F/OSS and the Public Sector

I've long held the view that the public sector stands to gain more from the adoption of F/OSS technology and principles than perhaps any other area. This is not just about saving on software licensing, reducing integration and support costs, fostering reuse and increasing the transparency of software projects. As if these benefits alone were not enough: costs are frequently an issue and transparency can be a serious problem where huge, complex, drawn-out and all-too-often-prone-to-failure public sector ICT projects are concerned. More than this, though, an opportunity exists to create, as Dan Bricklin called it in his article Software That Lasts Two Hundred Years, "Societal Infrastructure Software".

Rather than go into detail here I would instead recommend that those interested in the idea of software as a part of the infrastructure of society read Dan's excellent article. And consider how this proposition becomes even more compelling when you think through the positive implications of providing open, unimpeded access to the workings of such vital infrastructure to those in school, higher education and, well, any interested citizen for that matter. Transparency combined with the opportunity for public engagement through open collaboration has the potential to create an unprecedented sense of shared stakeholdership, and society stands to benefit from a great deal more than simply a reduction in infrastructure costs or an increase in the number of public sector ICT projects that deliver on time (or at all...) With greater transparency will come trust, and sustainability and social progress through public engagement.

So, increasingly I find myself, like many others, asking why it is that in general the public sector appears to be so far behind the curve when it comes to F/OSS adoption. We do hear that there is a faster uptake in places such as Brazil and where budgets are likely to be more limited than in, say, the UK or the USA. France's national police force has reportedly saved 50 million euros since 2004 through adoption of F/OSS and migrating a portion of its workstations to Ubuntu. Back in February the UK Government issued a document titled Open Source, Open Standards and Re-Use: Government Action Plan, in which it was reported that over the previous five years "government departments have shown that Open Source can be best for the taxpayer", and suggested that "we need to increase the pace". However, the overall picture with regards public sector F/OSS adoption, both here in the UK and elsewhere, remains seriously disappointing. Furthermore, we are still hearing reports that some government agencies are violating their own procurement policies by mandating vendor-specific technologies in their invitations to tender. Where instead they ought to be simply defining requirements - not solutions - and thus creating a more fair, diverse and competitive market, and one that has the potential to lead to better value for the taxpayer coupled with increased innovation.

Public sector adoption of F/OSS appeared to be one of the biggest concerns for attendees at the Accelerating Enterprise Adoption of Open Source conference hosted by BT earlier this month. At a roundtable organised by IBM, London School of Economics and OpenForum Europe I attended a few weeks later this topic came up once again and was the subject of much debate. There are likely numerous reasons why uptake in the public sector is slow compared to in other areas, despite the many clear and significant benefits use of F/OSS in place of proprietary software promises to afford. Factors such as the large size of many public sector organisations and their often deeply entrenched, complex ICT legacies must be taken into account. However, serious concerns have also been raised over the fragmented nature of public sector procurement, alleged non-adherence to policy and the suggestion that such organisations are simply not set up to procure F/OSS and the associated services.

Both F/OSS and related services can be, and frequently are, packaged for consumption in the same manner as with proprietary software and services. Although, this is not true of all F/OSS, and to solely consume it in such an unsophisticated manner will lead to severely limited access to the software commons and otherwise significantly diminished returns. It would appear therefore, that whilst F/OSS projects can always benefit from striving to make their software easier to consume and to lower the barriers to participation, that the procurement engines of public sector organisations must update their understanding. Top-down policy may help, and so will education, procurement guidelines and incentivisation, e.g. via robust TCO analysis and cost accountability.

I'd be very interested to hear the observations of others, additional perspectives on public sector F/OSS adoption issues and, of course, suggestions as to how the pace of adoption might be increased.