Abstract
Keywords
This article discusses the findings of a detailed content-based case study about the impartiality of television news coverage during the United Kingdom’s 2016 European Union (EU) referendum. We explore the successes and failures in the broadcaster’s attempts to provide news coverage that was both fair and informative, and, in so doing, suggest the need for a serious reappraisal of the way in which impartiality is interpreted by broadcasters.
UK broadcasters have to abide by ‘due impartiality’ guidelines in news and current affairs programming. During the 2016 EU referendum campaign, the BBC Trust oversaw BBC content, while the Office of Communications (Ofcom) regulated commercial broadcasters (since April 2017, Ofcom is also responsible for the regulation of BBC content). According to the (recently revised) BBC editorial guidelines, ‘Due impartiality is often more than a simple matter of “balance” between opposing viewpoints. Equally, it does not require absolute neutrality on every issue or detachment from fundamental democratic principles’. Similarly, Ofcom’s code states, ‘“due impartiality” does not mean an equal division of time has to be given to every view, or that every argument and every facet of every argument has to be represented’.
1
The BBC also issued specific EU editorial guidelines, which asked editors to pursue a ‘broad balance’ approach to reporting the campaign. According to the guidelines,
Due impartiality is not necessarily achieved by the application of a simple mathematical formula or a stopwatch, but the objective – in a referendum with two alternatives – must be to achieve a proper balance between the two sides … However, referendums are seldom fought purely on the basis of just two opposing standpoints – on each side, where there is a range of views or perspectives, that should be reflected appropriately during the campaign.
In other words, balance should not just be achieved by allowing the two camps to argue between each other, but must also consider debates
Broadcasters faced an interesting challenge in balancing the perspectives of both sides of the campaign. In one sense, like most referenda, it presented voters with a simple binary choice, with two opposing Remain and Leave campaign groups –
Behind these cross-cutting debates lay a series of substantive claims about, for example, the impact of the EU on employment, wages, public services and economic growth. While some of these claims were speculative and difficult to evaluate, others could be measured against expert evidence and testimony. Broadcasters could not assume high levels of public understanding about these issues: a Eurobarometer survey conducted in May 2016 found that only 27% of people in the United Kingdom could answer three basic questions about the EU – lower than in any of the other 28 member states except Latvia (Hix, 2015). There was, in this sense, a need to educate as well as inform.
Of course, broadcasters had a role to play in allowing both sides of the campaign an equal platform to articulate their message to voters. But the ‘due impartiality’ guidelines also encourage editorial judgements to allow journalists to rigorously test the claims and counter-claims of each campaign group. Our study offers a systematic assessment of how balanced the coverage was of both campaigns as well as how far journalists exercised their editorial judgement to verify the arguments put forward.
Interpreting impartiality: Constructing balance
There is an extensive academic literature focussed on examining ‘bias’ in news reporting (Groeling, 2013) – widely seen as a critical issue in maintaining a healthy and democratic public sphere. Most Western governments regulate their media systems, particularly in broadcasting (which remains a dominant source of information for most people – Ofcom, 2016), applying the notions of impartiality and objectivity to the regulation of news. While they intersect, both concepts reflect different empirical goals. Objectivity foregrounds the investigation of truth, whereas cruder notions of impartiality assume that truth is always contested. Editorial judgement allows journalists to plot their ways through this philosophical conundrum, searching for the most truthful version of a story against a backdrop of competing claims.
During elections, these editorial decisions are brought into sharp relief: broadcasters must give the key players the opportunity to put their case, while retaining the licence to examine those cases critically. The first of these is, of course, easier to measure. So, for example, Loughborough University’s studies of past UK general elections have included ‘stop-watch balance’ as a key criterion for assessing whether one party or leader receives more attention than their opponents (Deacon et al., 2005). This represents an important measure of representation, as well as revealing which parties and policies are marginalised. Regulators have previously allocated airtime to parties based on the level of past electoral success (Semetko, 2000) or other benchmarks (Hopmann et al., 2012), rather than rely on news values to determine coverage. Recent work on the 2015 General election (Cushion and Thomas, 2017) suggests that broadcasters exercise a degree of freedom to interpret impartiality according to their editorial judgements about news selection.
