Tomorrow could be the end of the United Kingdom of Great Britain and Northern Ireland. It could be the day Scotland votes to repeal the 1707 Act of Union and leave the UK.
Whether or not Scotland should become its own kingdom (First Minister Alex Salmond is insisting Elizabeth II of Great Britain and Northern Ireland would simply add Elizabeth I of Scotland to her titles) is a matter for the Scots to decide (and commenters to argue, as I suspect they will). This post is to explain why…they won’t.
Recent polls are saying the matter is close, a couple even had “Yes” (the secession option) in the lead, once the raw data was normalized, or “weighted.” In fact, though, both polls actually had “No” (the unionist option) winning with the raw data.
Now, to be clear, data normalization or weighting is a fine and proper thing to do when it comes to poll data. One has to make sure gender, economic status, regional residency, etc., are reflective of the population as a whole. The problem comes in when they weight based on political preference. This causes a whole host of trouble. They key one is as follows: party preferences can, and often do, change. Numerous American pollsters leave political preference unweighted for this very reason, while those that didn’t essentially imposed their assumptions about the electorate on their own poll – to disastrous results in 2012.
How much of the “weighting” in Scotland is due to political preference versus justifiable demographics is not known (at least to us in the public), but this issue is enough for me to predict that “No” is probably stronger than the polls indicate.
Another reason is the undecided vote (somewhere between 5 and 10 points in most polls). As it happens, the most prominent “Yes” man is the aforementioned First Minister, who has been a leading Scottish Nationalist for over two decades (and has been in his current job since 2007). Thus, his is the de facto incumbent in this vote (even though he isn’t on the ballot and his preferred outcome would be a seismic change). Undecided voters usually tack against the incumbent, unless a challenger is pushing dramatic change. With the incumbent himself promoting the major change, I’m guessing “No” will all but sweep the undecided vote.
Therefore, I’m predicting No wins by at least ten points and more likely twelve, 56-44 to be exact.
We’ll see how badly I’ve messed this up by Friday morning.
