Impact case study database

A well-functioning democracy requires well-informed citizens and responsive political parties. The BES team worked to improve media coverage of elections by integrating academic analysis into ITV’s election coverage, engaging with parliamentarians and the wider media. Its input into the British Polling Council’s inquiry into the 2015 polling miss, directly contributed to changes in polling methodologies used by pollsters, and key figures from both main political parties have attested to the impact the BES research has had in supporting the health of British democracy.

1. Improving media coverage of elections by putting academic analysis at its heart

Central to the BES team’s role in making a better informed democracy was the collaboration with ITV news, led by Green (University of Manchester until 2018) and Prosser. Drawing on the extensive research into British voting behaviour [1-4], Green and Prosser actively shaped ITV’s coverage of the 2015, 2017, and 2019 General Elections, and the 2016 EU Referendum, by putting the research results and analysis at their heart. Traditional television coverage of elections has relied solely on analysis of aggregate level data which can lead to false inferences about the behaviour of individual voters and groups. Green and Prosser introduced a new approach, using survey data collected before and during the election campaign [1] to better inform ITV’s audience about voter’s opinions and motivations. Summarising what the BES teams input and data brought to ITV’s elections coverage (which was nominated for Royal Television Society and BAFTA awards) the ITV News Editor wrote that this “has been central to our election and referendum programmes and provided unique benefits to viewers - informing them earliest of results - and crucially, playing a fundamental role in ITV's coverage in explaining results in a clear, compelling and rigorous way and bringing clarity to the most complex issues of our time for viewers around the UK” [A].

The UoM team placed the BES research as a principal voice in understanding the 2015 – 2019 elections and the 2016 referendum through extensive work with a wide range of media channels – informing the public about British elections and public opinion with high- quality survey data and analysis [A, B, C, D] that went beyond the headline polling numbers that dominate media coverage. The team engaged in events before and after the 2015 General Election, the 2016 EU referendum, the 2017 General Election, and the 2019 General Election (e.g. BES journalist events, Political Studies Association media briefings, the All-Party Parliamentary Group for Social Science and Policy, three major BBC Parliament events, one-to-one meetings with key broadcasters), and wrote commissioned articles for newspapers and the BBC online (related to the research in [2-4]). According to Factiva, in the six years since the Manchester team took over the BES, there was a greater than ten-fold increase in print media articles mentioning the BES (2014-2020) compared to the previous 6 years [B].

2. Putting political polling on a firmer methodological footing

Political polling plays a central role in modern democracy by informing political actors, the media, and citizens about voters’ preferences and opinions. Just as political events cry out for reliable data and greater public understanding, polling misses have undermined public confidence in survey research. BES research led the way in diagnosing the problems with polls in 2015 and identifying methodological improvements [5, 6]. As the Secretary of the British Polling Council wrote: “At a time when the whole concept of sample surveys was called into doubt, the ability of the BES to show that well-designed and well-conducted surveys can still represent the population with a high degree of accuracy helped polling organisations both by helping them understand what might have gone wrong and how to avoid similar problems in future” [C].

Drawing on the face-to-face survey data, research by Mellon and Prosser [5] was the first to suggest that unrepresentative samples were the cause of the 2015 polling miss [D], a conclusion later supported by the British Polling Council inquiry into the polling miss [6], in which Green took part, and accepted by the polling industry as a whole [B, C]. The Manchester team actively sought to inform the public and pollsters as to the reasons for the polling miss. Mellon and Prosser wrote a series of blog posts on the BES website and the influential Monkey Cage ( Washington Post) and actively promoted their findings in the media (BBC Newsnight, The Guardian, The Independent, Huffington Post). They presented at industry and academic conferences, and engaged directly with the heads of research at prominent polling firms, including Pew, YouGov, ICM, TNS, and GQR.

The Manchester-based research has caused leading pollsters to change their methodology. The importance of BES data in underpinning improvements to political polling is visible in its use in providing weighting targets for many polling companies [E]. The Head of Methods at Kantar Public wrote, “The single most important contribution for us was the suggestion of “turnout weighting” in Mellon and Prosser’s paper “Missing Non-Voters and Misweighted Samples: Explaining the 2015 Great British Polling Miss [5] ” [F]. BES data also plays a key role in new polling innovations, such as YouGov’s MRP (Multi-level Regression and Post-stratification) model which (almost alone amongst its competitors) accurately predicted the outcome of the 2017 election, in part by using BES data to estimate the likely turnout pattern [G]. The Secretary of the British Polling Council wrote “Individual BPC members have also made extensive use of the BES in conducting their own internal investigations into how they might avoid similar polling errors in future... the BPC as an umbrella body is happy to speak for its members and say that many have found the BES very useful in their own work” [C].

The research and engagement undertaken by the Manchester BES team has been recognised for its role in understanding the polling miss leading to more accurate polling methods adopted by leading pollsters. The BES was a finalist for the 2017 ESRC Celebrating Impact Prize, in the Outstanding Impact in Society category [H].

3. Supporting the health of British democracy through reliable data

BES research has impacted on British democracy by providing high-quality data [1] that informs the decision making of key political actors. Prosser and Mellon submitted written evidence to the House of Lords Select Committee on Political Polling and Digital Media, and Green gave oral evidence, and co-authored the World Association for Public Opinion Research (WAPOR) submission to the same inquiry. The behaviour of the two main political parties in the UK, the Conservatives and Labour, has been shaped by their use of BES data and analysis to understand voter movements, resource allocation during campaigns, and inform post-election analyses presented to stakeholders.

[Text removed for publication] [I].

Likewise, the General Secretary of the Labour Party, Jennie Formby, wrote, “Since the current BES scientific leadership group assumed management of the project, we have used BES information in planning and preparing for both the last two general elections” [J] . Specifically, she states that the BES research “has been used to inform Labour’s approach, strategy, and resource allocation in our campaigns… in particular, it has been used: to underpin targeting decisions, drive resource allocation, inform subsequent post-election reviews, and enrich understanding and discussions of political attitudes, trends and patterns in British elections” [J]. Summarising the importance of the BES, she wrote that the BES “ has an impact, far beyond the university sector, in ensuring that our democracy is healthy and well-served by political party activities which reflect the concerns and priorities of the British electorate” [J]. After the 2019 election the Labour Together Review commissioners wrote that BES data, “was incredibly helpful as our review aimed to understand Labour’s decline in support over a longer period…. Although they were hard to face, our conclusions were evidence-based and accepted across the party. We are certain that the deep respect for the work of the British Election Study from people across Labour’s political sphere helped us to achieve this” [K].