Imagine a country where representatives are elected by
plurality voting from single-member districts like in the US or UK. Suppose Labour wins 40% of the vote in every seat, and the remaining 60% is split evenly between Conservatives and Liberals. How many seats will Labour win? Every seat, of course, since the Labour candidate will have placed first in every seat.
In real life, such uniformity doesn't exist. Humans clump together for many reasons and create communities that have different resources and interests. Labour has dozens of "
safe" seats, as do, to a lesser degree I think, the
Tories, and to a considerably lesser degree the
LibDems. The
2005 result shows Labour winning a mix of urban and traditional working-class constituencies, while the Tories garner votes from the suburbs and countryside. The LibDem seats still come largely from places with historical Liberal leanings like the southwest and some parts of Wales and Scotland. I'm especially curious about the sociology of the remarkably competitive seat,
Ealing Central and Acton.
So, to answer your question, suppose we see a "swing" (to use the British term) away from Labour today. In many Labour seats that will have no effect on the identity of the MP. Even if 10% defect from a Labour incumbent who won with 70% of the vote last time, that same MP will still win, just with a diminished majority. Same holds true for seats where the other parties (like the Scottish National Party) have substantial majorities.
The BBC has thoughtfully provided a little
playground where we can simulate various results. Set the dial to 30% for each of the three major parties, and 10% for the SDP, Plaid Cymru, and the Irish parties, and see what you get for a distribution of seats in Parliament. I get a Labour plurality, but a minority government.
Now look at the opinion poll results. Click on each of them from ComRes to Harris and see what happens in Parliament. ComRes shows the largest Conservative margin, but it's still not a majority. Still if you bump up the Tories share by taking votes away from Labour, you can get to a Conservative majority around 39%.
So it's certainly possible that the Liberal Democrats could out-poll Labour yet still win
considerably fewer seats.
Like AutumnDemon, the
bookies, and these polls, I'm also guessing we'll see a Conservative plurality. In 1974, Labour edged out the Tories by a measly four seats, got the Queen to accept them as the Government,* then went back to the country eight months later and won a majority. I don't see that happening for the Tories. Even giving the Conservatives an implausible share of the vote in the BBC's calculator doesn't give the Tories a very large margin of seats. Holding another election may not yield a majority this time round.
No-majority is currently on 4/5 at Ladbrokes; not such a good deal as 8/3 in 1974.
*A hot political issue in some royalist circles was whether the Crown might have some say in the choice of Government. Cooler heads prevailed.