Abstract
This article examines the roles of the media in the process of political agenda setting. There is a long tradition of studies on this topic, but they have mostly focused on legacy news media, thus overlooking the role of other actors and the complex hybrid dynamics that characterize contemporary political communication. In contrast, through an in-depth case study using mixed-methods and multiplatform data, this article provides a detailed analysis of the roles and interactions between different types of media and how they were used by political and advocacy elites. It explores what happened in the different parts of the system, and thus the paths to attention that led to setting this issue in the political and media agendas. The analysis of the case, a partial policy reversal in the United Kingdom provoked by an immigration scandal known as the “Windrush scandal” reveals that the issue was pushed into the agenda by a campaign assemblage of investigative journalism, political and advocacy elites, and digitally enabled leaders. The legacy news media came late but were crucial. They greatly amplified the salience of the issue and, once in “storm mode,” they were key for sustaining attention and pressure, eventually compelling the government to respond. It shows that they often remain at the core of the “national conversation” and certainly in the eye of a media storm. In the contemporary context, characterized by fierce battles for attention, shortening attention spans and fractured audiences, this is key and has important implications for agenda setting and beyond.
“Why did it take an initial refusal by Downing Street to meet the Commonwealth Heads of State, the action of my right hon. Friend the Member for Tottenham and the many articles written by Amelia Gentleman for this national scandal to come to light? I have known about this matter for months; other Members have known about it for years. I do not understand why it took so long for us to realise that there was a severe and cruel injustice being meted out by the Government”
—(Emma Reynolds, Labour MP, in Parliamentary debate about Windrush triggered by e-petition).
“The key question is how policymakers prioritize issues for action given the flow of information into the system”
Why does an issue that is not new suddenly erupt on the media and political agendas? How does it burgeon into an important matter, when previously few people seemed to care about it? Are these eruptions significant influences on policymakers’ priorities, given the ever-growing tides of information constantly demanding their attention? Agenda-setting questions have fascinated political scientists and political communication scholars for a long time and for good reasons. Unless an issue gets into the political agenda, it will not be discussed, debated in the legislature or acted upon by the government. High salience in the media is not absolutely necessary, but research consistently shows that news coverage is an important factor in making policy change more likely (Walgrave and Van Aelst 2006), and this is especially marked during “media storms,” characterized by explosive increases in media attention (Walgrave et al. 2017: 550).
This article explores how agenda setting is affected by the contemporary political communication environment, where the flows of information are vast and fast, attention battles fiercer than ever, and there are a number of novel types of media and advocacy actors. We ask, “To what extent does our current understanding of political agenda setting still apply and what needs to be re-thought?” We approach these underresearched question through an in-depth case study design, focusing on a single policy event. We use a multiplatform mixed-methods approach, combining quantitative content analysis to identify the frequency, and hence salience, of the issue across different outlets and platforms; digital trace data of electronic petitions and Google trends; in-depth interviews with key actors; and document analysis about the case. This combination enables us to develop a detailed understanding of the complex dynamics between the different actors and platforms, and the conditions that enabled these practices to occur.
The selected case took place in the United Kingdom and became known as the “Windrush scandal”: the mistreatment of lawful citizens as a result of hostile immigration legislation. The case received much media and political attention during April to June 2018 although, as we shall see, the underlying issue arose far earlier. The attention resulted in the resignation of a senior minister and a partial policy change. For our purposes, it is key that the issue received explosive media attention, became highly salient in the political agenda, and that there were a number of different actors involved. All this makes it a rich case study to explore the complex dynamics we are interested in.
This article makes two key arguments. First, that a thorough understanding of agenda setting necessitates a broadening of focus. In the Windrush case, there were a number of actors and platforms—old, new and hybrid—that were important in the process. They offered a range of paths of attention to those advocating the case and contributed, to varying degrees, to putting the issue on the agenda. They were part of a campaign “assemblage” which comprises “multiple, loosely coupled individuals, groups, sites, and media technologies” (Chadwick et al. 2015). In this case, the assemblage combined a journalist from The Guardian, a legacy news organization that has been at the forefront of digital renewal; a small number of political and advocacy elites, including a handful of MPs (led by Labor’s David Lammy, born to parents from the Windrush generation) and equality, immigration and legal aid charities, digital advocacy organizations (change.org and 38 Degrees), crowdsourced opinion leaders, and a variety of digital platforms, including social media and electronic petitions. These dynamics would have been overlooked if, as is common, the research design had included only one type of media; alternatively, if focused on digital media, it would have overestimated their role.
This takes us to our second key finding. Legacy news media organizations were crucial in this case, showing that they still often are at the core of the “national conversation” and certainly at the eye of a media storm. To be clear: the distinctive role of legacy news media organizations is not defined by technology. Almost all outlets these days have digital dimensions in the production, distribution and promotion of news, and for some (including The Guardian), a combination of old and new media logics have become central to their business and journalistic models (see Chadwick and Collister 2014). But nonetheless, it matters that these organizations existed before the internet radically transformed the media environment, and are hence still strongly shaped by the logic, norms, and organizational forms of the previous era. This means that those at the top of this group still have the symbolic capital, user reach and loyalty, and the necessary resources (financial, expertise, and access) to produce quality content, and hence can command a central place in the media system (Nielsen 2016).
