Abstract
Muslim public opinion polls are mostly taken at face value as the direct and unbiased voice of British Muslims, but, as this article argues, most of the public opinion polls are commissioned by the media and suffer from similar framing effects to those seen in the general media coverage of Muslims. At a time of national crisis, following the London terrorist attacks in 2005, it has become especially clear that the media have been following their pre-existing narrative on Muslims rather than responding to public interest. We analyse all public opinion polls conducted in the 18 months following the 7 July attacks and all their broadsheet newspaper coverage to show that the media-framing effects influence both the creation of Muslim opinion polls and their reporting.
Introduction
All articles, questions, mentions, and polls commissioned per newspaper by its ideological tendency.
Source: Compiled by authors.
Muslim opinion polls are generally a major source of knowledge of British Muslims’ attitudes for the media, commentariat and even academics (for just a few examples see Field, 2007 and Saggar, 2006, 2009). However, these polls offer far from a straightforward picture of what Muslims think. With most of the questions asked in these polls determined by the media (who commissioned most of them and then reported the results), one has to ask to what extent did the British public see a picture of Muslim public opinion, and to what extent was it the media reflecting their own preferred narrative to create headlines and sell their newspapers?
The research question posed in this article is different to the usual investigation of media-framing effects. Most studies of media coverage of Muslims and Islam focus on the analysis of the value judgement aspect of the coverage, which forms a lesser part of this study, and which is heavily influenced by the media using a range of elite sources to inform their narratives (Chomsky and Herman, 1988; van Dijk, 2000). The media rarely use any statistical evidence of Muslims’ opinions and attitudes. Lewis, Mason and Moore (2011) show that in their sample of articles concerning British Muslims from 2000 to 2008, only 2 per cent used any survey/poll evidence at all. When used, however, quoting a public opinion poll is a powerful tool in the hands of the media outlets, as the media do not fall back on someone else’s authority and instead claim that they are directly reporting the views of the population in question. The existing studies do not take into account this difference between the picture of Muslims offered as a commentary and opinion from an elite source used by the media outlet, and the reporting of the results of the opinion polls. These are depicted as pure ‘news’, i.e. uncontaminated, ‘objective facts’ about Muslims and their opinions. This differentiation is crucial, as the public may look upon them as unrelated to the media outlets’ own preferred narrative tendencies, which in reality is not so.
The literature demonstrating an influence of media and other stakeholders in the production of public opinion artefacts is vast, but it mostly centres on media’s representations of the opinion polls (Bishop, 2004: 143–167) and their possible influence on these opinions in the first place (Entman, 2004: 127). However, in most cases, the main responsibility for framing public opinion at the stage of collecting it is deemed to lie with the pollsters. In the case of the opinion polls after 7/7, the engagement of the media in polling Muslims was more deep-reaching as the thematic selection of questions also fell on them as the commissioners of the polls: out of the 14 polls of Muslim public opinion conducted in Britain in the 18 months following 7/7, seven were commissioned by broadsheet newspapers and five by other media outlets (see Appendix). As a result our analysis concentrates on the role of the media in both the creation and reporting of such polls. We will call these two stages of media-framing effects commissioning effects (framing effect on what Muslims are asked in the public opinion polls), in contrast to the usually analysed reporting effects (which pieces of opinion get published and how they are presented).
To explore the commissioning effects of media involvement in production of the Muslim opinion polls we will tackle the selection and exclusion of themes not just from news stories, but at an earlier stage: the polls themselves. To test our hypothesis that this selection followed a pre-existing media narrative rather than news events, we will see if the selection of themes and topics responded to the news cycle. Finally, we will ask two questions: were certain themes presented as more negative than others, and did different newspapers present Muslims’ opinion poll evidence as more or less negative? We will then establish how the media’s ideological tendency influenced the commissioning and reporting effects.
