Do local elections predict the outcome of the next general election? Forecasting British general elections from local election national vote share estimates

Abstract

This paper outlines a method for forecasting British general elections from national level vote shares at local elections. Although local elections are notionally 'local', the evidence suggests that they at least partly mirror national electoral fortunes. A simple general election vote share on local election vote share regression model that accounts for partisan differences and incumbency effects fits past data with reasonable accuracy. Based on the results of the 2013 and 2014 local elections, the model forecast a 56% probability of hung parliament, with a 78% probability of the Conservatives receiving the largest share of the vote., local elections,

Do party leadership contests forecast British general elections?

“…If we classify the above-mentioned forecasting models of 2010 and 2015 by their lead time, we note that in 2005 the shortest lead time was one month, and the longest was six months (Gibson and Lewis-Beck, 2011: Table 1); in 2015 the shortest was less than a week, and the longest was twelve months (Fisher and Lewis-Beck, 2016: Table 1). Curiously, in both elections the most accurate model-in terms of forecasting the seat share of the incumbent party-was also the model with the longest lead time (Whiteley et al, 2011;Prosser, 2016). This underscores the above point by Lewis-Beck and Tien that models with longer lead time may actually forecast more accurately than models with shorter lead time.…”

“…The first column of Table 6 lists the different models. In addition to the Party Leadership Model, the list includes both structural models such as the local elections model (Prosser, 2016), the PM and Pendulum model (Lebo and Norpoth, 2007), and the citizen forecasting model as well as the PSA Expert Survey (Fisher et al, 2019a), polling averages with uniform national swing, and betting markets (Fisher et al, 2019b). The next two columns note the models' forecasted seat number for the Conservatives and Labour, respectively.…”

“…The fourth column reports the accuracy score. Note: Local elections (Prosser, 2016;Prosser and Fisher, 2019); Party Leadership Model Murr, 2019); PM and Pendulum (Lebo and Fisher, 2019;Lebo and Norpoth, 2007;Norpoth, 2019); Citizens (Murr et al, 2019;; PSA Expert Survey (Fisher et al 2019b); UNS from Poll Average and Betting markets average (Fisher 2019a); Accuracy = the score (0-3) on accuracy; Lead = the score (0-3) on lead time; Overall = the quality index calculated from this formula: (Lead x Accuracy) / 9.…”

The Impact of Candidates’ Profile and Campaign Decisions in Electoral Results: A Data Analytics Approach

“…For example, [54,55] present forecasting models for the 2015 parliamentary elections of Great Britain; for which the number of candidates predicted to be elected are 630 and 650, respectively, even when the total number of seats in the election is 632. A similar situation, for the same elections, is presented in [56,57], who forecast vote shares for the most popular parties participating in the election.…”

10 Downing Street: who’s next? Seemingly unrelated regressions to forecast UK election results

“…Many such models were applied to the 2010 and 2015 British general elections (see Gibson and Lewis-Beck 2011;Stephen and Lewis-Beck 2016). For example, Hanretty, Lauderdale, and Vivyan (2016) and Prosser (2016) both used, in a different manner, local election results to forecast national electoral fortunes. Lewis-Beck and Stegmaier (2011) as well as Murr (2016) exploited vote expectation surveys in order to produce "citizen forecasts" for the major parties.…”