate of £197 million considerablyunder-estimated the extent of customisation requiredwhich resulted in the programme taking two yearslonger than expected, and this is the main factor in thecost estimate increasing to £597 million 16 by 2007.In addition, the lack of a dedicated finance team resultedin costs being missed entirely (e.g. Management andSupport costs of £21 million).10 The 2007 Cost Model needs to be revised in thenew business case to reflect the changes in the Probationsystem and Data Share approaches which are both inan early planning phase so the cost estimates will beuncertain. We recommend allowing 40-50 per centcontingency to mitigate this risk, and then when therequirements are clear after detailed analysis re-estimatethe cost and negotiate a fixed price.Accenture Technical Report, September 20088 The data share component of the proposed NOMISsolution delivers far less than was originally envisagedfor end-to-end offender management across prison andProbation. We have concerns that the planned tacticalsolution is likely to require additional funding in thefuture to replace or improve it, and this should be aconsideration in the full business case.16 Figures do not included costs of capital and depreciation.THE NATIONAL OFFENDER MANAGEMENT INFORMATION SYSTEM33

APPENDIX THREEOffender informationexchange pre andpost NOMISPre NOMIS – key offender information exchangeHMI PrisonsHM Courts ServiceNOMSHM Prison ServicePrison to PrisonParole BoardYouth Offending TeamsHMI ProbationNational Probation ServiceProbation to Probation areaIntervention providersDrug Action TeamsVoluntary SectorLearning and Skills CouncilHousingC-NOMIS – key offender information exchangeHMI PrisonsNOMSHM Courts ServiceHM Prison ServicePrison to PrisonParole BoardYouth Offending TeamsHMI ProbationNational Probation ServiceProbation to Probation areaIntervention providersDrug Action TeamsVoluntary SectorLearning and Skills34 THE NATIONAL OFFENDER MANAGEMENT INFORMATION SYSTEM

APPENDIX THREERevised NOMIS – key offender information exchangeHMI PrisonsHM Courts ServiceNOMSHM Prison ServicePrison to PrisonParole BoardYouth Offending TeamsHMI ProbationlimitedNational Probation ServiceProbation to Probation arealimitedIntervention providersDrug Action TeamsVoluntary SectorLearning and Skills CouncilHousingCore offender information supplied via paper or telephoneCore offender information supplied electronicallyTHE NATIONAL OFFENDER MANAGEMENT INFORMATION SYSTEM35

APPENDIX FOURExternal reviewrecommendations oninformation exchangeSource Findings relating to information exchange 1 NAO analysis of progressHanson and White: Independentreview of serious further offence,HM Inspectorate of Probation,February 20061) The transfer of information between Probationand third party suppliers is crucial in thesupervision of offenders.2) OASys is fundamental in identifying highrisk offenders.3) Recording all case details in a centraldatabase is key to successful supervision.Transfer of information to third parties is limitedunder the revised NOMIS programme.Prison and probation OASys will be replacedwith a single offender assessment system, butmanagers will not be able to access up-to-datecase records.Rice: Independent reviewof serious further offence,HM Inspectorate of Probation,May 20061) A lack of information on an offender’sprevious sentence caused misjudgements.2) The identity of the Offender Manager andwho had lead responsibility was confused.3) As NOMIS enables better co-ordination ofinformation and visibility of assessments,it is essential.The data share strand of NOMIS should makeclear the designated offender manager, ashared offender assessment system will improvevisibility of information but the lack of singledatabase will mean information on previoussentences is not shared.Report of the Zahid Mubarekpublic inquiry, June 2006Paper-based records resulted in insufficientinformation being transferred. This increasedthe risk of high-risk prisoners not being identifiedand then going on to commit further offences.A shared OASys will mitigate the risk ofinsufficient information transfer, althoughmanagers in the community will not have accessto up-to-date prison case records. Within theprison system, up-to-date risk assessments andcase information will be available.NOTE1 Information flows were not the only or necessarily the main issues reported on in these cases.36 THE NATIONAL OFFENDER MANAGEMENT INFORMATION SYSTEM

