Summary of the impact

Based on his acknowledged research expertise in the areas of voting behaviour, electoral systems and survey methodology, John Curtice was widely consulted by the media before and after the 2010 UK general election to provide expert predictions on the likely outcome and to explain the results. By engaging with a variety of non-academic audiences, he informed public debate and understanding, and influenced pre-election planning by the UK Civil Service. He also contributed to the election night coverage by the three major UK broadcasters by accurately predicting the final result based on exit poll data. This informed much of the election night coverage, particularly on the BBC.

Underpinning research

Throughout an academic career spanning more than thirty years, Curtice has conducted extensive research into electoral systems, voting behaviour and survey methodology. Much of his research in the past 15 years has focused on developing innovative approaches for predicting and explaining the outcomes and implications of UK general elections. Two aspects of his research are of relevance here.

First, he has advanced understanding of how the electoral system used for UK general elections operates. The single member plurality system is often thought to ensure that one single party will win an overall parliamentary majority. The seats secured by the largest party are an exaggerated reflection of its voting lead over the second largest party, while third parties struggle to win any seats at all. However, these features are contingent. There needs to be a significant proportion of constituencies that are closely contested by the two largest parties, the system needs to be even handed in its treatment of those parties, while support for third parties should not be so geographically concentrated that they prove capable of winning many seats.

In his research, Curtice has demonstrated that these conditions have increasingly not been satisfied (reference 1). This finding allowed him to anticipate (reference 2) and subsequently explain (reference 3) the hung parliament produced by the 2010 UK general election.

Second, he has also been engaged in research on the methodology of exit polls. To be accurate, such polls must meet two main requirements. First, they need to be conducted in a sample of polling stations (precincts) that between them are representative of the country as a whole. Second, they need to translate the anticipated outcome in votes into an accurate forecast of seats won. These two requirements are particularly challenging in the UK because (a) the outcome of previous elections in individual precincts is not normally published, while (b) under single member plurality there is no arithmetic mechanism linking votes cast and seats won nationally. Together with David Firth, a statistician at the University of Warwick, Curtice developed a methodology to meet these challenges. They anticipated that an exit poll was more likely to be accurate if it estimated the change in each party's support since the last election rather than its level of support, because change in support varies much less from one precinct to another. However, this requires an estimate for each sampled precinct on the outcome of the previous election, which can only be supplied by a previous exit poll. Thus wherever possible, an exit poll should be conducted in the same precinct as last time. The resulting estimates of change can then be modelled and the resulting equations applied to the results of the previous election in each constituency in order to estimate the probability of each party winning each seat.

Curtice and Firth developed and applied this method to exit polls on the night of the 2005 election and subsequently analysed the reasons for its success (reference 4). Under Curtice's direction, this method was again implemented and evaluated in the 2010 election in collaboration with Dr Steve Fisher and Dr Jouni Kuhai of the London School of Economics (reference 5).

Key researchers at Strathclyde

John Curtice — appointed as Lecturer in 1988, now Professor of Politics in the School of Government and Public Policy. Director of the Centre for Elections and Representation Studies (founded in 2009).

References to the research

Details of the impact

Process from research to impact

Curtice has been a regular contributor to both print and broadcasting media coverage of elections and voting behaviour in Britain for over twenty years. He has also been a consultant to the BBC's election night programmes at every general election since 1979. This has ensured that his work on the British electoral system, and indeed his academic research in general, is widely disseminated to the general public and comes to the attention of politicians, civil servants, think tanks, political activists and journalists. At the same time, his understanding of the electoral system together with polling methodology has been called upon in the development and management of exit polls designed to enable broadcasters to forecast the outcome of an election as soon as the polls close. Curtice's research has had three distinct impacts.

A. Informing public debate and understanding of the British electoral system

In the run up to the 2010 general election, Curtice disseminated the key finding of his research, that the election was likely to result in a hung parliament where no one party had an overall majority, and the reasons why this was the case, to a range of non-academic audiences. He wrote regular articles for the The Independent, The Sunday Telegraph, and The Scotsman newspapers (e.g. source 1), wrote opinion pieces for The Economist and Prospect magazine, and made many broadcasts primarily, though not exclusively, for the BBC. Curtice's predictions and explanations of the electoral system were frequently cited in subsequent newspaper commentary about the likely results. For instance, Polly Toynbee in her Guardian column, 7^th April 2010, (source 2) drew on Professor Curtice's research in her assessment of the possibility of a hung parliament: "After all these years of waiting, the Liberal Democrats are closer to the brink of a hung parliament than for decades. There is, says Professor John Curtice, a 50% chance, as virtually every poll currently predicts it. There are reasons for doubt: the marginals appear to be swinging some 2% more in favour of the Conservatives. In quarter of those seats the Labour MP is standing down, losing the advantage a sitting MP usually gets of some 1,500 extra votes. But if ever there were a chance of no party winning outright, it should be now. The public welcomes a coalition: a Guardian/ICM poll showed 44% want a hung parliament, with only 29% preferring Conservative alone and only 24% Labour alone."

Alex Barker writing in the Financial Times, 27^th April 2010, (Source 3) made similar points based on Curtice's research:

"Experts are divided over what the rise of the Lib Dems will mean. Some argue that calculations suggesting Labour's set advantage are overstated because they are based on a uniform national swing. Others such as John Curtice of Strathclyde University say the rise of the Lib Dems — as it stands — will make no difference to the fact that Labour wins a higher proportion of seats to votes. Gordon Brown's proposal is to reform the system by moving to an "alternative vote" model where electors rank candidates according to preference. But this could actually accentuate the system's quirks, further exaggerating the bias towards the biggest party."

