Adopting or resisting 24-hour news logic on evening bulletins? The mediatization of UK television news 1991−2012

Making sense of the form, structure, and style of television journalism

Television news bulletins can be broadly defined as ‘a regularly scheduled program dedicated to the promotion of the top stories of the day and important issues for the audience, utilizing visuals, words and audio in a format not directly copied by another medium’ (Conway, 2009: 4; his emphasis). In the decades post–World War II, journalists had to adjust from the medium of radio to television, developing a format that emphasized its visual potential. Compared to the classic set of news criteria Galtung and Ruge (1965) proposed, television offers its own unique set of values, such as visual attractiveness and brevity (Golding and Elliott, 1979). Unlike print, radio, and, to a lesser extent, online news, television news is driven by moving images, and given the relatively short time span of bulletins, these need to be carefully packaged and edited.

However, establishing how far television news has modified its length and selection of routine conventions over recent decades is empirically difficult to pin down. Impressionistically, scholars have observed that the speed of television news storytelling has increased, with enhanced production values reshaping television conventions (Gitlin, 1987; Hallin, 1994). So, for example, writing about the 1980s, Hallin (1994) observed that the ‘character of the evening news … changed substantially. Its pace has come to resemble more closely the pace of the rest of commercial television, with television 10 soundbites and tightly packaged stories’ (p. 177). Indeed, there is a voluminous set of empirical studies tracing the shrinking of political soundbites, but far less is known about the changing form and structure of television news (Esser, 2008; Hallin, 1992). In a comparative study of UK television news bulletins in 1993 and 1996, Harrison (2000) measured the comparative length of stories (rather than conventions) and found shorter length items. Although the drop was not dramatic, she noted that ‘it is of some concern if it results in less information being transmitted or indicates a long term trend’ (Harrison, 2000: 165).

In our study, we attempt to systematically assess whether, in fact, a longitudinal trend is evident. In doing so, we categorize different television news conventions – such as edited packages or live two-ways, which we explain further below – and assess whether these different conventions have changed in length over time. Since the broad scholarly consensus suggests that the pace of television news has increased, with shorter soundbites and story lengths, our first hypothesis expects to find that:

The television news anchor is a central figure in television bulletins, presenting and introducing different items. However, historically producers were unclear about whether an anchor should have a strong visual or aural presence in the delivery of news (Conway, 2009). Between 1944 and 1948, according to Conway (2007), US network channel, CBS, employed over a dozen different anchors. But soon journalistic personalities emerged – most famously, Walter Cronkite, then Dan Rather – and anchors were embedded into the format of bulletins. According to Meltzer (2010), the personality of anchors has become an increasingly important part of US bulletins, stamping their authority and identity on the programme. In other countries, such as the UK, television news anchors have not occupied such a central focus in programming – and are sometimes known as merely ‘presenters’ or ‘newsreaders’ – to reflect the more formal, authoritative, and detached position in the narration of news (Hartley, 1982; Montgomery, 2007). This, of course, is shaped somewhat by strict impartiality requirements in UK broadcasting.

In understanding the role of news anchors in television bulletins, however, most scholarly accounts tend to be impressionistic, qualitatively exploring their role, characteristics, and performance. Our study, by contrast, will quantitatively explore whether their role in communicating news has changed over time, notably in how much they narrate individual news items compared to more edited packages or live reports. Given the increasing prominence of anchors in television bulletins, our second hypothesis is thus:

Needless to say, as more sophisticated technologies have evolved, the production of news has changed dependent on what is editorially possible – for example, sending an edited package from far-flung locations or broadcasting live through satellite. In doing so, this could editorially influence the type of conventions that routinely shape an evening bulletin and the information provided. Montgomery (2007) has examined broadcast news in detail – including news bulletins and 24-hour news channels – and his distinction between the two is instructive:

The role of the presenter/anchor on the dedicated news channels depends more upon unfolding the events through interview with correspondents, commentators or ‘newsmakers’. Their role is less scripted [than half-hourly news formats], more improvised, and more of a question of blending together, in the real time flow of broadcasting, the voices that are drawn upon to make up the news. (pp. 60–61)

Our study is designed to quantitatively explore the character of fixed time bulletins to ascertain whether, over previous decades, they have adopted a more improvised, fast-moving journalism with more live news conventions.

Many empirical studies of television news, however, focus on the agenda of news (e.g. crime, health or politics) as opposed to measuring the changing structure and format of bulletins. But there is a growing body of scholarship that has charted the journalistic thirst for more live, real-time, news action. Indeed, as technology has made instant forms of communication on television – including dedicated 24-hour news channels – far easier, this has fueled a wider journalistic culture of ‘going live’ (Seib, 2001). This 24-hour news cycle brings pressures to deliver news live immediately, rather than in a more edited format. This is particularly the case in political journalism, in which scholars have explored how the perpetual demand for news puts pressures on politicians to react in real-time.