In the case of Windrush, although most of the legacy news media came relatively late to the case, they played two key functions: amplifying attention and, crucially, sustaining it. Furthermore, it was the storm mode, that is, explosive and sustained attention (Boydstun et al. 2014) that, we will argue, was crucial in helping to maintain the issue on the agenda and pushing the government to act. In the contemporary context, characterized by fierce battles for attention, shortening attention spans, and fractured audiences, this is key and has important implications for agenda setting and beyond. It also provides an important reminder that we must not underestimate the continuing importance of legacy media organizations, but also that we must understand them and their implications in new contexts.
Dynamics of Political Agenda Setting in the Hybrid Media System
The roles of the news media in the policy process have been thoroughly studied. A key dimension, and our focus here, is their roles in political agenda setting. Whereas in communication studies, agenda setting is concerned with media influence on the issue priorities of the public, the main question for political agenda setting is “how policymakers prioritize issues for action given the flow of information into the system” (Jones and Baumgartner 2012: 7). Media political agenda setting refers specifically to the impact of the news media coverage on political priorities: to what extent and how do the media contribute to establishing the political agenda and, potentially, force a policy response (Walgrave and Van Aelst 2016, 2)? This question is often studied along with the counterhypothesis, that is, that the media largely follow political elites and hence have a weak independent influence. This literature shows, first, that the influence between media and politicians can apply in both directions and is often reciprocal, although recent studies tend to show stronger media than political effects (Walgrave et al. 2017). Second, the degree and kind of media influence is contingent on a number of factors, including type of issue (Soroka 2002; Vliegenthart and Walgrave 2008), who promotes it and “owns” it (Green-Pedersen and Stubager 2010; Thesen 2013), and the characteristics of the coverage, including prominence, persistence, congruence, and framing (Walgrave and Van Aelst 2006). Media storms appear to be of particular importance in this regard because they boost the influence of the news in the political agenda “into a higher gear” (Walgrave et al. 2017: 550). Storms, according to Boydstun et al. (2014), are identified by three criteria: a high level of attention (prominence), after an explosive increase (distinctiveness), with a certain duration (frequency). Finally, it has become clear that rather than conceptualizing this process as a battle of politicians against the media, it is more productive to understand it as actor-centered, that is, a range of political actors trying to use the media strategically to advance their positions in intra-elite power battles (Sevenans 2018; Van Aelst and Walgrave 2016). This includes politicians but also issue advocates such as corporations, interest groups, and nongovernmental organizations (NGOs).
However, while this body of literature is large, and insightful, it is also the case, that existing agenda-setting theories are only partially successful at explaining the dynamics of the contemporary environment, and then only by combining different literatures and approaches which are not integrated often. The first problem is that most of the existing a number of studies have focused exclusively on news media narrowly defined, that is, news coverage in legacy media, especially newspapers. This has, of course, historical reasons and is not without merit given that a number of studies suggest that these media are still important for agenda building and intra-agenda setting (e.g. Guo and Vargo 2015; Pfetsch et al. 2015). Nonetheless, their prominence can be neither assumed nor, crucially, discounted. Moreover, we need to know how they interact with other types of media.
In this regard, the emerging literature—mostly from the United States—on intermedia agenda setting in the hybrid media system shows that the agenda of different types of media are often homogeneous, highly interdependent and self-referential (Vargo and Guo 2017; Wells et al. 2018). As a result, a common national agenda still appears to exist, especially for high salience issues (Atkinson et al. 2014; Hellsten and Vasileiadou 2015; Meraz 2011). However, it is not clear, especially outside the U.S. and electoral contexts, how this homogeneity arises. Nor, more importantly for this article, do we know much about how it affects political agenda setting. Therefore, it is important to look at a case of political agenda setting across the media system. Moreover, most agenda-setting studies have taken macro-approaches based on time-series data. While this is helpful for exploring causality, it does not address the complex interactions underpinning this process (Jones and Baumgartner 2012; Jungherr et al. 2019). This leads to our research question as well as shaping our case study design: What were the roles and interactions of different types of media and actors in the process of political agenda setting in this case?
Second, it is not only about actors and their interactions, but crucially also about the changes in the conditions in which they do so. Actors and sites beyond newsrooms now strongly contribute to news production, from NGOs who increasingly provide “boots on the ground” for international reporting (Powers 2016) to Twitter users who, intervening in real time, shape the journalistic narratives of media events (Chadwick 2011). Moreover, there are also plenty of other actors vying for attention beyond mass media and professional journalists, including new types of both media and advocacy actors, a distinction that is in itself increasingly blurred. In fact, contemporary societies are characterized by acute information over-abundance and fractured audiences. As a result, the battle for attention, of both the public and political elites, is now fiercer then ever (Webster 2014). There can no longer be assumed to be one centralized media flow, but multiple flows; rather than A framing cascade, there are several (Entman and Usher 2018), and it is uncertain where the next tsunami might come from, or which ones policymakers will feel compelled to pay attention to. This is compounded by the fact that the rhythm of the “information cycle” in the hybrid media system has sped up at a rapid pace (Chadwick 2011, 2017). This makes the mismatch between the slow bureaucratic policy process and media’s short attention span even greater. At the same time, it creates a potential divergence between the rhythms of legacy news media—which even when publishing online are still heavily shaped by genres, professional norms and organizational patterns of the previous era—and the accelerated rhythms and looser norms of new forms such as blogs, trending hashtags, and viral memes.
