Whiteley, Paul. "Forecasting Seats from Votes in British General Elections". The British Journal of Politics and International Relations, Vol.7, No.2 (2005): 165-173.

Whiteley, Paul. "Forecasting Seats from Votes in British General Elections". The British Journal of Politics and International Relations, Vol.7, No.2 (2005): 165-173.

• There is a source mine of other voting predictions in British general elections available on page 165.
• Winning a general election in Britain does not mean winning a majority or plurality of votes, but instead means winning a plurality of seats in the House of Commons. A majority vote does not always translate into majority representation (166).
• In 1909, a predictive measure of votes to seats named the Cube Rule, or MacMahon's Law, was developed in predicts the following proportions of seats based on general votes:
• "The law, briefly, states that the proportion of seats won by the victorious party varies as the cube of the proportion of votes cast for the party over the country as a whole" (166).
• The Cube Rule worked well in the 1950 general elections, but has failed to deal with strong regional biases -- like those in the American South during the 1944 congressional election -- and cannot incorporate multiple parties (167).
• In 1968 through 1969, work was done to modify the Cube Rule so that it could be applied to multiparty systems, like that of Canada. They concluded that the seat share equals the votes for the party divided by the sum of all votes for that party within a predefined parameter (167).
• The author proposes the application of the Theil Formula to the British general elections, using elections from 1945 to 2001 as the given dataset. The formula predicts that the portion of seats [S] will equal the seat share of the previous election times the party's vote [P] share times a given error, all relating to preset parameters (168).
• The results of this formula applied to the given dataset are given from page 168 to page 170.
• The formula was tested on the results of the 2001 British general election and has mostly accurate, with the most differentiation coming from the Liberal Democrats. All deviations in seats gained, however, were not found to be statistically significant (171).
• The author makes a prediction about the seat percentages in the upcoming 2005 general election, predicting a strong Labour victory (171).
• This prediction came extremely close to the actual recording votes and seating, including the Labour government.