The ultimate democratic data mash-up

So, online petitions are back with us again.

X-factor Whitehall is back for a new series. The power of the people is now only a few clicks away. Etc. etc.

Amidst all the flurry of any new service settling into being, the following thought occurred to me.

The big, obvious problem with an online service as an expression of democracy is that it’s not evenly representative. It’s skewed in all sorts of ways. The most apparent being that not all of the population have online access (or want it), and that distribution isn’t consistent across political or demographic groupings.

But others include people’s willingness to use the Internet for different tasks, the dominance of a few powerful voices with well-developed networks, “bandwagons and herds”, and the relatively small sample sizes involved.

On this latter point, compare the number of “signatures” required to put an issue up for Parliamentary consideration, with the number of votes required to turf someone out of Celebrity Big Brother. The influence of just one mention in the traditional, mass media can dramatically sway the standing of an issue.

But there is a way that we could do better. Or at least shine a different light on the petitions that we are seeing.

Data exists, within the big demographic and market analysis companies (CACI and Experian spring to mind) which does precisely some of this heavy-lifting. Adjusting research samples for all sorts of biases. Correcting for factors such as representativeness of a particular geographical area, preferences or capability to use certain channels, and so on.

I’m not an expert in this type of data comparison, but I do know that such adjustments won’t ever give us perfection. They will, however, add to our insight. In what might prove to be compelling ways.

I wonder who’ll be brave enough to attempt this, the ultimate democratic data mash-up?