Why do major collaborative ICT projects in government fail? A post-structuralist interpretation, with reference to the Queensland shared services / health payroll fiasco

  • Michael Chisnall

    Student thesis: Doctoral Thesis


    Why have governments experienced “crises” and “failures” when implementing major collaborative ICT-enabled projects? And why have these problems continued despite extensive examinations of them across a wide range of disciplines? With few exceptions conventional explanations assume a rationalist paradigm, produce descriptivist observations, and conclude with (different) lists of critical ‘success and failure’ factors. Yet these accounts do not seem to get to the heart of the problem. This work uses a poststructuralist discourse-theoretical approach to develop new critical explanations of why these problems occur, extending this approach for the first time to apply to detailed empirical research on government ICT. Empirical evidence is drawn from two case studies. The first, smaller, case looks at tensions within a government workplace during a small-scale, but potentially disruptive, change. It illuminates contestation and the manipulation of information within hierarchies within a routinely operating public service organization (one that is not in crisis), drawing on interviews with the public servants involved.
    The second case takes up the whole of Part II and is a re-narration and discourse analysis of the "failed" Queensland Health Payroll System (QHPS) project. This extended analysis uses a comprehensive and detailed set of witness statements, submissions, transcripts and exhibits from an official Commission of Inquiry into the QHPS failure, held in 2013, together with contemporary media articles. The saga leading up to this began a decade earlier in 2003 with the launch of a Shared Services initiative, designed to save money by centralizing all Queensland State government corporate services in collective in-house units, instead of relying on the previous departmentalized arrangements. Four years later, after an internal audit uncovered overspending, delays and growing agency resistance, the top Queensland Treasury executive management sidelined internal expertise, deciding instead to contract with the private sector. Following external consultants’ advice, the complete project was outsourced to IBM, in the hope that their greater expertise and firmness could solve earlier problems.
    IBM soon encountered the same entrenched internal conflicts which had prevented Queensland Government Departments from properly engaging. Nonetheless, the project continued, plagued by arguments over scope, cost and schedule. An acute test of the new arrangements loomed over the payroll system for Queensland’s large healthcare system, where the existing software was going out of contract. IBM developed a system that was controversial and problematic, prompting examination of whether the outsourcing should be terminated altogether. Instead, the testing of the new system was pushed through (ignoring many warning signs), until a hugely flawed system was cleared for use in late 2009. After going live, chaos ensued as many healthcare staff went unpaid or were underpaid. The system was publicly labelled as a fiasco by Ministers. The State considered taking legal action against IBM, until top officials and the cabinet unexpectedly waived a considerable claim for compensation.
    This long-run and large-scale failure could have been averted at four different junctures, covered in detail in Chapter 4 to 8. Building on existing multidisciplinary academic and practitioner literature, including Project and Change Management, Planning, Information Systems Research and Public Administration, the analysis identifies fundamental hierarchical, communication, deference and leaderist logics that, together, explain why things went predictably wrong in ways that the actors and structures involved could not counteract. The logics observed, and the theory underlying them, extend the implications of this thesis across a wide range of major ICT (and, potentially, other bureaucratic) projects. It concludes with a proposal for two social heuristics, the use of which has the potential to lessen the risk of failure.
    Date of Award2018
    Original languageEnglish
    SupervisorDavid Marsh (Supervisor), Paul Fawcett (Supervisor) & David Carter (Supervisor)

    Cite this