ABSTRACT
The British election system is well known for producing disproportional results: most parties get different percentages of House of Commons seats than they do of the votes cast. It is also prone to produce biased results, whereby two parties with the same percentage of the votes get different percentages of the seats: on our measure used of this, the 1997 result was the most biased (favouring the Labour Party) in the last half-century. Previous analyses of that bias identified six geographical components involved in its production. This article explores those components further using graphical methods, employing three categorizations of votes — wasted, effective and surplus — to suggest why the 1997 outcome was so biased. Labour’s focused geographical campaign then improved its ability to translate votes into seats very substantially: it significantly reduced its ratios of both wasted votes to seats lost and surplus votes to seats won, and so achieved a geographically much more efficient distribution of support than before, whereas too many of the Conservative Party’s votes were cast in the wrong places.