A small number of studies have taken a multiplatform and actor approach (e.g. Chadwick and Dennis 2017; Graeff et al. 2014; Wright 2015) precisely in the assumption that “in a hybridized system, no one single account tells the whole story” (Wright 2015: 429). However, despite the popularity of the concept of hybridity, this literature is surprisingly limited, and mostly focuses on contentious politics (e.g. Bennett et al. 2018; Tufekci 2013) or electoral politics (e.g. Faris et al. 2017; Wells et al. 2018), as well as being very U.S.-centric. We know remarkably little about how it applies to the political agenda-setting process.
Finally, it is important to consider that the same changing conditions have also affected the advocacy environment. They have expanded the paths to attention (Tufekci 2013) and facilitated the formations of sociotechnical assemblages. Drawing on the work of Deleuze and Guattari (1983) and Latour (2005), 1 assemblages can be defined as a group of heterogeneous units—including individuals and groups but also nonhuman elements such as technologies, hence the emphasis on the sociomaterial relations—with permeable boundaries and interactive and interdependent relationships, that work together for some time for a common project. The units are often only loosely related and the assemblage ephemeral, but it is greater than the sum of its parts because “action results from linking together initially disparate elements” (Müller and Schurr 2016: 30). This includes organizing without organizations and connective action (Bennett and Segerberg 2012) but also the work and infrastructures of hybrid (Chadwick 2017) or “third generation” (Karpf 2012) digital advocacy organizations, such as MoveOn, 38 Degrees, change.org, and GetUp. These are characterized by their multi-issue focus and rapid response, and they are hybrid in the sense that they rely heavily on digital technologies, combined with offline campaigning and a dependence on linkages with the news media agenda (Chadwick and Dennis 2017; Vromen 2016).
In short, given these changes in the political communication environment, existing findings about political agenda setting cannot be assumed to apply. We must look at different types of media in interaction with other actors, paying attention to the role of assemblages, including new actors and platforms, while avoiding overlooking legacy actors, which are often still at the core of the system.
The Case: The “Windrush Generation”
The case is about the impact of hostile immigration policies toward a group of people who became known as the “Windrush generation.” They arrived between 1948 and 1973, invited by the U.K. government to fill acute labor shortages. However, for various reasons, many of these people and their descendants did not have the documents currently requested to prove their right to citizenship and, as a result, some were later treated as illegal immigrants. This was a consequence of new immigration policies in place since 2012 and exacerbated with legislation from 2014 and 2016. These policies have aimed to create, in the words of the then Home Office Secretary and later Prime Minister, Theresa May, a “really hostile environment” for illegal immigrants. As a result, elderly Caribbean-born people who have been living and working in Britain for decades were deprived of their citizenship rights and, in some cases, were deported. The numbers are unclear, but some have estimated that there are over fifty thousand people who could be potentially affected (House of Commons UK 2018). Moreover, there are other larger groups in potentially similar situations, including those born in other Commonwealth countries, as well as, post-Brexit, millions of EU citizens settled in the United Kingdom.
This problematic situation had been going on since at least 2012 (see Gentleman 2019 for a detailed history of the case). A number of NGOs and the High Commissioners of some of the affected Caribbean countries made several public warnings over the years to the government. There was no response, nor was the issue covered by the news media aside from the occasional mention. This changed from late 2017, and especially in early 2018 (see the summary of key events in Table 1). But why, and what were the roles of different actors? From November 2017, Amelia Gentleman, from The Guardian, started regularly publishing articles about Windrush, after being alerted to it by an NGO in October 2017. For months, she focused on detailed individual cases based on interviews with over twenty Windrush victims and featuring a consistent combination of investigative journalism and a strong human-interest focus to capture the readers’ attention (personal interview). Channel 4 also broadcast two of these stories in January and February. However, as we shall see, there was little repercussion until the first feature about Albert Thompson in The Guardian on March 10, 2018. Thompson (a pseudonym) had been living legally in the United Kingdom for forty-four years but was denied access to free treatment for prostate cancer, which is a right for citizens. The fact that Thompson’s case was about health care is significant because the National Health Service (NHS) is object of much national pride and it is an issue that is perennially high in citizens’ agenda. On that day, the first petition related to Windrush on change.org was launched, reaching 100,000 signatures within two weeks, followed a few days later by crowdfunding for his treatment. In the next two weeks, other petitions about the same person were created, on change.org and on 38degrees.org. On the day of the first article about Thompson, there was also a question in Parliament about the case at Prime Minister Questions (PMQs), as an attack on the government’s health policy, an issue “owned” by the opposition; the Prime Minister seemed unaware of the case.
Summary of Key Dates and Events.
Note. NGO = nonprofit organization; NHS = National Health Service; PMQs = Prime Minister Questions.
On April 6, Patrick Vernon, an equality activist who has been campaigning for the celebration of the Windrush generation for years, launched a Parliamentary e-petition. This was key because it was only at this point that what had been treated as a disparate set of individual cases became “branded” as the “Windrush generation” (Gentleman and Vernon, personal interviews), with its well-known allusions to the struggles and contributions of Caribbean immigrants in the United Kingdom. This petition was supported by a strategic coalition of several equality, immigration and legal aid NGOs, and a small number of MPs led by David Lammy. Just ten days later, it had gathered over 150,000 signatures, hence comfortably passing the 100,000-threshold required to trigger a debate in Parliament. It was also at this point that the other petitions gained substantial support, including the change.org petition for Thompson, which reached over 400,000 signatures. Moreover, it was then that 38 Degrees decided to throw institutional support behind its citizen-initiated petition, which until then had been languishing below fifty signatures (personal interview with the petition’s creator). Given the number and pace of petition signing, this might appear at first as an example of “organizing without organizations.” However, it had elites and organizations at its core: a combination of political and advocacy actors taking the window of opportunity to push for an issue they have been campaigning on for a long time. In this campaign assemblage, however, new forms of digital participation were also important, for gathering support and, as we shall see, helping to amplify the overall visibility of the case.
