Irish Lisbon referendum update

Here's a quick update on the Irish Lisbon Referendum. There was a new poll from Millward Brown on Sunday that put YES at 53% and NO at 26%. However, getting rather more attention is a poll from Gael Poll which claims to show YES and 41% and NO at 59%.

As far as I can tell, this poll is hokum. The company don't seem to have a website so I can only go on what I've got, but the sampling of the poll seems to have been conducted at just ten sample points, suggesting a face-to-face survey with no attempt at a broad representative spread of sample. Compare this with a professional Ipsos MORI face-to-face poll, which uses in the region of 200 sampling points. Worse, a couple of sources indicate there is no attempt at weighting the poll.

The poll is being given some creedance because their poll before the first referendum got the result pretty much spot on, when polls carried out using what should have been solid methodology got it wrong. I would caution against reading too much into that. Even a stopped clock tells the right time twice a day.

Irish politics isn't my thing, I've no idea how the referendum campaign may go in the last week, or how fluid opinion is. However, respectable polling outfits aren't showing any real drop in that lead yet, and while some of the polls in 2008 got it wrong, they were showing the two sides very close - as opposed to the sort of towering lead the Irish polls are currently showing for the YES camp.

UPDATE: A mea culpa from Mick Fealty at Slugger O'Toole for reporting it. He has more detail from the methodology which really does bury this: "researchers were friends of the organisers who in turn interviewed people in their social groups, paying some attention to the spread of social class…" In other words, they asked their mates.