To answer these questions we analysed all the polls conducted during the 18-month period between July 2005 and December 2006 and all the articles in the British broadsheet press reporting their results (in a period for up to a year after each poll was first released). First we analyse what was asked, and then how it was reported. The inclusion of a stage of media control over Muslim public opinion prior to reporting is the main contribution of this paper to the existing body of knowledge. We find that media commissioning and reporting effects were visible in the questions asked and the way in which they were reported, with the narrative of problematic integration of Muslims, with special attention to cultural difference, prominent beyond what could be expected considering the events that took place during that period.
Muslims and the media
Whereas the studies of what is asked of Muslims, by whom and why are virtually non-existent, there is abundant literature on media representations of Muslims. It is generally agreed upon that these are predominantly negative and very selective. In fact, any ethnic minority suffers from a more negative coverage than the white majority in most countries (van Dijk, 1991). This negative bias has been measured in many different ways from association of the minority group with predominantly negative topics such as crime or cultural deviance, placing such content in more visible places in the newspaper, to using different syntax when describing minority and majority groups (van Dijk, 2000).
Following the terrorist attacks in New York in 2001 (9/11) the media narrative on terrorist threat changed with significant consequences for the public perception of Muslims (Kundnani, 2004; Moore, Mason and Lewis 2008; Nacos and Torres-Reyna, 2003; Spigel, 2005; Zelizer and Allan, 2002). The media’s influence over the interpretation of events is done through what van Dijk (2000) calls ‘news structures’ or what Norris (2003) calls ‘news frames’: patterns of selection and exclusion (of topics or stories), emphasis and language used in creating news stories and reports of events that form a coherent picture with the attached value judgements that support the consistency of media narrative. The existence of these frames and structures is not, in its own right, a negative thing. To an extent the role of the media is to introduce such explanatory frames in order to give their readers not only news out of context, but also an understanding of the origins, importance and consequences of events. To perform this role, selection and exclusion are inevitable, but they also are crucial mechanisms of manipulation of stories.
In creating these news frames in relation to Muslims, the media predominantly focus on two topics: terrorism and cultural differences. Coverage of terrorism is an answer to topical events such as 9/11, the Bali bombings, 7/7 and many attacks in Iraq, India, Israel or Afghanistan. The proliferation of the coverage of cultural issues and differences, however, reflects a preoccupation about integration (Jaspal and Cinnirella, 2010). This is a result of culture being presented as a determining factor of Muslims’ relations with the West (Huntington, 1996; Said, 1978) and the main focus of this debate is on Islam as a religious identity rather than a more complex political phenomenon (Pargeter, 2008). Said associates this trend of focusing on cultural difference with ‘Orientalism’ in which Muslim culture is presented as backward and at odds with Western values (Said, 1978), which is often associated with the (as some argue) ‘new’ forms of racism, in which socially unacceptable notions of biological inferiority were replaced by socially tolerated concepts of cultural inferiority and other negative stereotypes (Barker, 1981, Pettigrew and Meertens, 1995). Some have argued that it differentiates Islamophobia from more general forms of racial prejudice (Allen, 2005; EUMC, 2005; Modood, 1992; Open Society Institute, 2010; Runnymede Trust, 1997).
It therefore comes as no surprise that the negative media picture of Muslims is fuelled by the notion of cultural difference (van Dijk, 2000), also found in the UK (Poole, 2002, 2006). Whereas sometimes the media focus on culture may be dictated by current events (such as instances of violent crime or forced marriage) or explicitly cultural debates (such as the role of women and headscarves), it is often just part of the media narrative. The media are generally painting a picture of culturally different, and therefore more threatening, Muslims who stand in opposition to Western (and British) values (Poole, 2002, 2006; Said, 1997). Poole also underlines the generally negative image of religion in the British context, which exacerbates the situation for Muslims, a group mostly defined by its religious affiliation (Poole, 2002). Most research concluding that culture is not a dominant theme for coverage of Muslims in the media does not include religion, thereby separating religion, cultural difference and failure to integrate (Lewis et al., 2011) and comparing them individually rather than together to the narrative of terror threat. As a result, terror threat has been deemed – until 2008 – the most dominant theme of media narrative (Moore et al., 2008), with cultural difference only in more distant second place. However, including other narratives linked to cultural difference: integration-related difficulties such as religious identity and national loyalties (Lewis et al., 2011), make the theme of integration the most dominant media frame already before the 2008 benchmark.