As the petitions were fast attracting signatures, a different type of elite got involved: diplomats, specifically the High Commissioners of twelve Caribbean nations. It was not the first time that they had raised their concerns, but they took advantage of a new window of opportunity and joined the elite campaign alliance, and broader assemblage, that was forming. On April 12, they used the incoming Heads of Government Commonwealth summit to put pressure on the government, “demanding action from the Home Office” in what Barbados’ high commissioner called “guerrilla diplomacy” (Hewit 2018). The government confirmed to The Guardian, on the 15th, that they have rejected the meeting request. On April 16, the first day after Parliament’s Easter recess, Labor MP David Lammy delivered a letter to demand action from the PM signed by 140 MPs from all parties. It was the first but not the last MPs’ actions to draw attention to the case in Parliament and, by doing so, attract the media’s interest. Parliament then became a key site for the assemblage. Over the next few weeks, the opposition MPs strategically used a number of Parliamentary instruments, including PMQs, Urgent Questions and summoning the Home Office Secretary to the Select Committee, to keep pressure on the government and enhance the newsworthiness of the issue.
It is from this point in mid-April that the issue of Windrush gained momentum on the political agendas (both Parliament’s and government’s) and, ultimately, provoked a storm in the news media. Moreover, within a short period, the government’s response escalated from apologies, to the announcement of bureaucratic measures, the launch of an independent commission, to the resignation of the Home Office Secretary and, eventually, the announcement of a financial compensation scheme for those affected worth up to six hundred million pounds. 2
This combination makes Windrush a potent case for the aims of this article. In addition to high and explosive salience, it led to policy change even if, as the literature shows is most often the case (Walgrave and Van Aelst 2006), change was to an important extent symbolic. Moreover, there are two other characteristics that make it particularly interesting. First, a range of different types of actors were involved in a campaign assemblage (political elites, traditional and hybrid advocacy organizations, activists and digitally enabled crowds and opinion leaders), exploiting platform affordances and news media logics to support their aims. Second, it was to an important extent driven by an investigative journalist in a legacy news organization at the forefront of digital reinvention.
This demonstrates how the media are both an arena for other political actors, and actors themselves. Moreover, it reflects important features of the current media environment. Investigative journalism and its use to advocate a cause are of course not new. But there were some features of how it developed that are distinctive to the contemporary communication environment. First, there is the use of this type of investigation by media organizations not only for general brand-building, highly necessary in the current hyper-competitive environment, but also as a potential route to income generation, via regular donations and crowdfunding initiatives. Second, there is the influence of analytics in the coverage, which is in part why the issue was relentlessly covered by The Guardian: “the interest from our readers was very high” (Gentleman, personal interview) as shown in internal metrics, as well as emails and Twitter. Finally, typical of The Guardian’s hybrid approach, the stories were published in the print version but with additional features online, and were promoted by Gentleman’s and other Guardian accounts on Twitter. At the same time, it is not all about the new. The fact this was a legacy organization—with high prestige, expertise and resources—was also crucial. The Guardian not only had a senior journalist working relentlessly on this, but also assembled a team to investigate Amber Rudd’s involvement and received leaked information that was key to her resignation.
As any single case study, the case has clear limitations in terms of generalizability. But for the aims of this article, the richness of the case and the depth we can achieve are key. The next section will first discuss the issues related to research design, and then the methodology, explaining the approach and detailing the process of data collection.
Approach, Methods, and Data
To examine the roles and interactions between the various actors, a first crucial step was to uncover when the case received attention in different elements of the media system. To achieve this, we used a combination of quantitative content analysis to uncover the patterns of salience over time across different platforms and, to better understand the roles of different actors, we carried out a small number of elite interviews (see Supplementary Information) and complementary document research, including the analysis of parliamentary proceedings and historical background about the actors and issues that led to the case. In combination, this helped us to develop an in-depth and rich understanding of the case as a whole. For the content analysis, we used a combination of data sources, covering legacy and new types of media (digital native, alternative, and social). We first identified key legacy news outlets: the two main national TV news bulletins in each of the three main channels (BBC1, ITV, and Channel 4), 3 and the print versions of all top-selling national newspapers, including tabloids and broadsheets. Second, we added key digital native news sites (Buzzfeed and Huffington Post) and “alternative partisan brands” with the highest reach in the left and right (Media Reform Coalition 2019; see Supplementary Information). Third, we include social media, specifically Twitter. The rationale is twofold: first, plain data accessibility; second, Twitter’s reach. Their users are nothing like the population: it is disproportionally used by political and media elites, and those who are more politically engaged; in this case, however, this is an advantage as these groups are likely to be key in amplifying the salience of the case (Harder et al. 2017: 280).