Once again, however, there is little systematic evidence to demonstrate a shift in live reporting over recent decades. While empirical studies have demonstrated rolling news channels increasingly using a ‘breaking news’ tag (Lewis and Cushion, 2009) or selecting what have been labeled live event news stories (Livingston and Bennett, 2003), the proportion of live news in fixed time bulletins has not been subject to any sustained longitudinal analysis. We are left, in other words, with anecdotal accounts rather than a precise empirical record of changing live conventions or the impact of the 24-news cycle.

Tuggle and Huffman (2001), however, have carried out a study of 24 local television news stations in the US and found all of them ‘aired more stories containing a live element than standard reporter packages’ (p. 339). More recently, cross-sectional studies examining the amount of edited news compared to live news in US, UK, and Norwegian evening bulletins have been carried out (Cushion et al., 2014a, 2014b). It was discovered that in the US and UK especially, journalists routinely reported live in different contexts, most prominently by way of two-ways. Moreover, in a more detailed UK study, live news was found to be disproportionately about politics, with political editors occupying significant ‘air time’ when reporting politics (Cushion and Thomas, 2013).

Our article will build on these studies by interpreting whether the proportion of live news has grown steadily since 1991 and, further still, whether political reporting is the topic most likely to be covered live. Our third and fourth hypotheses therefore expect that:

If these hypotheses are confirmed, it will mean a shift in the form of news conventions, with a greater use of broadcasting live on location, thus asking journalists to not only pre-edit stories but to explain them live ‘on-air’. This has become known as a live two-way, where – as Montgomery (2007) explains – a ‘short exchange live from the studio with a correspondent at the scene’ often follows an edited package involving the same reporter (p. 119). Within the studio, a more recent convention identified in television news is ‘the structured panel discussion’ (Patrona, 2012: 147; her emphasis) involving an exchange between anchor and reporter. Patrona (2012) suggests this is ‘a new form of broadcast news interaction, where prominent journalists engage in live conversation on a designated news story’ (p. 158). In other words, television journalists appear to increasingly operate in real-time contexts, moving away from tightly edited and scripted conventions toward more improvised and live formats.

In doing so, it could potentially deliver greater interpretation in news coverage, a trend well documented in journalism studies and political communication (Patterson, 2000). The precise meaning of how interpretive journalism is measured can differ between scholars, but a review of empirical studies using this term defined it as a form of journalism that goes ‘beyond descriptive, fact-focused and source-driven journalism’ (Salgado and Strömbäck, 2012: 154). While the rise in interpretive news has been evidenced across different media, on television most empirical studies relate to changes in political reporting or, more specifically, in elections over time. In many cases, this is seen to reflect a journalistic goal to de-spin the increasingly sophisticated machinery of party politicking (McNair, 2000). So, for example, Steele and Barnhurst (1996) identified a significant reduction in factual information – from 24.6 per cent in 1968 to 2.4 per cent in 1998 – on US network presidential coverage. De Vreese (2001), meanwhile, found more analytical reporting in successive elections on the main public service broadcaster in the Netherlands – on NOS Journaal – with a greater propensity to interpret party political spinning. Likewise, in the UK, journalists covering elections have gradually spent greater time interpreting politicians’ motives and covering party political maneuvering (Deacon and Wring, 2011; Semetko et al., 1991).

The evidence suggests that as television news reports less factually about events and issues – notably in the world of politics – journalists are placed more centrally in the narrative of storytelling, for they increasingly appear to be asked to form judgments and volunteer opinions, correspondingly leaving less time for politicians’ voices to be heard. Indeed, comparing coverage of election news on Danish television between 1994 and 2007, Hopmann and Strömbäck (2010) confirmed that the visibility of journalists on-screen and acting as commentators had risen, with political sources marginally decreasing. This, of course, is consistent with previous studies showing political sources have been given less time and space to air their views (Esser, 2008; Hallin, 1992).

