"We can't allow things that are inaccurate to stand." — The Word of Our Dan, February 19, 2008.

Friday, February 09, 2007

The morning after

The only huge surprise of the night in last night's by-elections was in Port au Port. For a supposedly "tight" race, it was nowhere near being so. In fact, people in the district were so furious at Danny, on whose watch the Stephenville mill would never, ever close, that the margin spread between the two parties increased by 14 percentage points.

And for all the slagging that Danny Dumaresque has taken by Liz's Rapid Reaction Battalion of planted VOCM callers — the only Rapid Reaction Battalion in the province — he was spot-on in predicting a substantial increase in Liberal support in Ferryland, where the margin swang over 20% in the provincial Liberal's favour.

Of course, that's about as meaningful to the final result as a 20% swing towards the federal Liberals in Wild Rose.

In Kilbride, the vote-share change (Liberal down -1.2%, split equally among the Tory and NDP candidates) is of no great statistical significance, as is the three-district total overall: Liberals up 2.3% in the three districts combined, PCs down -2.6%, NDP up 0.9%.

But consider this: According to the recent CRA polls, Danny Williams and his party have been polling up to 14 points above their actual vote share in the 2003 election. In 2003, he took 58.8% of the popular vote. CRA has had him up to 73%.

So why doesn't this Smallwood-esque level of support show up in actual voting results?

That's a question Danny might want to ponder, although it's largely academic: you only need to win by one vote, after all. It's certainly something the opposition parties need to ponder.

And it's especially a question CRA needs to ask itself. Is there a flaw in its methodology when it comes to fieldwork or weighting in its Newfoundland and Labrador quarterly?

On to Humber Valley and Labrador West!


Post a Comment

Subscribe to Post Comments [Atom]

<< Home