Viewed in this context, impartiality is not simply a measure of who is able to speak, but which side is allowed to set the agenda. For our purposes, this is complicated further by the nuances of the EU referendum debate: while both campaigns tended to focus on particular issues, there were clearly different agendas
Our findings should also be considered against the wider backdrop of how UK news media have reported the EU over several decades. The EU has long been the bête noir of a significant section of the UK press. While far less vituperative, research has shown that broadcast news has
Method
We examined UK evening television news bulletins, including BBC News at Ten, ITV News at Ten, Sky News at Ten, Channel 4 at 7 pm and Channel 5 at 6.30 pm. Each channel, of course, broadcast a wider range of programming over the campaign in formats that allowed greater analytical detail or longer interviews than a daily bulletin can reasonably supply. But many of these programmes are watched by small audiences, whereas the evening bulletins are the most watched format of television news and taken to be programmes of record. Since millions of people in the United Kingdom are regularly exposed to nightly bulletins, it makes them one of the most important sources of information about the EU referendum campaign. Our long experience of analysing broadcast news output also suggests that while there are undoubtedly variations between different news programmes in style and tone, the main evening flagship bulletins reflect the channel’s considered editorial position.
We monitored these news bulletins over the 10-week campaign (from 15 April to 22 June 2016). The unit of analysis was the ‘news item’, categorised by the type of convention (a live two way, say, or an edited package) rather than by story topic (this allowed us to isolate each distinct reference to the EU). We examined both on-screen contributions (when they are seen and speak on camera) as well as indirect sources (when a journalist refers to them, such as ‘the Leave campaign said today …’).
Our content analysis began by quantifying the range of issues addressed by broadcasters and the types of sources informing coverage. We then considered the role and treatment of every statistical claim made during the campaign (instances where a source or journalist referred to any data or figures in the context of the EU referendum). Our aim here was to both quantify the presence of statistics in coverage, and assess
Overall, we examined 571 news items, 1582 sources and 517 statistical claims. 2 Approximately 10% of the sample was recoded to check the reliability of the variables. The level of agreement between two coders for all variables was above .92 and, according to Krippendorff’s alpha, achieved high levels of reliability (with each variable scoring .88 or above).
Interpreting the impartiality of issue and source balance
We began by considering the balance of issues in news items about the EU referendum. As we have suggested, debates about leaving or remaining in the EU were potentially wide-ranging. Of the 571 items we identified in television news coverage of the EU referendum, around four in ten were about the
Story subject of news items about the EU referendum.
EU: European Union; NHS: National Health Service.
If we include economic forecasts and trade agreements, the dominant topics on all bulletins related to the economy (around 20%). This covered a wide range of economic indicators, and sometimes intersected with other issues, such as the free movement of people being contingent on access to the EU’s single market. If we exclude these sub-topics, the economy and immigration received roughly the same level of attention in television news coverage (about 11%) – far more than other topics. There were some minor differences between bulletins, the BBC being the most immigration-oriented, while Sky News and ITV carried more items principally about the economy. But perhaps more striking was how similar broadcasters were in balancing items about the economy and immigration, particularly Channels 4 and 5 which were proportionally identical on these two issues (14.7% and 10.4%, respectively). Loughborough’s study of EU coverage found a similar dual focus – at the expense of most other issues, ‘including the environment, taxation, employment, agricultural policy and social welfare’, as well as the potential consequences of a Leave vote, not least on the possible departure of Scotland from the United Kingdom (Deacon et al., 2016).
In short, the United Kingdom’s evening television news bulletins gave roughly equal attention to those issues – the economy and immigration – that both sides of the official campaigns chose to emphasise. This suggests that broadcasters allowed the main campaigns to set the agenda – and our analysis below will explore how often bulletins challenged them on these issues.
The next part of our analysis examined the sources informing television news coverage, including the balance of voices between advocates for Leave and Remain. By this measure, the broadcasters were fairly evenly balanced, with the total average difference of only 3.7% between Leave and Remain. Most – apart from ITV – gave marginally more time to Remain sources (see Table 2). We see a similar pattern with indirect sources (Table 3) proportionally balanced between opposing camps (with a difference of just 3.2% in television news coverage generally, in this case tilted in favour of Leave). There were greater disparities between channels, but these were not dramatically different or uniform (roughly between 4% and 12%). Overall, our findings indicate that actors representing the Leave and Remain campaigns were fairly well balanced on UK television news over the 10-week campaign. It also shows that
On-screen sources supporting, opposing or unclear perspectives towards continued EU membership.
EU: European Union.
Indirect sources supporting, opposing or being neutral towards continued EU membership.
EU: European Union.