The timeframe for all sources was the same, based on the chronology of the case: March 1 to May 30, 2018. 4 Newspaper articles and broadcasting transcripts were retrieved using LexisNexis and The Guardian’s API. 5 Relevant articles in alternative news sites were identified using a Google search and retrieved from the individual media sites. For all media types, we followed similar—but platform-tailored—keyword search strategies. For press, broadcasting, and alternative news sites, we searched for “Windrush” and “Albert Thompson” (details in Supplementary Information). After cleaning for false positives and duplicates, the final sample comprises 1,771 pieces: 1,263 newspaper articles; 245 TV segments; and 263 pieces from alternative digital outlets. We purchased a full data set from Twitter’s Historical PowerTrack, using a query developed from an earlier sample we had obtained from Twitter’s REST API as the scandal unfolded. After removing false positives from the data (see Supplementary Information), the final Twitter data set contains 1,769,940 tweets posted by 251,459 unique users.
The second main sources of data were electronic petitions. We focus on the e-petition to Parliament as the data provided by the U.K. Government Digital Service is highly detailed, but also use the interviews to reconstruct the evolution over time of the other petitions. Finally, we included the analysis of Google trends, as a proxy of public attention to the issue compared with other search terms. Although undoubtedly not a representative measure of public opinion (Trevisan et al. 2018), it offers a reliable “public driven gauge of salience” (Boydstun et al. 2014: 526) that has been used in a number of studies of agenda setting (Bennett et al. 2018; Boydstun et al. 2014; Graeff et al. 2014).
Analysis
The key question for political agenda setting is at what point do issues get on the political agenda, breaking the equilibrium that had meant that until then, although the issue had been already identified as problematic by some, nothing had been done about it. As explained in the case section, we know that the government’s response started, or certainly accelerated, from April 16. We also know that The Guardian published several articles over a few months before then. But what happened in between? What was the tipping point for the shift in attention-allocation and why?
First, we focus on the evolution of the issue for the legacy news media. As Figure 1 shows, Windrush had barely been covered until April 16, despite several earlier articles in The Guardian and some coverage by Channel 4. There is some coverage at the end of the preceding week when the high commissioners of the Caribbean countries complained, but it is still sparse.
Timeline of key events combined with daily number of newspaper articles, main TV news bulletins, alternative sites articles, and tweets referring to Windrush and/or “Albert Thompson.”
From April 16, there is a sudden surge; so much so that it can be categorized as a “media storm” following Boydstun et al.’s (2014: 518) definition that it should last at least seven days, have an explosive (at least a 150 percent) increase in attention to an issue from one week to the next and, over one week, 20 percent of front-page stories must be devoted to the issue. In this case, as shown in Table 2, there is indeed an explosive increase: the number of pieces referring to the issue grew, in that week, twenty-three times: from nineteen to 445 units. It is also enduring: the attention remains high for over three weeks. Only in the fourth week is there a substantial decrease—more slowly in newspapers, with its larger news-holes, than on TV. Moreover, a similar pattern of explosiveness is also present in the number of front-page stories/broadcasting opening headlines.
Number of Articles Per Week before and During Storm.
Note. This includes daily number of newspaper articles and main TV news bulletin items in selected outlets referring to Windrush and/or “Albert Thompson.”
The presence of the storm is crucial because in “storm mode, the media agenda matters more for the governmental agenda”: it leads to “a powerful political stampede … as political actors try to outrun each other” (Walgrave et al. 2017: 550). And indeed, when combined with the timeline of events and Figure 1, we can see that this is the start of the issue breaching the attention tipping point on the political agenda. More specifically, it starts when the then Home Office Secretary is summoned to attend Minister Questions in Parliament, apologizes and promises several light reparative measures, and the PM reverses her decision not to discuss the issue with Commonwealth leaders. Importantly, this was also the day after the delivery of the letter signed by 140 cross-party MPs and the day that the Commonwealth summit was opening.
The salience of the issue on the political agendas (both Parliament and government) remains, as the political fallout of the issue continues, pushed by the opposition using several parliamentary instruments. This, in turn, feeds the news media’s criteria of newsworthiness: not so much about the Windrush issue itself, but about the political ins and outs. In fact, the timeline clearly indicates that salience in legacy news media was associated with typical news values: elite involvement; the partisan battle for blame allocation in Parliament, party conferences and local elections; the fallout of a senior minister; and the political scandal dimensions of the issue. It continues, and in fact peaks, when Rudd resigns. Ultimately, the key is that there was a reciprocal influence between elites (advocacy and opposition) strategically feeding the newsworthiness of the issue, the governmental response, and news media attention in ratcheting effect and at higher pace because of the storm mode.
The second step in our analysis is to incorporate newer types of media, specifically a combination of digital native news sites (i.e. Huffington Post and Buzzfeed), the most popular online partisan news websites, and Twitter. The analysis reveals that, for this case, there was little “alternative” in the patterns of salience: new and legacy media’s are similar. Figure 1 (above) shows that, like most legacy news media, these new media sites only started focusing on the issue from April 16. Twitter is somewhat different: it shows higher activity earlier and has a clear peak on April 13, both related to the petitions. Despite these noteworthy differences, the overall pattern is similar to that of the other media types. Indeed, there is, as shown in Table 3, a high degree of correspondence in the evolution of the coverage, as shown by the high, significant correlations in daily frequencies (March 1–May 31) between all four media pairings (TV, newspapers, alternative, and Twitter). In this case at least, these newer types of media did not appear to have played a leading role, but they did contribute in volume and pervasiveness to the storm. Twitter, as further discussed below, was important at an earlier stage, but subsequently not that different.
Correlations between Daily Numbers of Alternative Site Articles, Newspaper Articles, and Main TV News Bulletin in Selected Outlets as well as Tweets Referring to Windrush and/or “Albert Thompson” (March 1–May 31, 2018).
