Forecasting the Vote for a Third Party: The British Liberals, 1974–2005
| Author | Éric Bélanger,Richard Nadeau,Michael S. Lewis-Beck |
| Published date | 01 November 2010 |
| Date | 01 November 2010 |
| DOI | http://doi.org/10.1111/j.1467-856X.2010.00427.x |
| Subject Matter | Article |
Forecasting the Vote for a Third Party:
The British Liberals, 1974–2005bjpi_427634..643
Éric Bélanger, Richard Nadeau and Michael S. Lewis-Beck
While for multiparty systems the development of vote function forecast models for the incumbent
party and the official opposition party is commonplace, only rarely do these models try to forecast the
vote for a third party. Can third party vote shares be forecasted with reasonable accuracy? We
explore this question within the context of British politics. Our model proposes that the British
Liberal party vote is mostly driven by the extent to which the UK electorate approves or disapproves
of the official opposition leader. Our results are consistent with the idea that once the decision has
been made to punish the incumbent government, a voter must then decide whether to support the
official opposition party or another smaller party.
Keywords: electoral behaviour; UK; British Liberal party; forecasting models
By nature, electoral forecasting is primarily concerned with predicting whether or
not the governing party will be re-elected. That is to be expected as election
forecasters are interested first and foremost in knowing whether or not there will
be change in the governance of the country. In two-party systems such as the
United States, one does not need to know much more than that since there is only
one partisan alternative to the incumbent; if it is forecasted that the incumbent
party will not be re-elected, then that automatically means that the other party will
come to power. In multiparty systems such as the United Kingdom, the situation is
a bit more complex as there is more than one party in opposition. If the governing
party is predicted to be kicked out of office, how sure can we be that it is the official
opposition party that will step in? For this reason, most of the forecasting models
developed for the UK have typically tried to predict the official opposition party’s
vote share in addition to the incumbent’s (e.g. Lebo and Norpoth 2007; Nadeau
et al. 2009).
While for multiparty systems the development of vote function forecast models for
the official opposition party is very useful, only rarely do these models try to
forecast the vote for a third (or minor) party. A few models have been developed
that forecast support for the Front National in France (Auberger 2008; Evans and
Ivaldi 2008; Jérôme and Jérôme-Speziari 2010a) or for similar radical right parties
in other European countries (Evans and Ivaldi 2010). But in the UK example at
hand, the vote for the Liberal party—now the Liberal Democrats—has never been
forecasted. Vote intentions for the Liberals have been the object of forecasts (Sanders
2005a and 2005b) as has been the seat share going to the Liberals (Whiteley 2005
and 2008), but not the party’s actual vote share. Likewise, a number of studies have
been published that explain support for the Liberal party (e.g. Alt et al. 1977;
Lemieux 1977; Studlar and McAllister 1987; Clarke and Zuk 1989; Eagles and Erfle
doi: 10.1111/j.1467-856X.2010.00427.x BJPIR: 2010 VOL 12, 634–643
© 2010 The Authors. British Journal of Politics and International Relations © 2010
Political Studies Association
To continue reading
Request your trial