The focus on difficulties in integration and cultural difference is especially damaging for the public view of Muslims in Britain because it generalises from cultural differences to difficulties of integrating with the wider society. From there, a link is being drawn between lack of integration – alienation – and the threat of terrorism. It is usually hard to pick up in a single article because of its short form; however, many books published by journalists offer a more in-depth understanding of the logic behind it. This literature links three issues: alienation (cultural and social), religious extremism as an attractive alternative to Western identity and the resultant support for violence, including ‘potential’ or ‘tacit’ support for violence (Bawer, 2006; Phillips, 2007; Shore, 2006). The apparent link between exclusion, or self-imposed segregation from mainstream British society, and the threat of crossing to the ‘other’ side – religious fanaticism and then support for violence – has been to a large extent a built-in response to the stories of the 7/7 bombers, all of whom were British-born young Muslims who appeared to have chosen fanatic Islamism over Britishness. Even though the narrative of disillusioned Muslims seeking ‘revenge’ on Western society is widespread in the media (Ali, 2007), generally it does not find much evidence in the existing empirical research, as most of the academic evidence on Muslim public opinion in Britain, using academic and government surveys as opposed to commercially commissioned public opinion polls, finds that Muslims have high levels of belonging to Britain, high levels of political trust and trust in British institutions and are equally or more civic-minded than the majority (Maxwell, 2006, 2010; Sobolewska, 2009). Similarly, research definitively shows that both the successful and would-be terrorist are either average or above average in terms of class and education than the general public (for example see Krueger, 2008), rather than alienated as is often assumed.
Despite the evidence to the contrary, the image of ill-integrated and threatening Muslims is widespread in general British public opinion. The common interpretation of Muslim public opinion is that ‘Muslims in Britain are slow to integrate into mainstream society, feel only a qualified sense of patriotism and are prone to espouse anti-Western values that lead many to condone so-called Islamic terrorism’ (Field, 2007: 466), and Field argues it is a justified opinion of Muslims. In addition to the general issues of validity of the concept of public opinion and its nature as a social and political construct (Bourdieu, 1979, Lippmann, 1946) and largely an artefact of measurement (Bishop, 2004), we want to raise the possibility that the picture of Muslim public opinion is to a large extent a reflection of media’s overarching narrative. We believe this is especially evident from the polls conducted after 7/7, which were predominantly commissioned by the media. Commissioning of polls is, therefore, in addition to reporting, a tool with which media impose their own coherent narrative and shape the public’s understanding of Muslim integration, terrorist threat and other aspects of Muslim public opinion. There is a very old truism that ‘you can only get an answer to the question you ask’ and nowhere is this truer than in public opinion. The way in which we pose a question determines which answer we receive and therefore there is an obvious opportunity for creating bias in the answer by creating a bias in the question (Clark and Schober, 1992).
Although the right-wing conservative press is usually more at fault in misrepresenting minorities, mostly in terms of criminality, than the left-leaning media outlets (Said, 1978), in the case of Muslims, even the left-leaning ones engage in negative representations, mostly grounded in the left-wing outlets’ secular and liberal stance, which is contrasted with ‘irrational’ – i.e. religious – and culturally ‘backward’ behaviour of Muslims (Poole, 2002, 2006). 1
Data and research design
Because this project goes beyond observing the media framing effects in the usual way –based on publications of all material relating to Muslims – and focuses on representation of Muslim public opinion, we need to differentiate between commissioning and reporting effects, as previously defined. The media’s role as commissioners of Muslim public opinion polling is rarely addressed; in the case of Muslim polls following 7/7 they were particularly pronounced. Most of the commissioners were media actors: television channels, broadsheets and other newspapers. The only two non-media commissioners were think-tanks. We classified the bodies commissioning the polls into three categories: left-wing (The Guardian and The 1990 Trust), right-wing (The Times, The Sun and Sky News) and independent (BBC and Channel 4 – both recipients of public funds and therefore under an obligation to provide balanced coverage – and PEW Global, a non-partisan think-tank).
