Over the last forty years the concept of swing has transformed the interpretation of elections, first in Britain and later in Australia, India and other countries. But Americans have been oddly reluctant to apply this most useful of simplifying formulas to election reporting or to the analysis of voting trends.
Swing is a device for reducing to a single statistic all the complex movements in party support between one election and another. It offers a quick and easily comprehended indication of trends and of the extent to which voting behavior differs from one area to another.
On election night in Britain, as soon as the broadcasters announce the first constituency counts, they also report the local swing, the net percentage movement between Conservative and Labour since the previous election, and on this basis they suggest what will be the final outcome ‘if the swing in the remaining seats resembles those already declared.’ Although this health warning is dutifully repeated, in fact the projection from the early results has always given a remarkably accurate pointer to the overall division of seats.