This could make an interesting topic for a paper – but I need to make sure it's not too general. Others can do polical philosphy far better than I.
There is (will be) a lot of literature and claims around this subject. from both theoretical and practical viewpoints. The argument is particularly well rehearsed in the field of eVoting, though the introduction of Shared Source by Microsoft and others has muddied the waters to some extent.
It will be best to find a specific example where the open/closed decision was made by a government body, and to explore its consequences and implications.
The focus must be on what's happening in the real world. That is, a tool that is being actively used, not something that was funded, developed and is now comatose. Though it might be interesting to look at the factors behind why some e-government applications fail – which might or (more likely) might not be related to the licensing regime.
One possibility would be a discussion of the issues around a system that our department has developed. There are some serious practical reasons why releasing the code would be difficult.
Among them I can see:
- Ownership: Responsibility for the code still has to rest somewhere
- Quality Assurance: The users would want some sort of assurance that the code hadn't been hacked and was stable
- Support: Who would support the users, in configuring, maintenance/bugfixing and development of new feature
- Funding: Who would pay for all this, and how?
Solutions to these problems have been found before, and applications have been successfully transferred to the FLOSS domain. But are they applicable in the circumstances of say an internally developed e-petitioning system? Would moving to FLOSS make any difference to the perceived or actual openness and accountability of the application and service that is provided? How? Indeed, how much is openness and accountability a requirement in e-government decisions?