Looking at the reporting effects, using the best-case approach (Entman, 2004: 77) we chose to look at the reporting of the Muslim opinion polls by the broadsheet newspapers only. Following the logic of this approach, we look at the broadsheet newspapers’ fulfilment of their claims and reputation for the quality of their news coverage. The broadsheets were also uniquely suited to our study because of their active role in commissioning of the polls, whereas The Sun was the only tabloid to have commissioned a poll in this period. 2 We looked at eight broadsheets and have coded them with respect to their ideological tendency as right-wing (The Times, The Daily Telegraph, Sunday Times and Sunday Telegraph) and left-wing (The Guardian, The Independent, The Observer and The Independent on Sunday). Previous literature shows this ideological division has important consequences for these papers’ coverage of issues such as immigration and minorities (Poole, 2006).
Between the 14 polls conducted in the 18-month period following 7/7 (details in the Appendix) there were 408 questions asked (we were able to find the wording for 405 of these questions). 3 We have coded these using a two-tier coding frame: theme and topic. We developed a coding frame with four broad themes: integration, security, foreign policy, and political freedoms and liberties. Within these themes we further identified 70 more specific topics, the nine most popular of which will be discussed in analysis. 4 Within the broader theme of security the most popular topics were overt support for terrorism, ‘tacit’ support for terrorism, support for terror prevention and awareness of radicalism in the Muslim community. Within the theme of integration, the most populous topics were issues of culture and cultural difference, loyalty and belonging to Britain, discrimination, religion and alienation. The coding reliability test was performed by three coders on a sample of 94 questions and the rate of agreement was 90 per cent, the inter-coder reliability achieved, as measured by Cohen’s kappa, was 0.92, which is outstanding. We found that our broad themes and topics corresponded closely to the findings of other studies of media coverage.
In the second stage of our analysis we looked at the reporting of the Muslim public opinion polls in the broadsheets in the period of 12 months after each of the polls was published. Overall we found 344 mentions; out of the total 405 questions in 14 polls, 142 – from 11 polls – were mentioned at least once. The three polls ignored by the broadsheet press were the 1990 Trust poll and the two polls conducted by The Sun – a lower quality newspaper. Altogether, these mentions spread across only 83 articles (see Table 1). We coded each of these mentions into four evaluative categories: positive, negative, ambivalent and neutral using a set of criteria including language used and the context in which the mention was provided. The coding reliability test was performed by three coders on a sample of 73 mentions, the percentage agreement was 94 and the coder reliability (Cohen’s kappa) achieved was 0.89, which is very good.
As the media do not directly shape the questions themselves, thus excluding from this study possible sources of bias such as question wording and other sources of unreliability of the polls (Bishop, 2004), we will focus on the aspect of poll production that media do control when commissioning the polls: theme and topic selection.
Analysis
Theme selection and exclusion
The theme selection and exclusion is one of the main aspects of media framing of issues surrounding Muslims and Islam. As we argued earlier, this can be done not only through selection of stories that the media prints or broadcasts, but also in how they commission Muslim opinion polls. First, therefore, we looked at the media as commissioners of the polls: what questions were asked and how the ideological tendency of the outlet influenced what was asked.