***p < .01.
The importance of the storm is also clear from the analysis of the data from Google trends, which we use as a proxy for general public interest on the issue. To have comparative points of reference, in addition to Windrush, we also searched for immigration (the more general issue) and Rudd (the name of the former Home Office secretary). In Google trends, numbers express search interest in a topic (or query shares) relative to total queries for selected topics at a given time and place, in this case the United Kingdom. The 100 is given to the maximum day-peak for the term with the highest volume of searches in the chart; the other numbers are essentially percentages relative to this peak (Bennett et al. 2018: 665; Boydstun et al. 2014: 526; Graeff et al. 2014).
Figure 2 shows that, whereas the search patterns for immigration were spread across the timeframe, there is a sudden and substantial rise in interest in Windrush. The surge materialized at the exact time, from April 16, when the news media got into “storm mode” and the government started to respond. Although the resignation of Rudd led to the highest search volume, it is short-lived. In contrast, the query share for Windrush remained high for almost three weeks, closely matching the period of high salience in legacy media.
Google search volumes from Google trends data.
In short, the analysis of the case so far shows that legacy news media continue to be crucial in the process of political agenda setting, especially in storm mode. They are key in the way that they rise and feed on attention from political elites, as well as for creating substantial interest from the public at large. That said, there were also other dynamics at play. We will now turn to the role of petitions, hybrid digital organizations, and digitally enabled “crowds” and their leaders.
We focus on the Parliament petition as we have complete data for it. Unlike public interest at large (as indicated by Google trends above), Figure 3 shows that, for a small but significant number of people, the Windrush victims were on their agenda since at least early April, which is consistent with the earlier attention on Twitter (Figure 1, lower panel). By April 10, the Parliament petition had reached over twenty-five thousand signatures and by April 15, that is, before news coverage got into storm mode, it had over 120,000 signatures. This puts the petition among the top 1 percent in terms of signatures in the history of Parliament (UK Government and Parliament 2019). Equally, petitions on 38 Degrees and, especially, change.org—about Windrush in general and Albert Thompson specifically—had high levels earlier, well before the start of the storm.
Daily number of signatures for the Windrush Parliament petition.
So, on one hand, sustained legacy news media attention to the issue came later than that of a digitally enabled crowd of engaged citizens. The differences between Figures 1 and 3 are confirmed by the Pearson correlation, which shows that petitions have risen relatively independently from patterns of salience in news media coverage (r = –.156; p ≥ .10; see details in Supplementary Information). On the other hand, the highest peak in the number of signatures comes the day legacy news media started to pay substantial attention to the issue (16/4), showing the importance of news media in amplifying interest. It was also only then that the government’s response and the public interest, as measured by Google trends, materialized. Moreover, legacy news media (especially the press but not just The Guardian) continued to cover the issue longer, including in a number of front covers (Table 2), while interest from the public (Figure 2) and digitally enabled crowds in Twitter (Figure 1) started to dwindle faster, essentially flattening after May 4.
But we should not disregard the earlier popularity of the petitions, which started when the news media—other than The Guardian—were barely covering the issue. Digital petitions are rapid response mechanisms for activists and digital advocacy organizations (Karpf 2012; Vromen 2016). Like other forms of so called “click-activism,” they are “interventions in the attention ecology” (Tufekci 2013: 854) that provide novel and fast pathways for raising attention to advocacy causes. Enabled by social media platforms, they can create an integrative mechanism between a small committed group of advocacy actors and sections of the public at large, who sign and share it. Table 4, based on data about the sources of the referral to the Parliament’s e-petition page, provides clear evidence of this: most of the petition signers arrived from social media (first Twitter closely followed by Facebook). This is consistent with a previous study of two hundred petitions that shows that two thirds of all visitors to the petitions’ site arrived via these platforms (Hale et al. 2018). Access to the petition directly from links in legacy news media sites played a role, but small and comparable with those of search engines.
Source of Referrals to Parliament’s Windrush Petition.
Source. Authors’ analysis based on data from the U.K. Government Digital Service.
Note. Referrals from sources that could not be identified are excluded from the sample.
The petitions then went on to play an important role in interdependence with other actors. Even if they represent a very small proportion of the population, signatures, as other kinds of attention metrics, elevate perceptions of the worthiness of an issue, creating further attention from other actors such as politicians, journalists, and platforms (Zhang et al. 2018: 3163). And indeed, the first piece published on BBC online about the Windrush scandal referred to the petitions’ metrics as one way to explain the importance of the story, as did several politicians. Moreover, other outlets, including the best-selling Daily Mail, directly promoted the petition in their coverage. But it is not just about numbers: petitions are also an opportunity for activists to both get a sense of what might be popular and to re-frame the issue. In this case, the 38 Degrees and change.org petitions indicated the high resonance of Thompson’s case as it quickly became associated with the cherished NHS. Moreover, the Parliamentary petition—set up by Vernon—was key to re-branding the previously disparate individual cases as the “Windrush generation” and quickly gathered support through social media networks.
The final element of the analysis is based on Twitter data. We first identify the opinion leaders in the conversation (Papacharissi and de Fatima Oliveira 2012) measured by the number of retweets, and second, we analyze the hyperlinks included in any tweets about the issue. We coded the top 20 accounts in four broad categories based on their bios: politician, professional journalist or media organization, activist (which also includes campaigner and blogger), and celebrity. We chose to redact some nonorganizational twitter handles due to privacy concerns. The logic behind the redaction is that we did not want to expose Twitter users who are not otherwise public figures (e.g., who chose to appear on TV, radio or in newspaper articles with their real name) or have given us explicit consent to feature their names.
