The political economy of UK party support, 1997–2004: Forecasts for the 2005 general election

Abstract

Aggregate‐level models of party support patterns increasingly seek to evaluate both “economic” and “political” explanations of election outcomes. This paper explores the evolution of support for Labour, the Conservatives and the Liberal Democrats between 1997 and 2004. A number of politico‐economic models are developed and tested against monthly timeseries data for the 1997–2004 period. A variety of different assumptions are then made about likely movements in the key independent variables in these models in the run‐up to the next general election. Finally, these assumptions are used to inform a series of forecasting models in order to produce forecasts of likely party support levels in the early summer of 2005, the presumed time of the next general election. The analysis suggests, on balance, that the best performing statistical model is one that takes account of Blair’s leadership ratings, voters’ personal economic expectations, the “costs of ruling” and the continuing effects of the Iraq war. The central forecast from this model, for an assumed election in June 2005, produces vote shares of 39% for Labour, 31% for the Conservatives and 22% for the Liberal Democrats. Although a wide range of forecast election outcomes can legitimately be derived from the existing data, most of the forecasts reported suggest that a Labour victory is the most likely outcome of the next UK general election.