Commissioning bias in theme selection
Figure 1 shows what the main four themes asked about in Muslim public opinion polls were in the 18 months following 7/7, and whether their distribution varied by ideological tendency of the commissioner. The two dominant themes were integration (52 per cent of all questions) and security (30 per cent), with themes of foreign policy (9 per cent) and political freedoms (2 per cent) far behind, and very few questions outside of any of these categories and thus grouped in the ‘Other’ category (7 per cent, not shown). Given the nature of this period, we expected a lot of the questions were determined by the public concern with security threat related to terrorism; however, we were surprised by the number of questions on issues of integration. For the independent and the right-wing commissioners, the questions relating to integration were most popular, forming a clear majority of all questions asked by independent commissioners (66 and 49 per cent of all questions commissioned, respectively). The left-wing commissioners were the only ones more interested in security (security questions outnumbered integration ones by 8 per cent).
Popularity of themes by ideological slant of poll commissioner.
These patterns are hard to disentangle, and, pursuing them further, we intended to present a comparison between the media and non-media commissioners, but with a couple of exceptions the differences were not visible. The first observed difference was related to the huge interest of the television channels – that are classed as independent – in asking about integration. We believe this phenomenon drives the overwhelming pattern seen for independent commissioners asking predominantly integration-related questions. Another theme that differed in popularity between media and non-media commissioners was foreign policy. Despite it being a plausible explanation for Muslims’ disgruntlement with the British government, it was largely ignored by most polls. The foreign policy questions were mostly asked by think-tanks – 1990 Trust on the left and an independent Pew Global. It shows that these non-media commissioners were the only ones to go beyond the narrative of integration to seek explanations for Muslims’ disgruntlement.
Figure 2 presents a more detailed picture of what topics were present within the two dominant themes: integration and security. Within the theme of integration the dominance of the issue of culture (29 per cent of all questions on integration) was consistent with other literature on Muslim press coverage. Other well-known topics such as discrimination and Islamophobia (19 per cent), loyalty to Britain (21 per cent), religion (15 per cent) and alienation (13 per cent) were significantly less popular. The impact on the distribution of these topics by ideological tendency of commissioner is again interesting. The left-wing-commissioned polls were dominated by two polls from The Guardian that were conducted shortly after 7/7 and on its first anniversary – and did not ask any culture-related questions (which brings us to the issue of newsworthiness and following a news cycle, which will follow in the next section). Left-wing commissioners were most likely to ask questions about Muslims’ loyalty to Britain and Islamophobia. The independent commissioners, in contrast, were mostly preoccupied by culture (41 per cent), with religion second (18 per cent), loyalty and belonging (16 per cent) and discrimination and alienation adding up to 15 per cent between them. The right-wing commissioner’s interests were almost the exact reverse, with the topic of discrimination being the most popular (25 per cent) alienation and culture in tied second (22 per cent each), and loyalty and belonging and religion amounting to 10 and 12 per cent respectively.
Popularity of integrationist and security topics by ideological slant of poll commissioner: proportion of all integration and all security: (a) Integration and (b) Security.
Although we saw that commissioners with different ideologies preferred certain question themes and topics, it is hard to see any definitive pattern. The preference of right-wing commissioners for the topic of discrimination may be of interest, however, as it confirms the earlier-discussed more sympathetic approach of the right-wing media. We will discuss the evaluative interpretations at a later stage when we look at the reporting of polls, but at this stage it is worth noting that – as existing literature shows – different themes and topics have potentially different interpretations. At this stage, therefore, we can look at only their evaluative potential. The issue of Muslim loyalty is generally open to becoming part of a negative narrative, as is the issue of culture and religion. In all these cases, poll questions have the potential to reassert the differences in culture and importance of religion in British Muslims’ lives, with the implied possibility that, in a clash between Britishness and religion, it will be the latter which will win the loyalty of Muslims. Although potentially some of these questions may result in drawing similarities between Muslims and other Britons, but in fact the limited comparison with white public opinion polls made it unlikely (with only 2 out of 14 polls offering such comparisons, and not on all questions). Topics of alienation and discrimination are more than potentially ambivalent. On the one hand, these questions may seek to defend Muslims from negative consequences of increased suspicion and Islamophobia (by making the public aware of the high costs of the extra vigilance), but the focus on discrimination also can be perceived as negative as the Muslim communities are often attacked in the press for their airs of ‘victimhood’ and the demands they place on British society to placate them (Lewis et al., 2011). We will revisit this issue in our last section, in which we explore the evaluative reporting of polls by the broadsheets.