The dominance of supportive or oppositional voices towards EU membership indicates that, once again, the broadcasters (especially the BBC and Channel 5) chose to cover the referendum by focussing on the two official campaigns, rather than using more independent sources that were less adversarial. While journalists might argue no source is truly independent, they could (and did) draw on competing expert voices to help form a judgement about the veracity of claims made by both sides of the campaign. Moreover, where possible, they could interpret the weight of expert knowledge about particular issues. This is, after all, common practice in news reporting, where economic expertise is widely used, and coverage is often based on assumptions held by a majority of economists (Lewis and Thomas, 2015). However, Table 4 shows that knowledge-based professions, such as think tanks, economists and academics, made up a small share of contributions, whereas the clear majority (62.2% of sources) were politicians representing the Leave or Remain campaigns.
Type of on-screen sources in television news coverage of the EU referendum.
EU: European Union.
If we examine the party affiliation of politicians appearing across all television news bulletins, a significant imbalance emerges: an overwhelming majority – 71.2% – was from the ruling UK Conservative party, compared to Labour’s (the Official Party of opposition) 18.4%. UK Independence Party (UKIP) – a right-wing Euro-sceptic party – represented 7.6% of sources, with just 2.8% left for other parties to contribute (particularly notable here is the absence of the Scottish National Party (SNP), whose pro-EU stance combined with a suggestion that a Leave vote might trigger another independence referendum). The distribution of party political sources was broadly similar between bulletins, with Conservative party sources most dominant on Sky News and Channel 5 (79.2% and 78.2%, respectively; or 86.7% and 84.6% including UKIP). Loughborough’s study across press and broadcast sources found a similar imbalance (Deacon et al., 2016).
So while the balance between Remain and Leave contributions was relatively even across evening bulletins (see Tables 2 and 3), party political voices were
‘I think people in this country have had enough of experts’ (Michael Gove, Leave campaign) – Accepting or challenging statistical claims
During the campaign, journalists and commentators often complained about exaggerated or dubious claims made by rival camps, with statistics often misused to support or counter opposing arguments. The parliamentary Treasury Committee, for example, criticised some of the claims made by both campaigns, while the chair of the UK Statistics Authority singled out the Leave campaign for its misuse of statistics (BBC News, 2016). For both, the most glaring example was the Leave campaign’s oft-repeated claim (painted on the side of their campaign bus) that the UK government sends £350 million to the EU every week – a figure that took no account of the United Kingdom’s rebate or the various grants and subsidies flowing from the EU to the United Kingdom. Despite independent sources challenging the Leave’s claim about the UK government spending £350m per week on EU membership, an Ipsos MORI (2016) survey found that almost half of respondents believed this was true just days before the election.
In order to explore whether bulletins accepted or challenged statistical claims made by sources over the 10-week campaign, we examined every reference to a statistic either by a (on-screen or indirect) source or by a journalist that attributed the source. By statistical claim, we refer to instances where a source or journalist referenced a statistic (or statements related to figures), such as the Leave campaign’s claim about the cost of EU membership.
Overall, we identified 517 statistical claims, with a high proportion of all EU related news items – 42.4% – featuring a statistic. Our research on the broader use of statistics in broadcast news established that this is around twice the average proportion of statistical references in news items (Cushion et al., 2016). The use (or misuse) of statistics, in other words, was very much a feature of this campaign. Table 5 shows there was some variation between bulletins, with Channel 4 featuring the highest proportion of statistics and Channel 5 the lowest.
Reference to a statistic within a television news item about the EU referendum.
EU: European Union.
Table 6 shows that 43.7% of statistical references were made by politicians on the Leave or Remain side, or the official (and unofficial) campaigns. If we include UK political institutions (which was overwhelmingly made up of UK government treasury department sources, which appeared to favour the Remain campaign) then over half – 54.7% – of all statistical references were made by politicians or political bodies. If we exclude journalists, this figure rises to 69.0% of all statistical claims.
Statistical references made by sources or journalists in television news coverage of the EU referendum.
EU: European Union; ONS: Office for National Statistics.
To make sense of these figures, it is useful to compare them with statistical claims made during routine news coverage (outside the referendum period). Our research on the broader use of statistics in broadcast news shows that the overwhelming majority of statistical references – 81% – are made by journalists (Cushion et al., 2016). During EU referendum coverage, this proportion dropped dramatically to 15.3%. In other words,
When journalists did use statistics, they were usually unattributed and, again, typically involved attempts to ‘balance’ one statistic against another. So, for example, a report on BBC News at Ten on 14 June stated that
In the middle of the campaign we got those figures showing that last year
This statement asserts more than it explains, particularly about the benefits of immigration (something that, surveys demonstrate, is poorly understood, with many people assuming that immigration has a negative impact on the economy). But it is fairly typical of attempts by reporters to use statistics to balance competing perspectives about immigration.
