Devil's Advocate: Open source for government

Despite the UK government's reluctance to embrace open source, Martin Brampton points out the many ways they're suited to each other - the least of which is cost.

The government regularly claims it will reduce the cost of administration through the use of IT. Such a claim is inherently implausible, given the sector's track record in IT projects. Quite apart from that, government appears to be going about things in entirely the wrong way.

Recent thinking has involved emphasis on shared services. They are central to the Cabinet Office eGovernment Unit's strategy and also to the Treasury's efficiency review. The concept relies on the assumption that money will be saved through the sharing of administrative functions such as finance and human resources.

This depends on a number of prejudices that are popular with politicians but of highly dubious veracity. One is the belief that the private sector is inherently more capable of running an organisation efficiently. This is probably untrue but is made more damaging by politicians frequently having an out-of-date understanding of what happens in the commercial sector.

Were this not so, government would be making more use of the concepts developed by the open source movement. Scarcely any commercial organisation now has a blanket policy of ignoring open source, and a substantial number are happy to treat it on its merits against closed source alternatives.

Yet not only does government make little use of open source, it has done nothing to utilise the fruitful ideas that have vitalised the movement. In fact, the failure to grasp the significance of open source is widespread but it is a particular loss to the government sector, where enormous gains could be made.

Open source developments are becoming a significant economic force, with strongly positive effects. They are akin to the equally strongly positive effects of the free exchange of scientific knowledge, something that is being increasingly stifled by commercialisation. Sadly, the most important factor about open source is often seen to be that it is free.

In fact, the most important factor about open source is that it encourages a far more efficient model for the creation of software. Quite apart from the huge gains that are made in reducing sales and marketing costs, the actual process whereby open source is written is mutually beneficial and promotes efficiency.

Open source projects are not islands. Many have active links with other projects in complementary areas. Even those that don't usually place high value on the open source community and frequently seek solutions to problems through the use of material already developed in another project. This is the kind of thing that has huge potential for government.

It has been argued that government cannot simply commission the software it needs as open source because public money is at stake, and benefits would accrue beyond the immediate goals of government. Quite apart from the short-sightedness of the argument, it also ignores the possibility of adapting the spirit of open source to the sphere of government.

Even if the standard open source agreements were to be adapted to restrict the openness to the whole of national and local government, that would still be a large enough world to be significant. Moreover, the concept could be applied not only to software itself but also to knowledge about how best to deploy automated systems.

Such a move would be likely to cast doubts on the process of sharing resources. This involves too much emphasis on products and too little on processes. It requires considerable reorganisation, and increases the complexity of the requirements, a factor that is highly likely to increase the software cost disproportionately. It also relies on another fallacious belief, the idea that technology solves people problems.

Success is far more likely to come from allowing different parts of government to meet their own needs while encouraging reuse of both technology and expertise. Present efforts look likely to have much the same results as past grand schemes. It is time to try a more radical approach.