Within the theme of security, the ideological tendency had a clearer impact on what was asked (Figure 2). On the left the narrative of ‘tacit’ support for terrorist and the issue of awareness of terrorism and extremism were dominant (31 per cent each). These were questions yielding potentially the most negative results as the ‘soft’ support for terrorism questions contained vague questions on ‘understanding’ the motifs of terrorists and ‘sympathising’ with their cause, thus increasing proportions agreeing to these fuzzy formulations. Similarly, questions about awareness of terrorist activity yielded higher proportions of Muslims agreeing than direct questions on support for terrorism, thus creating an atmosphere of fear. Conversely, questions on direct approval of terrorists yielded very small – often within polling error – proportions saying they support terror; conversely, questions on support for terror prevention resulted in many Muslims offering their support in the fight against terrorism. These two types of questions were less popular with the left (15 and 22 per cent, respectively), but more popular with the right wing (26 and 35 per cent) and independent commissioners (20 and 41 per cent), once again underscoring the left’s unsympathetic tendency towards Muslims.
Reporting bias in theme selection
We now turn to the selection of themes in the reporting of these poll questions in the broadsheet press to see if ideological patterns shown by poll commissioners persist. As half of these polls were commissioned by broadsheets, we may expect that the security and integration themes would also dominate newspaper reporting of the polls. Figure 3 shows that this was the case, with integration being a dominant theme for most broadsheets regardless of ideology (over 50 per cent overall). The available questions on foreign policy and political freedoms and liberties were almost completely ignored in almost all cases. The left-wing broadsheets covered the issues of foreign policy relatively more extensively (between 7 and 9 per cent) than the right-wing ones (between 1 and 2 per cent).
Integration and security questions mentioned by newspaper: proportion of all poll questions mentioned.
The coverage was dominated by integration items, with The Guardian and Daily Telegraph most heavily preferring the theme of integration (56 and 60 per cent, respectively). Figure 4 shows a more detailed breakdown of this theme by topic. The questions on culture and cultural difference amounted to more than 50 per cent of integration-related coverage for The Times, The Guardian and Daily Telegraph, with the general preference for this issue on the left side of the ideological spectrum. Again, however, differences between newspapers trump ideological differences in the selection of themes and topics reported.
Mentions of Muslim public opinion polls by broadsheets by topics: (a) Integration and (b) Security.
Looking at the topics within the theme of security, the questions on ‘tacit’ support for terrorism were least popular in The Times (19 per cent) and most popular in The Guardian (57 per cent). The reporting of questions on awareness of extremism, however, was entirely inverted, with The Guardian barely mentioning them (4 per cent) and The Times mentioning them a lot (52 per cent). Despite the right-wing newspapers reporting more items on support for terrorism, the main differences between left- and right-leaning newspapers are mostly led by the difference between The Times on the right and The Guardian on the left.
Responsiveness of theme selection to the news-cycle
The fundamental question in assessing framing is to first agree what should have been asked and why what was being asked was a result of media framing. Even though most of the existing research makes a justified claim that presenting Muslims predominantly in the light of terrorism constitutes bias on the part of the media, in the immediate aftermath of 7/7 we can hardly argue that focus on a timely and newsworthy issue at this time would constitute bias. The events on the scale of 7/7 are reflected by the press and the interest in Muslims and Islam, not just the terrorists themselves, peaks. However, presence of less-relevant topics or absence of other timely and newsworthy themes can be construed as media framing. Hence we will look at whether or not questions asked in Muslim polls, and their reporting in the press, followed the news cycle. More specifically, we will see if the media fell into their preferred narrative by explaining Muslim terrorism with an integration-related narrative.