GLOSSARYC-NOMISCOPSCustody PlusData ShareDelius / National DeliusEarned Value ManagementEDSIAPSIISThe project to develop the cTAG/iTAG commercial software package and itsdeployment in both the Prison and Probation Services. The project became theNOMIS Programme after plans were revised during the Ministerial moratorium.Contracted-Out Prisons. At present there are 11 privately managed prisons inthe UK.A short period in custody of up to three months followed by a longer period undersupervision in the community of a minimum of six months. The court specifies thelengths of the two parts and attaches specific requirements to the supervision part ofthe sentence to address the rehabilitative needs of the offender.A common repository of data, fed from case management systems that enableoffender managers and staff within Offender Management Units in the NationalProbation Service and public prisons to quickly and simply identify the location ofan offender, and the location and contact details of the offender manager/offendersupervisor for that offender.Delius is one of seven case management systems operating in parts of the NationalProbation Service. As part of the Probation Case Management System project,Delius will be upgraded to become National Delius and will be rolled out to allprobation areas.A project management approach which provides more accurate status information.It enables the project team to establish how much of the project has been deliveredand at what cost, compared to how much was originally planned.Electronic Data Systems is a US company, specialising in business processoutsourcing, information technology application development and thetransformation services using IT.Interim Accredited Programme Software is a national IT application that records anoffender’s attendance, completion and reasons for non-completion on an accreditedprogramme. Some Probation areas also use it as a case management tool.Inmate Information System. A management information system updated daily withinformation drawn from LIDS. It provides statistical information to NOMS, policeand the Parole Board and also allows day-to-day management of the DiscretionaryConditional Release (parole) and Home Detention Curfew schemes.THE NATIONAL OFFENDER MANAGEMENT INFORMATION SYSTEM37

GLOSSARYLIDSNet Present ValueOASys/e-OASysOMNISunk costsPRIMEPrison NOMISQuantum ContractSteriaSysconTAGThe Local Inmate Data System is a legacy system that supports each prisonestablishment in managing Reception Processing, Activities, Prisoner Movementand Inmate Finance.An investment appraisal technique that is used to calculate the present value offuture cash flows from an investment. Future amounts are discounted back totoday’s values to reflect the risks facing the project and the erosion of the value ofmoney over time.An evidence-based tool for assessing the risks posed by offenders, and to providea systematic basis for planning their management within the overall objective ofreducing re-offending. OASys is the generic name for the system, e-OASys is theversion used by the National Probation Service.The Offender Management National Infrastructure (OMNI) contract was signed on1 January 2006 by the National Probation Service and Steria, a French IT servicesprovider. The agreement’s duration is six and a half years.Past expenditures which cannot be recovered.The Prisons Information Management Environment project was the forerunnerof the C-NOMIS project. Its scope was to develop a replacement for the PrisonService’s legacy LIDS systems.A project within the NOMIS Programme begun after the moratorium. It is intendedto replace the Prison Service’s LIDS legacy system with the TAG application.A Private Finance Initiative agreed between HM Prison Service and Electronic DataSystems (EDS) in February 2000 that will run until 2012. It provides an integratedand networked information system to improve access for HM Prison Servicepersonnel to prisoner information.The IT services company undertaking infrastructure improvements in the ProbationService under the OMNI contract.Syscon Holdings Limited, a Canadian company, specialising in offendermanagement systems. Their TAG application forms the basis of the C-NOMIScase management software. In June 2007, the company was acquired by SecurusTechnologies, Inc of the US.The generic name for SYSCON’s commercial off-the-shelf software product calledcTAG/iTAG.Printed in the UK for the Stationery Office Limitedon behalf of the Controller of Her Majesty’s Stationery Office6075546 03/09 7724038 THE NATIONAL OFFENDER MANAGEMENT INFORMATION SYSTEM

Design and Production byNAO Marketing & Communications TeamDP Ref: 008884This report has been printed on ConsortRoyal Silk and is produced from acombination of ECF (Elemental ChlorineFree) and TCF (Totally Chlorine Free) woodpulp that is fully recyclable and sourcedfrom carefully managed and renewedcommercial forests. The range ismanufactured within a mill which isregistered under the BS EN ISO 9001accreditation, which provides the higheststandard of quality assurance.