I am not sure that anyone has ever actually thrown a baby out with its bathwater, but the old saying came to mind when we at Brunel University looked at recent reports on two government projects. Studying patterns in the events related by them raises questions about whether good practices built up over decades of harsh experience have been discarded in the dash for digital.
We looked in detail at the NAO and PAC reports on the CAP-D project, and the report published by the Institute for Government on Universal Credit (UC). Our goal was to identify underlying research questions to support funding bids. We managed to do that, but the bad news is that the questions relate to major fissures that may have opened up in the way government approves, manages and assures big projects with an IT component. Our note on our findings is at http://bura.brunel.ac.uk/handle/2438/13368.
We took as one of our references the Cabinet Office report from 2000, “Successful IT”, which looked at project failures of the 1990s and was one of the foundations for the Gateway review process for project assurance. Shockingly, the first few chapters of it describe exactly the scenarios of CAP-D and UC that we deduced from the reports: distributed authority and responsibility, a loss of focus on purpose and approach that should have been underpinned by a proper business case, and failures of programme and project management.
These are all issues that should have been identified and addressed through the Cabinet Office’s major projects assurance processes that were built around Gateway, the Treasury’s business case development and approvals process, and decades of learning the difficult art of project management. The fact that they were not, despite a strong Cabinet Office presence in both, begs the question of what happened to all these established procedures.
To give one example, the reports discuss the issue of the Senior Responsible Owners (SROs). They focus on the high rate of turnover, but what is significant is who was in the role and what they thought their role was. The original purpose of the SRO role, as set out in Successful IT, was to provide a firm anchor of overall responsibility for the achievement of the benefits intended for the project within the organisation’s core functions. So typically an SRO would be a senior civil servant responsible for the policy or administrative functions within which the project lay, giving him or her the highest responsibility and power over the project to avoid split authority. Yet with both CAP-D and UC, a Cabinet Office official was brought in at the end of a succession of SROs, seemingly in the different role of project director. Gateway reviews are a tool for the SRO and use of the business case as a foundation: weaken or remove those elements and the system falls apart.
Evidently, and possibly related to the Cabinet Office IT spend controls process imposing particular approaches to IT procurement and development, there was much debate and contention about how to source technology components, and great confusion about what should be done in an “agile” manner. The arguments seem to have obscured the policy and operational objectives of the projects. Again, assurance processes should have been able to draw on a clear business case and help the projects get back on track. Instead, we see a dominance of technology-centred arguments.
The reports of the projects point towards them suffering from being tagged as “digital by default”, which rightly or wrongly had them being treated from the outset as if they were web based developments to be done using agile methods — something irrelevant to the primary goals of both. Whether that too led to the collapse of the approvals, management and assurance processes is hard to tell, but we need to ask if it was a contributing factor. There is certainly some evidence from both cases that measures taken at the centre of government relating to managing IT, and its interventions in the project departments, caused disruption. They were intended to improve the implementation of projects and the public’s experience of dealing with administrative processes, but seem to have achieved exactly the opposite. Now, where did that baby go?