In the 18 months under analysis, there were a few salient events concerning British Muslims and Muslims in general. The relevant event in the news cycle that we expected to contribute towards the increase in numbers of security questions asked was obviously the July 2005 bombing, followed by the fatal police mistake during an anti-terrorist operation the same month; crackdown on extremist Muslim clerics in August 2005; violent protests in London against Danish cartoons thought to be offensive to Muslims in February 2006; passing of terror prevention legislation in March 2006; another fatal police mistake in an anti-terror operation and a foiled bomb plot in June 2006; and finally the first anniversary of the 7/7 bombing and violence abroad relating to the offensive cartoons in July 2006.
The events we expected to cause the increased interest in integration-themed questions and their mentions were the cartoon controversy in February and September 2006 and the debate over Islamic dress in March 2006 (Shabina Begum’s case) and October 2006 (Jack Straw’s remarks) and finally after Tony Blair’s speech about integration in December 2006 and controversial Channel 4 documentary about mosques in January 2007.
The results in Figure 5 show that whilst there has been some following of the news cycle in the frequency of asking and reporting security-related questions, little evidence of this can be seen for integration-related items. The proportion of security questions asked in Muslim opinion polls was steadily falling after the 7/7 bombings in order to rise again in response to the failed bomb plot in June 2006 and culminate on the first anniversary of 7/7 bombings. The patterns of reporting security questions followed closely, until a small peak of asking security questions was recorded to coincide with the cartoon protests in October 2006 but not matched by reporting more security questions in broadsheet press.
News cycle and the polls: security and integration.
In contrast, the proportion of questions asked – and reported – on the theme of integration does not follow the expected cycle. Instead, more integration questions appeared in polls and in press in August 2005 – seemingly to coincide with Muslim cleric crackdown – and then when the terrorist plot was foiled in June 2006 and arrests were made in August 2006. The last peak in asking and reporting integration questions coincided with Blair’s speech on integration. Only two of these events, the crackdown on extremist clerics and the prime minister’s speech, were directly related to integration. However, the other two were security-related events and it is highly indicative that the foiled bomb plot originated with British Muslim bombers threw into question the integration of Muslim community in the eyes of British media. As a result it is clear that the poll commissioners, and then the reporting newspapers, conflated issues of integration and terrorism. This strongly indicates that the narrative of integration has been a preferred explanation for the Muslim terrorist threat, despite its lack of support in empirical research (discussed earlier).
The good and the bad: Evaluative interpretations
In a final test of media framing, we now look at how these polls were interpreted and framed by the broadsheet that reported them. We are interested in the overall positive or negative presentation of results, whether they varied by theme (with the contentious issue of integration getting more negative coverage) and whether right-wing papers were more sympathetic than left-wing ones. Here therefore, we look at only the reporting of Muslim public opinion polls.
There were multiple mentions per question, with significant inconsistencies in evaluative framing between the multiple mentions of any single item. We recorded 219 mentions as having negative evaluative interpretation, 73 as having positive, 35 were reported as neutral (without any interpretation) and 11 were coded as having ambivalent interpretation. Out of the 142 mentioned questions, 38 (more than a fifth) were reported in an inconsistent way. Figure 6 shows how the interpretations varied by themes and, as we expected, we can clearly see that the integration questions were presented in a more negative light than security questions, which were more evenly reported.
Positive and negative mentions of poll questions by theme and newspaper.
