Voters in marginal constituencies: average or different? An analysis of the demographics, values and ideology of voters in UK marginal seats

Background
One often critiqued facet of Britain's First Past the Post electoral system is it divides up the country into 'safe seats', said to be almost impossible to win, and 'marginal seats', said to be the true battlegrounds of British elections. The Electoral Reform Society has used this distinction to point to the relatively large changes in election results that can flow from shifts in small numbers of votes in marginal seats, a common feature in plurality systems (Garland & Terry, 2017), (Terry, 2013).

The outsized attention paid to marginal seats been examined by academics analysing campaigns. There are important democratic and political ramifications to this concentration of attention and activity on marginal seats.

Academics have discussed whether voter behaviour in marginals is different. Some studies have suggested that tactical voting is more common in marginal seats, given its higher efficacy (Lanoue & Bowler, 1992). Many studies demonstrate higher turnout in marginal seats (Denver & Hands, 1974), (Selb, 2008). Research has also shown that politicians tend to respond to this incentive with higher constituency service (Heitshusen, Young, & Wood, 2005).

If, as electoral reformers argue, the focus on marginal seats created by the British electoral system means voters in safer seats are underrepresented, with party and government policy aimed at winning the marginals, who is it being underrepresented?

Research Questions
The central hypothesis is that voters who live in more marginal seats have different demographics, attitudes, behaviour and ideology to those who live in safe seats.

Do voters in marginal seats share the views of voters outside them? Are marginal constituency voters representative of national ones in terms of their values and positions? Electoral reformers critique that safe seats are unrepresentative, but do safe seats have unusual mixes of voter value groups and are some underrepresented as a result?

Impact
This question has major ramifications for the literature on party campaigning, policy formation and electoral systems, and for organisations outside academia.
At least 2 reports will be produced for the Electoral Reform Society, with accompanying events and dissemination through their networks as appropriate. Work will also be targeted at academic journals.

Methodology
The BES datasets will form the core data for the study. Variables which measure constituency marginality provide the central measurement necessary for this analysis. With sample sizes of 30-35,000, BES datasets provide significant amounts of individual level data which can be used to study voters in marginal seats.

The key goal is to examine the views of the electorate in marginal seats. The BES provides a variety of question banks which can be used to measure these. The package of values questions will be used to analyse how voters in marginals differ on these measures. Questions on issue importance and self-placement on scales will also be used for comparison with safer seats. These comparisons provide the heart of the analysis.

The proposal is to take a case study approach, looking at the UK, allowing for greater depth. The work will cover between 2010 and 2017 election. This period has multiple advantages. Firstly, boundaries have not changed since 2007, removing an important source of variation in seat marginality, this making it easier to isolate and track political and social differences between marginal and non-marginal seats. Secondly, volatility in this era means that it is possible to observe major changes in the party system, starting with the 2 and a half party system that characterised British politics since the 70s in the 2010 election, through to a more multi-party distribution of votes at the 2015 election and finally the two-party dominance of 2017.