The final polls before the British election today have been breaking quite nicely for the Conservatives as the Lib Dem boomlet fades, while Labour has been getting no traction and still may finish third in votes.
The Real Clear Politics average has the final polling average as:
Conservatives 36%, Liberal Democrats 27.4%, Labour 27%
Right after the first televised debate, the Tories had been down to a 2-3 point lead with many individual polls showing Nick Clegg's Lib Dems in the lead for the first time in the party's history. Ever since, the Tories have been gradually extending a lead at the expense of the Lib Dems.
If every seat swung uniformly, the encyclopedic Electoral Calculus would give us the following seat breakdown:
Conservatives 299, Labour 233, Liberal Democrats 86
The Conservatives need 325 to give them an outright majority, and despite these predicted numbers, I think the Tories will manage to avoid a hung Parliament and (just) clear the bar of being able to form a Government without the support of other parties.
Behind this assumption is the deep suspicion that all seats won't swing uniformly. Some will swing more than others, and those are more likely to be the critical marginal seats where voters feel the greatest burden of deciding who gets into Number 10 and with what kind of Government.
Already, polls are showing the marginals with an overall swing to the Conservatives of 1 to 2 points greater than the national average. Meanwhile, the Lib Dems might not be able to siphon off Tory seats despite their momentum and may make most of their gains at the expense of Labour.
In 1997, Labour were predicted to get in with a 370-seat majority. They won 416. Massive tactical voting by Liberal Democrat supporters in competitive Labour vs. Tory seats juiced their total.
The Tory-Lib Dem alliance in the electorate isn't quite as strong, but it's still a factor against the flailing Labour government: the Conservatives have been gaining completely at the expense of the Lib Dems in the last two weeks, showing there's some overlap in their base.
Meanwhile, Cameron has been closing strong and making a strong pitch against a hung Parliament. Judging by his party's rising poll numbers, people seem to be buying it:
An example from over here would help illustrate the point.
In 2008, Indiana was competitive in a Presidential election for the first time in decades, eventually going to Obama. Yet a uniform national swing would have had McCain still winning the state by 10 points. Its proximity to Obama's home base of Illinois and its economically distressed status can only explain part of this shift. Take a look at this map showing the county-by-county shift in votes from 2004 compared to other neighboring states:
Indiana was an outlier. The next counties over in Michigan, Illinois, and Ohio did not swing as strongly for Obama despite their demographic similarities and Great Lakes status.
The only thing that can explain this single state swing is a competition effect. When a state is competitive when it wasn't before, turnout surges as new voters come to the polls. This effect is particularly strong in the handful of states that are portrayed as neck-and-neck or decisive to the outcome. A certain inertia dictating a too-close-to-call race kicks in to produce a single-state outcome that can't be explained by demographics or regional differences. Case in point: Florida was Ground Zero in the 2000 election and it swung less strongly to Bush than would have otherwise been expected. Ohio was Ground Zero in 2004 and it swung differently than Pennsylvania and Michigan which were similar economically and demographically but weren't quite as competitive. Subconsciously, voters in those states seemed to take seriously their role as the "deciders" of the Presidential election and voted accordingly.
This effect will be magnified in Britain where the governing unit is a Parliamentary constituency one sixth the size of a typical Congressional district here. There you get swings that are well outside the national norms depending on which party is perceived as having the best shot at throwing the bums out.
One of my favorite stories of this General Election is the race in Morley and Outwood between the unfortunately named Gordon Brown henchman Ed Balls and Conservative challenger Antony Calvert.
Labour won this newly created seat by 21% at the last general election, yet the race is portrayed as neck-and-beck, with the betting sites now pegging Calvert as the favorite. Calvert has been running a Scott Brown-style Internet campaign fueled by small donations gunning for a "Were you up for Portillo?" moment -- a last-minute come from behind decapitation (or castration?") of a high-value Cabinet minister and potential Party leader.
What's happening in Morley and Outwood will likely be repeated up and down Great Britain today as people calibrate their votes to do the greatest possible damage to Gordon Brown and Labour, for whom the infamous "bigoted woman" incident was the last straw.
My prediction: The Tories wind up with 328 seats and the chance to form a Government outright.