Opinion Leaders on Twitter in the Windrush Debate.
Note. Ordered by number of retweets; numbers are taken from the data set and may have changed since. NHS = National Health Service.
a. This name has been altered due to concerns with the privacy of the account owner(s).
Table 5 indicates the prominence of two different type of elites in the Twitter conversation. On one hand, there are a number of activists, with thousands of followers and a history of digital campaigning (especially on issues of race and the national health system), who became “crowdsourced elites” (Papacharissi and de Fatima Oliveira 2012) on the issue. The activists among the top 20 accounts were responsible for 10 percent of the retweets in the overall Windrush conversation and a remarkable 26 percent when looking only at the prestorm period, which indicates how important these few individuals were early on (see Supplementary Information). Second, there are several politicians on the list, led by David Lammy, whose tweets were shared so widely that they made up a remarkable 10 percent of all retweets in the Windrush conversation. Lammy is a Labor MP and child of parents from the Windrush generation who was highly involved in the issues from day one. He took the window of opportunity—and his strong digital presence—to campaign for the Windrush victims, and race and immigration issues more generally, as well as using the issue strategically for the purposes of party-competition. Lammy aside, the number of politicians among the opinion leaders in the Twitter conversation shows their enduring importance in policy debates in the media, not only, as consistently shown in the literature, in professional news coverage, but also in the Twittersphere.
In contrast, with the exception of The Guardian’s Amelia Gentleman at #2, journalists and news legacy media accounts are not that prominent, with only one other journalist (@mrjamesob, a talk radio host with a strong record of speaking-out online) in the top 20. This does not mean, however, that legacy news media did not play an important role in the Twitter conversation. But they did so not so much through the journalists’ or media’s Twitter accounts but, as we shall see, via the news items they published which were widely shared.
Table 6 shows analysis of the hyperlinks included in any tweets about the issue between March 1 and May 31. First, the role of The Guardian is again striking, not only in breaking the story but also in amplifying it: over 30 percent of the URLs in tweets came from the newspaper. Second, it was other legacy news media organizations who were the object of the majority of the links on Twitter, with the BBC playing a leading role. This is an important indicator of their enduring importance in creating and sustaining a national conversation, even, or rather especially, in the highly fractured audience environment that characterizes contemporary media systems. Finally, fascinatingly, this list shows the range of organizations and platforms that play a role in the process in the hybrid media system, ranging from the electronic petitions and YouTube—where mostly videos from legacy media about the topic were shared—to digital native news sites, partisan blogs, and digital advocacy organizations. Altogether, the analysis of the most visible twitter accounts and the hyperlinks highlights the diverse set of actors involved, including both new and hybrid, but also the enduring importance of political elites and legacy news media. It also shows the interdependence across platforms, including between electronic petitions, social media sharing, and coverage in legacy news media.
Most Often Linked Domains in Tweets Referencing to Windrush.
Note. URL = Uniform Resource Locator.
a. Different URLs with the same organization domain were aggregated. Some tweets did not contain hyperlinks or only referenced twitter, that is, quoted other tweets, and so are not included (1,553,097). Of the above, some included one URL (210,878) and others more than one (5,983).
Discussion and Conclusions
We set out to examine the roles of different types of media in political agenda setting in the hybrid media system. How do they interact with other actors and platforms? How do they contribute to reaching the attention “tipping-point,” pushing the issue into the government’s agenda and help to overcome inaction? Our case study shows the continuing importance of legacy news media in political agenda setting, especially in storm mode. It also uncovers the distinctive roles of other actors as well as clear evidence of similarities in patterns of coverage across media types.
The analysis shows that, in the case of Windrush, the issue was first brought into the public sphere by persistent coverage by an investigative journalist in a legacy news media organization at the forefront of digital reinvention, joined by a strategic alliance of advocacy and political elites. For these elites, this was not “news,” there was no revelation from the media coverage (cf. Sevenans 2018). But the early coverage by The Guardian was important as a signal of a window of opportunity, and so was the attention and re-framing of the issue enabled by the digital petitions gathered by actors and technical artifacts in the campaign assemblage, including crowdsourced leaders (Papacharissi and de Fatima Oliveira 2012) and hybrid advocacy organizations.
But, perhaps surprisingly, given the emphasis on the new and digital in some of the literature, the case shows that legacy news media—such as the BBC, Channel 4, and Daily Mail, among others—played a crucial role. It was only when they reached “storm mode,” and both media and political actors tried to outrun each other in a sudden frenzy, that friction against action was overcome, and there was a government response. We uncovered three key roles for legacy news media in this process: initiating, amplifying, and sustaining attention to the issue. First, a news legacy outlet (The Guardian) raised the issue and gave it persistent coverage when practically no one else was paying attention, helping to gather and energize the campaign assemblage and to generate a sense of ongoing-ness about the issue. It was later also crucial in cross-media diffusion, by providing both the material about individual cases for news media rehashing in their coverage, and the most popular hyperlinks for social network sharing. Arguably, a similar initiating role could be played by a different type of media. But The Guardian is a remarkable example of organizational hybridity among news outlets and was hence in some ways uniquely positioned. It combines, in particularly successful ways, old and new media logics, including the resources (budget, journalistic expertise, and access), prestige, and brand recognition and loyalty of top legacy news organizations with the enhanced reach, timeliness, promotional opportunities, and constant feedback of digital. Moreover, its current business model partly relies on donations, some project-specific, which encourages sustained investigative stories like Windrush.