If journalists are excluded, Table 7 shows, once again, most broadcasters (notably on BBC, ITV and Channel 4 bulletins) were doggedly even-handed in balancing the statistical claims supporting Leave and Remain. This, does, however, conceal a notable difference between the breadth of the make-up of those making claims. Our findings indicate that a significant proportion (42%) of the reported claims favouring Remain came from business, the academy, trade unions, financial institutions or other professions. Anti-EU sources, by contrast, were composed overwhelmingly of politicians or Leave campaigners (97.8%). In other words, the Remain side had a range of actors expressing viewpoints using data or figures, while any statistical claims favouring the Leave camp came almost entirely from political figures. In this sense, the ‘imbalance’ favouring Remain on Sky News actually
Statistical references supporting, opposing or unclear perspectives towards continued EU membership.
EU: European Union.
This finding, we would argue, demonstrates another problem with the narrow definition of impartiality broadcasters chose to adopt. In most sections of civil society – economists, business people, trade unions, academic experts and so on – there were clear majorities (in some cases, such as among economists, overwhelmingly so) in favour of remaining in the EU. So, for example, 88% of economists (based on a survey of more than 600 respondents from academia, the city, industry, small businesses and the public sector) believed leaving the EU would harm the British economy – Sodha et al., 2016). And yet this consensus got caught up in the broadcaster’s balancing act, and the weight of this support was squeezed into one side of the ledger, giving them equal weight with (less well supported) counter-claims.
The decision to represent independent expertise as, in effect,
The preponderance of statistical points made by politicians rather than independent sources – especially on the Leave side, which depended almost exclusively on political sources – raises the question of journalistic scrutiny: to what extent were their competing facts and figures challenged or contextualised in news coverage? There were undoubtedly occasions when statistics when journalists did question or contextualise claims, as this example indicates,
In streets up and down the country, people are trying to work out if that figure of 350 million is accurate. Well, in 2015 we received about £85 million pounds a week thanks to a rebate negotiated in the 1980s. We also received £88 million in payments to sectors like agriculture, and on top of that it was about 27 million pounds paid to the private sector. Add those together and take them away from that 350 million figure and you end up with a net contribution to the EU of more like 150 million pound a week. (9 June 2016, Channel 5)
This, however, was not the norm. Table 8 shows just under one in five statistical claims were put into some context (with Sky News being the least likely to do so, although, as Table 3 shows, Sky were also most likely to use independent sources). Most of the context was supplied by journalists or by politicians/campaign groups – 57.6% and 17.9%, respectively – with few instances of reporters using independent expertise used to explain or analyse competing claims.
The proportion of statistical claims challenged or contextualised in television news coverage of the EU referendum.
EU: European Union.
Around the same proportion of statistical claims – just over one in five – was challenged in some way by a journalist, political or independent source. However, in keeping with the overall flavour of coverage, most of these challenges – 65.2% – came from politicians or campaign groups themselves, with few instances of the use of independent sources to verify claims. When less partisan sources Institute of Fiscal Studies spokesperson: We’ve already looked at the numbers and our conclusion is that the most realistic numbers to look at in terms of direct contribution we make, net of how much is spent here is a 150 million pound per week, about £8 billion per year. (9 June, 2016, Sky News)
As we have seen in the narrow rubric used by broadcasters, however, such independent scrutiny would have been notched up as a contribution favouring the Remain campaign – thus requiring a counter-claim from the other side.
Journalistic challenges to statistical references were not routine – made in 17.6% of claims – with very few occasions when a number or figure was We’ve done our own bit of research to find out if membership really does cost each of us £252 a year, as claimed by many Brexit campaigners. Are we giving Brussels £350 million a week, whilst receiving nothing in return? Here’s our latest fact check. [statistics appear on screen without voice over] Does the EU cost every person in the UK £250 a year? Leave Campaigners say the membership fee is £350 m a week – Leave campaign. That would be 0.6% of national income. Or one seventh of UK health spending (13.9%). That’s £252 a year per person in the UK, They estimate this because last year the UK gave £18.8 bn to the EU. But what they leave out is that we get quite a lot of money back, there is a rebate every year, negotiated by Margaret Thatcher. Billions are given to the UK to spend on things like farming (£9.8 bn). Even more comes back in grants for universities and business (£5.7 bn) – IFS. Bringing the cost down from £252 to £89 a year. (19 April 2016)
However, this more detailed statistical breakdown was very much the exception rather than the rule.
The campaign was thereby dominated by
The proportion of news items with a statistical claim including a tit-for-tat in television news coverage of the EU referendum.
EU: European Union.
