Skip to main content

The Impact of Coalition on the 2015 General Election

The British Election Study Team

By Jane Green and Ed Fieldhouse

There has been a lot of talk about how the 2015 election is different: more voters than ever before look set to vote against the major Westminster parties; the fragmentation of the two-party system enhances uncertainty about the assumptions of uniform swing; and the three major parties are all unpopular. There has been a lot of commentary on all these things, but what hasn’t been analysed is how the impact of a five year period of coalition might change the nature of the electoral contest and result come May 7th. This particular political context could have very important implications.

Britain’s majoritarian electoral system has increased the probability of single-party government, whereas a proportional electoral system would favour votes for challenger parties and increase the probability of coalitions. We contend here that the experience and expectation of a coalition government may be increasing the incentives and likelihoods of voting for challengers, even under – and in despite of – the incentives of a majoritarian first-past-the-post electoral system.

There are four reasons to expect that the British public’s experience of coalition may be encouraging voters to switch to challengers.

1. Coalition government blurs policy and ideological distinctions between coalition parties, making a party with a clearer policy-based platform more appealing to voters. This might be based on a clearer and distinctive position in left-right terms, or on a clearer appeal on cross-cutting issues such as immigration, Europe or devolved powers. The likelihood of a party entering a coalition may further blur the dividing lines for voters – and that may also encourage them to seek a better defined alternative.

2. The prospect of another coalition – and of uncertain coalition negotiations – makes it less likely that a voter will get their most preferred outcome if they vote for one of the major parties. In this sense, coalition government may make some voters think a vote for a major party is in fact wasted.

3. The chance of a hung parliament means that a minor party with minor representation may have disproportionately high influence in parliament. In this sense, coalition alone (without a proportional electoral system) makes a minor party vote less likely to be wasted than when single party government is anticipated.

4. Green, Liberal Democrat and SNP voters are more positively predisposed than other voters to coalition government in general. We find this, for example, using BES survey questions asking respondents what they think about the effectiveness of coalition government. These voters may see they have the most to gain from coalition. In the case of SNP supporters in Scotland – the actual experience of coalition might inform this positive attitude. These voters won’t be easily encouraged to vote for a major party that ensures a single-party government. But on the flip-side, UKIP voters see coalition governments in general much less favourably. This may cause UKIP voters to punish the parties currently in coalition – and those parties who would consider forming a coalition government. Ironically, UKIP voters may thereby enhance the probability of an outcome they least favour (because a large scale vote for minor parties makes the prospect of a hung parliament more likely).

British Election Study data enables us to explore the potential consequences of coalition government. We devised a question designed to examine two expectations – how close the race is expected to be and whether respondents think Labour or the Conservatives will form a government on their own. For this analysis we look at the latter expectation. We use an early preliminary sample (around 16,000 respondents) of pre-election online data gathered between 6-13 March 2015 (the initial sample of wave 4 of the BES internet panel). Respondents were asked, ‘How likely is it that either of these parties will win more than half of the seats in the General Election so it will be able to form a government on its own?’ Respondents placed the likelihood for Labour and Conservative separately, on two scales where 0 = very unlikely and 10 = very likely. We divide the sample into a) respondents who think either of the major parties look likely to form a government alone (points 6 -10 on the scale) and respondents placing both parties at 6 or below (they expect a hung parliament). 41% of our sample thinks neither the Conservatives nor Labour will win a majority. We then compare vote intentions according to these expectations. Table 1 reveals the differences in vote intentions for the major parties (each party loses votes if respondents point to a hung parliament) and for the challenger parties (each gains votes).

Table 1: Vote intention by government outcome expectations

Vote intention  and coalition

Note the scale of the differences. Support for Labour and the Conservatives is lower by about a quarter amongst those expecting a hung parliament. The Liberal Democrats and Greens double theirs; UKIP adds a third and the SNP see a quadrupling of support if respondents think a hung parliament is the likely outcome after 7th May. These simple findings suggest that the experience of coalition – and the likelihood of coalition – have likely had a profound impact on the tendency of voters to consider voting for a challenger.

The ongoing significance of this is two-fold. (1) There are unintended consequences of coalition for the major parties: they lose votes and also make coalition more likely – because fragmentation of the two-party vote leads to greater hung parliament likelihood; (2) The campaign is likely to enhance the view that neither party can win an overall majority, so there could be more vote losses away from the two major parties and towards their rivals, if more voters become convinced of a knife-edge outcome.

There is a further way coalition – and the expectation of coalition – may influence voting decisions. Different party voters will prefer different coalition bedfellows, and this may influence strategic voting and the likelihood of switching from different challenger parties to mainstream parties, should different outcomes of the May 7th general election emerge as more likely. We ask respondents to rank the outcomes they prefer: A Conservative led coalition, Conservative majority, Labour led coalition, Labour majority, or ‘some other government’. Table 2 reveals the rank orders by 2015 party vote intention, again using the preliminary wave 4 BES sample.

Table 2: Preferred Governments by 2015 Vote Intention

Preferred govt by vote intention


Table 2 reveals fascinating results. Liberal Democrat voters would prefer a Conservative led coalition over all other alternatives – including the catch-all ‘some other government’ category, which we presume is interpreted as a government led by the respondent’s preferred party, or a combination of parties. SNP, Plaid, UKIP and Green voters would all first prefer ‘some other government’. But among the realistic outcomes in Table 2, the second preferences of these party voters are highly illuminating. SNP voters would prefer a Labour led coalition. UKIP voters would prefer a Conservative majority. Green voters would prefer a Labour led coalition. It is also interesting that not all Conservative voters would prefer a Conservative majority, and not all Labour voters prefer a Labour majority. But only small proportions of Conservative and Labour voters currently therefore look likely to be happy with the outcome of the May 7th result.

The potential significance of these findings comes from understanding the voters likely to be disappointed by different coalition negotiation outcomes, and what those negotiating parties might therefore take into account. The experience of the Liberal Democrats in a formal coalition between 2010 and 2015 serves as a caution to any possible coalition partners after May 7th. These parties will be keen not to alienate their supporters, and so we can understand, for example, the SNP’s position vis a vis the Conservatives and Labour (to rule out any deal with the Conservatives but to leave open the possibility of a confidence and supply arrangement with Labour). Liberal Democrat voters look fairly evenly split in their potential acceptance of either a Con-Lib coalition or a Lab-Lib coalition, although marginally more in favour of a Conservative minority government. This makes sense given that remaining 2015 Liberal Democrat voters are those who are disproportionately comfortable with the Con-Lib coalition of the last five years whilst their more left-wing supporters have long since deserted them. What is particularly intriguing is the preference of UKIP voters for a single-majority Conservative government. This most likely reflects three things; (i) the fact that the majority of UKIP voters are former Conservative voters;  (ii) a desire not to see a Con-Lib coalition or a Lab-Lib coalition (or other configurations), given UKIP’s ideological distance from these left-of-centre parties; and (iii) a dislike towards coalition government in general. As well as having implications for parties and their potential coalition negotiations, the prospect of coalition changes the decision calculus of voters. We cannot anticipate the degree to which this will be the case, but we can anticipate with some confidence the importance of voter’s experience with coalition for understanding the 2015 general election and their expectations of future coalitions, as well as the impact of these expectations on the governing outcome post May 7th 2015.