Second, when legacy news media other than The Guardian eventually devoted significant attention to Windrush, they were key to amplifying the visibility of the issue and compelling the government’s reaction. It was also only at this point that public attention increased substantially. This response was very much in unison, that is, at the same time and following very similar patterns across legacy news media outlets. Moreover, it quickly moved into “storm mode,” receiving longer sustained attention from the news media and an explosive response from government. This leads us to the third, and key, role of legacy news media: sustaining attention over time, and hence putting continuing pressure on the government both directly and indirectly by facilitating the opposition’s role. In turn, this fed the newsworthiness of the issue. The key is that there was a reciprocal influence between strategic elites (advocacy and opposition) feeding the newsworthiness of the issue, the governmental response, and the legacy news media attention in ratcheting effect and at a higher pace because of the storm mode.
Interestingly, it is in fact some of the most-criticized characteristics of legacy media that helped to sustain the attention: first, their “obsession” with process, partisan battles, and political scandal; second, how in mediatized fashion elites strategically fed this (news)media logic; and finally, the pace of the news cycle. Although historically criticized for its short attention span, the legacy news media (including those like The Guardian with a strong digital presence) do not tend to move into and out of stories as quickly as digital actors in the system. Stories that “go viral,” due to digital networked sharing by multiple nodes, receive fast large-scale attention that tends to fade quickly (Nahon et al. 2011). In contrast, although news media storms are also characterized by high explosive attention, the attention is more likely to be sustained, which is to do with professional norms and allocation of resources, especially when in storm mode. Specifically, there is inter alia a lowering of gatekeeping thresholds for the story and related issues, resources re-allocation to the coverage of the topic, and higher imitation—and thus competition—across multiple news outlets (Boydstun et al. 2014). It is then in their interest to stay “locked” in the storm, especially, as in this case, if elites keep talking about it and providing fresh newsworthy angles on the issue. In contrast, new digital actors (e.g. whether crowd-based such as viral memes or hashtags, or blogs and small digital media organizations) do not have the same institutional constraints to move fast into a breaking story nor, conversely, the same incentives to “stay locked.”
Although these dynamics can make legacy news media slower to react, they also give them potentially larger and more sustained discursive power (Jungherr et al. 2019), especially as the battles for attention are becoming fiercer, collective attention spans shorter (Lorenz-Spreen et al. 2019) and audiences more fractured. Clearly, legacy news media are no longer the only path to attention. But, as in this case, in many media systems, they are still at the core of national conversations, directly due to enduring reach and resources, and indirectly by providing a large proportion of the original news content referred to on other platforms. This role is key and has significant implications for agenda setting and beyond, both in terms of what is included—as in this case—but also excluded, if not fitting with the legacy news media norms. Moreover, it indicates that, in this communication environment, the dynamics of media storms (and the legacy news media’s roles in them) are likely to become ever more important.
These insights are the result of deep-diving into one case. This approach also means that our analysis has some limitations in terms of generalizing beyond the particularities of the case and the political and media systems in which it took place. But it is an important starting point for comparative work, and hence has cross-national relevance. The impact of digital disruption on the communication environment, and thus on the relationship between media and politics, is global, but its effects are undoubtedly mediated by national characteristics. This article uncovered the enduring significance of legacy news media organizations and their relationship with elites, especially in storm mode. But the features of the U.K. media system, with its very strong legacy news media core that has now translated also into digital consumption patterns (Majó-Vázquez et al. 2019), were key to the dynamics we uncovered. This is different from the United States, the focus of most of the previous studies, but not unlike other countries in Western Europe, that also have strong public service broadcasters and other legacy media organizations with large reach, prestige, and resources. In which ways these dynamics are shaped by the characteristics of different media systems, including the strength of legacy organizations but also, for example, the degree of political parallelism, is a key question for future research to pursue. Equally, what happens in cases where, although there is (digitally enabled) activism, there is no strong elite involvement nor partisan and process dimensions, potentially dampening the interest of legacy media? We hope that the insights from this article and methodological approach are applied in the future to other cases to answer these questions.
Supplemental Material
sj-pdf-1-hij-10.1177_1940161220925023 - Supplemental material for Political Agenda Setting in the Hybrid Media System: Why Legacy Media Still Matter a Great Deal
Supplemental material, sj-pdf-1-hij-10.1177_1940161220925023 for Political Agenda Setting in the Hybrid Media System: Why Legacy Media Still Matter a Great Deal by Ana Ines Langer and Johannes B. Gruber in The International Journal of Press/Politics
Footnotes
Acknowledgments
We would like to thank Amber Boydstun, Lucas Graves, and Maggie Scammell who gave us very useful feedback in earlier versions of the paper, as well as the four very helpful IJJP anonymous reviewers. We would also like to thank our interviewees for giving up their time to answer our questions and for fighting, in different ways, for the victims of the Windrush scandal.
Authors’ Note
This paper was first presented at the annual conferences of the European Consortium For Political Research (ECPR), International Journal of Press/Politics (IJJP), and the International Communication Association (ICA); we would like to thank discussants and participants for their comments.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
Supplemental Material
Supplemental material for this article is available online.
Notes
Author Biographies
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