Through these various means his research reached a large audience and informed public debate in the run up to the election. In the immediate aftermath of the election, Curtice contributed to public understanding of the result and the implications for future elections and the future of the single member plurality electoral system. He wrote several articles explaining how the hung parliament came about in 2010 and why it could happen again. The key findings of his research and their implications for future elections were presented to MPs and others in a lecture at Portcullis House which was aired on BBC Parliament (source 4). In that lecture he explained why the electoral system can no longer be relied upon to deliver a single party majority in the House of Commons. Curtice's research was widely drawn upon by politicians and media commentators in the run up to the 2011 referendum on whether to change the UK electoral system to the alternative vote system (Source 5).

B. Influence on pre-election constitutional planning by the UK Civil Service

In addition to shaping public debate before the 2010 election, Curtice's pre-election predictions about a hung parliament were also taken seriously by government officials. Following a request by the then Prime Minister, the Head of the Civil Service published for the first time a detailed statement on the constitutional conventions that should be followed for the formation of government after the election. This statement, which has since been incorporated into the Cabinet Office manual (source 6), took direct account of Curtice's predictions. The then Cabinet Secretary Sir Gus O'Donnell confirms that:

"As Cabinet Secretary over the period 2005-2011, I was responsible for coordinating civil service preparations for the outcome of the 2010 General Election and specifically for the eventuality of a hung parliament and a potential coalition government. In doing so, we used a range of scenarios and purposely steered clear of accepting any predictions from polls. Nevertheless, we were aware that the polls meant we needed to ensure we prepared for all outcomes including a hung Parliament. We knew, from media reports, of the predictions of Professor John Curtice that the election was likely to result in a hung parliament in which no one party had an overall majority and why. As one of the leading UK academic experts on elections, Professor Curtice's research had a significant influence on our recognition of the need to prepare for a hung parliament." (Source 7)

C. Contribution to 2010 General Election night coverage on BBC News, ITN and Sky News

Supported by Fisher and Kuhai, Curtice directed the analysis and modelling of raw exit poll data collected by Ipsos-MORI and GfkNOP on polling day in 2010 on behalf of all three main UK broadcasting organisations (BBC, ITN and Sky), using the approach previously developed in collaboration with Firth. To an average election night audience of 8.202 million viewers (Source 8), at 10pm this analysis was broadcast simultaneously by all three broadcasters, accurately forecasting that: (i) the Conservatives would win 307 seats (resulting in a hung parliament); and (ii) that the Liberal Democrats would win fewer seats than in 2005. Though the first prediction did not come as a surprise, the second contradicted most expectations. The results of Curtice's analysis informed much of the broadcasting coverage of the initial result of the election including on-air statements by prominent politicians (Source 9). Subsequently Curtice was responsible for providing for the BBC analysis of the election results as they were declared, including above all their apparent implications for the likely eventual overall outcome. Sue Inglish, Head of Political Programmes at the BBC confirms that:

"Professor Curtice was commissioned to undertake these tasks because of his expertise in voting behaviour, electoral systems, the geography of party support and survey methodology. The work of Professor Curtice and his team made a significant contribution to the election night coverage of all three broadcasting organisations in general and the BBC's in particular. The exit poll provided us all with the invaluable intelligence that (a) there was likely to be a hung parliament, and (b) that the Liberal Democrats' performance was not going to meet most people's expectations, intelligence that shaped the content and tone of our coverage, including not least in interviews with senior politicians during the course of the evening. Meanwhile, his subsequent commentary enabled the BBC to explain clearly and accurately to its many audiences, the significance and implications of the election results as the story unfolded over the following 24 hours" (Source 10).

Sources to corroborate the impact

1. J. Curtice, `Battle narrows down to the key marginals', The Independent, 10^th March, 2010,http://www.independent.co.uk/voices/commentators/john-curtice-battle-narrows-down-to-the-key-marginals-1918835.html
2. P. Toynbee, `Liberal Democrats could be tainted by Tory association', The Guardian, 7^th April, 2010, http://www.theguardian.com/commentisfree/2010/apr/06/liberal-democrats-nick-clegg-hung-parliament
3. A. Barker, `Creaking poll system set for biggest test', The Financial Times (London), 27^th April 2010, http://www.ft.com/cms/s/0/f5f60016-5193-11df-bed9-00144feab49a.html#axzz2eh54kLa9
4. Hansard Society, Britain Votes 2010, public lecture by Professor Curtice at Portcullis House, Westminster, 15^th September 2010, http://www.hansardsociety.org.uk/blogs/recent_events/archive/2010/09/17/2708.aspx
5. J.Purnell and J. Forder, `AV: For and against', 23^rd March 2011, http://www.prospectmagazine.co.uk/magazine/av-for-and-against/#.UjHu58aTjng
6. Cabinet Office, Chapter 6: Elections and Government formation, 2010, https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/60646/election-rules-chapter6-draft_0.pdf
7. Statement from Sir Gus O'Donnell, former Cabinet Secretary to the UK Government
8. J. Deans, `TV ratings: Channel 4's alternative election night outshines ITV in ratings', 7^th May 2010, http://www.theguardian.com/media/2010/may/07/channel-four-election-ratings-outshine-itv
9. BBC Election 2010 Broadcast, 6^th-7^th May 2010, available at: http://www.youtube.com/watch?v=BiF1D7Uh1QA. Key segment is `Election 2010 - Part 1' at time 4:52 to 7:51.
10. Statement from Sue Inglish, Head of Political Programmes for BBC News