Over half of these statistical tit-for-tats involved rival camps either challenging or contextualising their opponent’s claim – 53.8% – without any reporter mediation (a proportion broadly similar across broadcasters). Approximately half of these exchanges were not centred on a specific figure or data set, but on distinct or generalised statistical claims, as this example on Channel 4 illustrates,
Without additional analysis or journalistic interpretation, it would be hard to see how this kind of statistical tit-for-tat would enhance viewers’ understanding of EU membership or the consequences of leaving. The dominant message conveyed by this particular form of impartiality was that the weight of evidence on both sides was broadly comparable – something that many independent observers (notably economists) would dispute. The decision by broadcasters to push most independent voices onto one side of this simple binary may have amplified viewers’ confusion about the relative merits of EU membership.
Statistical tit-for-tats, constructing balance and (re)interpreting impartiality
In recent years, there has been greater recognition among broadcasters that impartiality should not be translated into simply balancing the competing sides of a debate or issue. It is now widely acknowledged that a binary notion of balance can distort coverage when the weight of evidence clearly falls on one side (Lewis and Speers, 2003) – most famously in the coverage of climate change or the reporting of the Measles, Mumps and Rubella (MMR) vaccine.
And yet, our study of television coverage of the EU referendum campaign found that while broadcasters understandably balanced the airtime granted to official Leave and Remain actors, they did not consider or scrutinise the veracity or weight of opposing arguments. In applying this editorial judgement to ‘due impartiality’ and ‘broad balance’, we also found broadcasters did not reflect the full range of party political opinions and left little space for analysis beyond politicians and campaigners. In particular, the UK Conservative Party – who were split on leaving or remaining in the EU – made up by far the largest share of party political sources. Labour – or, indeed, any left of centre party voices, who represented a more left-wing perspective on EU membership - were marginalised as the electoral contest was often interpreted through the prism of Conservative party infighting or a future leadership content. In short, while the Conservative case for Remain was featured prominently, a more left-wing case for EU membership – despite intense campaigning by many senior Labour, Liberal Democrat, SNP, Plaid Cymru and trade union figures – was side-lined. The striking failure of the Remain campaign in many traditional Labour heartlands raises the plausible possibility that the lack of political balance in broadcast coverage – notably the absence of a left-wing case for Remain – may have been significant. While broadcasters might claim Labour’s marginal coverage was due to its leader being a reluctant participant in the referendum campaign, there were nonetheless many other senior figures in the party (and other parties) that could have been used to counter-balance Conservative perspectives.
Our analysis of statistical claims made during the coverage also shows how the narrow balancing act between the (generally Conservative) advocates for Leave and Remain
While the preponderance of expert testimony and statistical evidence favoured the Remain campaign, the proportions of statistical claims made on television news were finely balanced between Leave and Remain campaign actors. In practice, this meant that evidence from a number of highly respected economists, for example, were given equal rate with a simple rebuttal from a campaigner on the Leave side.
Similarly, most statistical claims were not subject to challenge or put in context. Instead, we found coverage was dominated by
In the aftermath of the referendum, several journalists questioned the way impartiality had been interpreted. Justin Webb, presenter of the BBC’s Radio 4 Today show, argued that ‘One of the clearest messages during the referendum campaign was that audiences were hungry for real knowledge. People wanted to go beyond claim and counter-claim so that they could work out what was true’ (cited in Plunkett, 2016b). Yet, the operational definition of impartiality the broadcasters chose to adopt ruled this kind of journalistic scrutiny out of bounds. If, for example, reporters had repeatedly made clear that the Leave campaign’s central statistical claim was widely discredited, they would have fallen outside the narrow confines of their own, self-imposed rubric. Getting to the truth, in this sense, was trumped by a narrow version of impartiality.
As a consequence, our study indicated that journalists often appeared reluctant to make judgements about the veracity of some of the campaigns’ claims or turn to alternative sources of knowledge for analysis and verification. Despite the fact that, as Radio 4 presenter Eddie Mair pointed out, voters were asking ‘basic questions’ about the EU membership that politicians
Although the United Kingdom’s referendum campaign lasted 10 weeks, days before the vote, one survey showed that less than one-third – 31% – of people ‘felt well or very well informed about their EU vote’. 3 This was revealed by an Ipsos MORI (2016) poll that found close to half of respondents believed that the UK government spent £350m per week on EU membership, despite independent experts repeatedly challenging this statistic. Adopting a more evidence-driven approach to impartiality, where journalists independently seek the most truthful version of events – as well as taking their mission to inform more seriously – may have gone some way in better informing people before they cast their vote.