The Party Leadership Model: An early forecast of the 2015 British General Election. (18th May 2015)

• Record Type: Journal Article
• Authors: Murr, Andreas Erwin
• Is Part Of: Research & politics. Volume 2:Number 2(2015:Apr.)
• Publication Date: 2015-05-18
• Subjects: Bayesian inference -- citizen forecasting -- election forecasting -- British general elections -- party leadership contests -- 2 × 2 contingency table; Political science -- Periodicals; 320.05
• DOI: 10.1177/2053168015583346
• Languages: English
• ISSNs: 2053-1680

Abstract

British political parties select their leaders to win elections. The winning margin of the party leader among the selectorate reflects how likely they think she is to win the General Election. The present research compares the winning margins of party leaders in their party leadership elections and uses the results of this comparison to predict that the party leader with the larger winning margin will become the next Prime Minister. I term this process "the Party Leadership Model". The model correctly forecasts 8 out of 10 past elections, while making these forecasts 4 years in advance on average. According to a Bayesian analysis, there is a 95 per cent probability that having the larger winning margin in party leadership elections increases the chances of winning the General Election. Because David Cameron performed better among Conservative MPs in 2005 than Ed Miliband did among Labour MPs in 2010, the model predicts Cameron to become Prime Minister again in 2015. The Bayesian calculation puts his chances of re-election at 75 per cent.