Paul Clarke has a nice summary of systems complexity generally, and the eternal battle between getting things done, and pleasing everyone. Fortunately I haven't seen the original stories on the Christmas Tree in question, so have no idea what the context is. I prefer it like that.
I think Paul is right to highlight the role of open data as we move forwards into a technical democracy, and the possible solutions/problems coming in as a result - I think there's a good chance that transparency can lead to ever-decreasing circles of receipt-checking, process justification etc, and the whole country implodes in a swamp of exclamation marks and daily mail headlines.
These 2 questions seem rather pertinent, IMHO: "And with what discretion? Authorised by whom?" - Are these the same issues we've been grappling with for years anyway, in the form of representative democracy? On a broad picture, it's not necessary for all citizens to be involved in all decisions all of the time - so we vote for the person we think we can trust most with power. We, as voters, are handing over discretionary power so that someone else is creating a world we want to live in. I call this "trust", because even today there's no way I can know (or want to know) everything my MP is up to. I have a wife and kids and a job.
I've yet to be convinced that the drive for more transparency isn't just a way of getting us to trust politicians less. The over-riding message from on high seems to be that transparency is there to hold people to account - which I think is a real shame, as open data is far more powerful as a platform for collaboration than accusation.
Transparency as distrust leads to a bizarre situation in which people we've "trusted" via our vote are then afraid to apply that power - especially considering a vote is local, while headlines are national or global. Worse, it can drive important decisions further into obscurity and complexity to avoid such scrutiny (and here it's hard not to draw comparisons with the banking industry as a warning).
Perhaps part of the problem is believing that cost is the deciding factor in how accountable (and hence transparent) a decision-process should be. But cost says nothing of either complexity or impact - both of which are much more important in deciding the "suitability" of decisions, I would say.
Cf. two other realms - banking, as mentioned already, and open-source software.
On one hand, the uselessness of auditors in predicting the collapse of banks serves to show how bad it is to have systems that can rapidly create complex models around themselves. Compare this to how open-source software operates - for a project to be sustainable, it is vital that complexity is managed, and that the code is readable by anybody. If the code is unreadable, it grows more slowly, is more prone to bugs and security risks, and is less maintainable. Both designing and refactoring code are essential to ensure a solid output.
Can we apply these lessons to government decision-processes? If transparency is the way forwards, then I think we have to - sure, there are fundamental differences between software (which, for instance, can be forked) and a democracy (which can't, quite so easily). But as things become more open and "many eyes" start taking peeks, the productivity-gains and effectiveness of open data mean that we cna't just assume that openness is enough. Openness needs to be accompanied by feedback - the same constant re-factoring process that goes into software engineering.
In other words, it is not enough to use transparency to justify decisions already made, and to prevent bad decisions being made in future through the threat of later accountability. Openness in data needs to go hand-in-hand with an openness to change - to influence new ways of contributing, of collaborating, and of voting for those who we trust. Even new ways of thinking and feeling about why the decisions are being made in the first place.
