Two weeks after the deliberative poll, and around a month after the initial poll, we now appear to have all the results we need to assess the representativeness of the Tomorrow's Europe poll, with the release of the results of the initialal poll of 3,500 (PDF). This was the group from which the sample of 362 was taken - and so their representativeness should be compared to this initial poll. I've already noted my suspicions about the demographic representativeness - but, of course, what's needed as much as anything in an exercise like this is representativeness of opinion.
So as one of my suspicions was that pro-EU types would be more likely to take part, let's take that as a first point of comparison - especially as this is surely the most easy to measure political opinion when it comes to any aspect of EU politics. As a bonus, support for the EU is one of the things tested in the regular Eurobarometer opinion polls (most recent PDF), and so this makes for an easy point of comparison to a larger sample.
So, those who thought their country's membership of the EU is a bad thing?
Eurobarometer - 15% ; 3,500 sample - 12.3% ; 362 participants - 9.4% ; 362 (post-deliberation) - 2.8%
And those thinking it is a good thing?
Eurobarometer - 57% ; 3,500 sample - 69.4% ; 362 participants - 79% ; 362 (post-deliberation) - 89.9%
This already looks a tad concerning. Eurobarometer (a survey of 27,000 people - 1,000 in every EU member state) would seem to show that people with a negative opinion of the EU are under-represented by 5% among the Tomorrow's Europe poll's participants, while those favourable to the EU are over-represented by 22%.
But wait - the figures are not directly comparable. Because the results we have for the Tomorrow's Europe polls doesn't distinguish between "slight" and "very" favourable or disfavourable - unlike the Eurobarometer poll.
That 15% who think the EU has been bad for their country in the Eurobarometer survey is actually 15% who think EU membership has been very bad for their country, while in the Tomorrow's Europe poll, the 12.3 and 9.4% takes into account the full range of negative opinions, right up to (but not including) "have no opinion one way or the other". And, just to give an indication of how large a proportion the "no opinion"contingent may be, in the last ten years of Eurobarometer surveys, the percentage polled thinking EU membership is neither good nor bad has stood somewhere around the 25-30% mark.
So, those results may be out by as much as a third. How can we tell? We need the raw data - because the Tomorrow's Europe participants were asked to rate their support/opposition to the EU on a sliding scale, so these figures do exist. But why haven't they been released?
It also, it must be said, seems a bit odd that support for the EU rose so much amongst the participants - and rather unfortunate, as that's hardly likely to quell the suspicions of the eurosceptics that this was an exercise in brain-washing...