Figure 6 also shows how ideological tendency of the broadsheets influenced their overall evaluative interpretation of the polls. On the whole, the right-wing broadsheets were more balanced, giving more positive interpretations to issues of security and integration and less negative ones. The Times gave more positive coverage than any other newspaper to the findings of the polls (25 per cent of their coverage of polls was positive), including the crucial issue of integration, which The Times was by far most positive about (positive coverage of integration was 16 per cent of their coverage). The other right-wing broadsheet, The Telegraph, which carried a lot of mentions of Muslim polls, also gave 26 per cent to positive mentions, but was more critical of integration – 40 per cent of its coverage contained negative mentions of it. The left-wing press was significantly more critical. The left-wing paper that has given most attention to the polls, The Guardian, referred to these polls as positive only 17 per cent of the time. More than half (54 per cent) of its coverage was occupied by negative mentions of integration-related questions; a figure exceeding any other newspaper’s negative coverage of this theme. It also mentioned security questions negatively 25 per cent of the time, in comparison with 17 per cent by The Times. The other left-wing broadsheets mentioned polls in a positive light only 14 per cent of the time, making the average ideological difference between papers very pronounced.
Conclusions
The main aim of this analysis was to identify whether media framing was present at the stage of both creating the Muslim public opinion polls and then reporting them. First, looking at the selection and exclusion of themes and topics that Muslims were asked about and those which were later reported in print, we saw most of the Muslim public opinion as heavily influenced by the media narrative both at the stage of data collection (commissioning of polls) and at the stage of reporting of the results. We did not find much evidence of systematic and coherent differences between the ideological tendencies of the commissioners nor of print newspapers’ reporting, suggesting that the selection and exclusion of themes and topics is part of a universal narrative on British Muslims. Integration (including culture) and security constituted an overwhelming majority of issues asked of Muslims, whereas other relevant topics such as foreign policy and political freedoms were largely ignored. We found, in agreement with previous research, that the media narrative rests heavily on the inter-connection between terrorism and cultural difference and integration. This was supported by the fact that, as we found, the asking and reporting of integration-related questions did not follow the news cycle of integration-relevant events, but instead also increased during periods of security threat. We argue that this indicates that the kinds of questions Muslim opinion polls have asked (and newspapers reported) followed the pre-formed media narrative, in which terrorist threat is caused by the difficulties to integrate and other cultural difference, a link which we showed in the introduction most research generally dismisses as spurious. This has serious consequences for social cohesion in Britain, as it is safe to conclude that the British are given a very selective insight into what their Muslim counterparts think about Britishness and how what they think influences the risks of terrorism.
The dominance of integration of Muslims as the main media frame is crucially important also because, as we saw, integration is a more negatively interpreted and reported theme than security. Negative integration-related mentions of Muslim public opinion were a majority of all negative mentions published by the broadsheets during the period under analysis. Overall, all themes in all broadsheets were given more negative than positive coverage, but especially mentions of questions regarding support for terror prevention and the rejection of the 7/7 as an unjustifiable act of violence made the theme of security relatively more positive. This once more underlines the risks of conflating terrorism with issues of integration as it creates a generally more negative picture of Muslims in the British press.
Our results also show that the right-wing press was more sympathetic towards British Muslims at this time of crisis following the 7/7 attacks than the left-wing press. Existing research does point to the fact that left-wing press may be more critical of Muslims than they are of other ethnic minorities and immigrants in general, because of Muslims’ perceived cultural traditionalism (Poole, 2002). Conscious of the fact that the coding of mentions of Muslim public opinion into negative or positive may be open to human error and an undue influence of personal prejudices and judgements, we are keen to underline that, even without the coding of stories into negative or positive, the left-wing press published more mentions of the questions pertaining to integration and cultural differences between Muslims and majority population. This alone points to the fact that the framing of Muslims in terms of a problematic and culturally alien minority has been more prevalent on the left. The contrasting good news on the right is that The Times newspaper has been found to be rather more balanced, not only publishing close to an equal number of positive mentions of Muslim public opinion, but also within the contentious theme of integration mentioning more stories on discrimination against Muslims than other outlets.
Footnotes
Acknowledgement
The authors want to thank Dr Laurence Lessard-Phillips for her time and effort as the third coder, as well as her invaluable comments on earlier drafts of this paper.