Not so fast, Mr Montgomery
The 'Danish surprise' is not in Mecom's interim results

Blog buzz revealed election results

Want to know the results of next year's US election? Follow the blog buzz. At least if this Norwegian experiment is anything to go by:

A blog measuring the blog buzz around Norway's political parties and key political issues proved to be quite accurate when it came to predicting the winners and losers in yesterday's local election.

I must admit that I was very sceptical, and remain sceptical, to whether measuring the frequency of keywords, or correlations of keywords, in the blogosphere, is a reliable way to predict election outcomes.

Still, just as political scientist Dag P. Svendsen predicted on his blog, the election winners were indeed The Labour Party (AP), The Progess Party (Frp) and The Conservative Party (H), although the differences between their gains from the last local election were marginal (currently 2,1pc, 1,2pc and 1pc). As for the losers, Svendsen's predictions were correct for The Socialist Party (SV), but incorrect for The Liberal Party (V) and the Christian People's Party (Krf).

Before the election I said I had little faith in using quantitative analysis of blog buzz to predict election results or to measure how concerned people are about different political issues:

A quantative analysis doesn't look at what values people attribute to the party or issue, and is hampered with methodological problems such as the risk of measuring spurious connections, how the blogosphere may not be representative for the population at large etc.

However, in this case the blogosphere proved to be an excellent mirror of the country's population. Now, THAT is interesting. So for all those out there who thought the blogosphere was the exclusive domain of nutters, crakpots and losers: at least in Norway, bloggers seem to be quite representative of the population's overall voting pattern.

I have also said I think blogs are comparable to digital versions of the conversations people have over a cup of coffee or a pint of beer, and as such the blogosphere can be a great resource for politicians who want to know more about what issues people are concerned about; how the political parties and the way they deal with these issues are perceived etc.

In other words, the blogosphere is an interesting place to keep an eye on both for forward thinking politicians and companies, but they will need to apply some sort of qualitative analysis in order to get the most valuable input.

Not to mention how the blogosphere offers fabulous opportunities for politicians to engage directly with their potential and existing voters, unmediated – especially in a local election in such a small country such as Norway, I might have added.

Sadly, I didn't see one top politician grasping this opportunity. Yes, quite a few of them had blogs at this election, put out some videos on YouTube even, but it bore every hallmark of being something they'd been told they should do – yet, with the exception of one or two youth politicians, didn't have a clue how to.

Two interesting tools for filtering the US political blogosphere in a meaningful way here (via Poynter's E-media tidbits).


The Norwegian local election is not the only event where the use of blogg monitoring tools have predicted the results of an election, i used the tool successfully in monitoring the French Presidential election earlier this year. There are a number of studies that have confirmed the relationship between the validity of aggregated information from the www and the link to the ex ante measurable results from the real world, the monitoring of elections and election results beeing only one example among many.

The comments to this entry are closed.