Since our review of longitudinal empirical studies suggests television news has become increasingly interpretive over recent decades – largely based on the analysis of election reporting – with journalists adopting a more active role in communicating what is happening in the world, two more hypotheses include:

In reviewing empirical studies about television news – which have explored the changing form, structure, and style of its conventions – it should also be acknowledged that many scholars consider underlying commercial forces and market-driven logic to be central agents in explaining changes in journalism. So, for example, packaging shorter news items, using news anchors more prominently, broadcasting more live reports, and sourcing less external sources, as well as reporters interpreting and reacting to events live on-air have been observations directed more explicitly toward commercial media systems as opposed to public service broadcasters (Semetko et al., 1991; Strömbäck, 2008). Our final hypothesis thus anticipates that:

The mediatization of news and journalistic interventionism: an analytical and methodological framework for measuring fixed time television news bulletins

Taken together, many of the trends identified in our review of the previous literature – of a greater interpretation in television news reporting, of journalists being more central to the news narrative, or adopting live, rolling news conventions – represent characteristics of the process scholars have labeled mediatization. Broadly defined, this concept has been used to interpret the degree to which the media influence different facets of society leading to changes in how they function over time (Lundby, 2009). Put another way, as the media have become a more pervasive force, the extent of their collective power in different countries manifests itself in a media logic that informs the behavior of institutions. Altheide and Snow (1979) are often credited with first using the term ‘media logic’, but today it is applied in more specific contexts. So, for example, in debates about the mediatization of politics, political coverage is seen as being mediatized if journalists rather than politicians increasingly appear and interpret news since this reflects a media logic superseding a political logic (Strömbäck, 2008). Alternatively, if television news bulletins began to adopt formats increasingly dominant in broadcast or online journalism – including quicker, live, and instant news conventions – it could similarly be interpreted as a mediatization of television journalism, a triumph of 24-hour news logic trumping fixed time news programming. While the media mediatizing itself might appear counterintuitive, it suggests there are multiple logics operating in media culture, rather than a singular force.

In order to operationalize the changing format and conventions of television news bulletins between 1991 and 2012, we draw on the concept of journalistic interventionism. The concept broadly refers to what Semetko et al. (1991) have labeled the ‘discretionary power’ of journalists, and it has primarily been used to characterize interventions in peace journalism or, more relevant to this study, to interpret how journalists cover elections (p. 3). In our view, journalistic interventions can be operationalized beyond election campaigns and can be identified in the routine use of editorial conventions. In this study, we interpret routine conventions as ‘journalistic interventions’ – for example, making an editorial decision about reporting a news story in a live or edited format – with the aim of evaluating whether fixed time news programming has changed over recent decades and adopted a media logic more reflective of 24-hour news values or other forms of instant journalism. Or, put more succinctly, we ask whether there is any evidence of a mediatization of television news bulletins by examining whether bulletins conform to a fixed time or rolling news logic.

As previously acknowledged, evening television news bulletins have historically been defined as reporting ‘the top stories of the day’ (Conway, 2009), rather than – as many 24-hour news channels now claim – to ‘bring news when it happens’. We can theorize the competing logics of fixed time programming versus rolling news accordingly: on the one hand, a fixed evening time bulletin logic would deliver mostly edited news packages, relying on many external sources or anchors carefully narrating a story with the help of an autocue. On the other hand, a rolling news logic would be less pre-scripted, with anchors narrating more news stories at a rapid pace while live reporters react, deliver updates, or even interpret the latest news action (Montgomery, 2007).

To empirically operationalize different journalistic interventions, we developed four routine television conventions previously piloted and tested in other studies (Cushion et al., 2014a, 2014b; Cushion and Thomas, 2013): (1) edited packages from a reporter out on location sent back to the studio; (2) standalone items that involved just television news anchors, where they talk over an item, often with moving images or a picture in the backdrop; (3) a reporter or guests having a live discussion with the anchor in the studio; and (4) either a live two-way between anchor and reporter or where there is no interaction, just the journalists reporting from outside the studio. These types of interventions range from a more tightly edited, structured, and externally source-reliant format to a less scripted, live, improvised, and potentially more interpretive form of television journalism. The length of different interventions will also be measured over time to interpret whether news has become shorter – reflecting a quicker pace inherent in 24/7 media culture (Rosenberg and Feldman, 2008) – or lengthier, dealing with less stories but reporting them in more depth.

To explore UK evening television news bulletins longitudinally, we carried out a quantitative content analysis of BBC and ITV early evening bulletins over 3 weeks in 1991/1992, 1999, 2004, and 2012. ² The BBC is a public service broadcaster regulated closely, whereas ITV is a commercial public service broadcaster operating under a more ‘light-touch’ regulator. ³ Channel 5’s early evening bulletin was also examined in 2012. It is a commercial public broadcaster launched in 1997. Since ITV is a long-standing public service broadcaster, attracts larger audiences than Channel 5, and has both national and regional commitments as part of its broadcast license, it could be argued that Channel 5 is subject to less regulatory baggage than ITV. We can thus ask – as our final hypothesis predicted – whether a greater degree of mediatization is present on commercialized and more lightly regulated media systems.

Overall, the study generated 2040 news items with the four different types of journalistic interventions acting as our unit of analysis. All news items were examined over the course of the bulletin, along with the story subject reported (politics, crime, education, health, etc.) and the volume of on-screen and off-screen sources (e.g. journalists referencing a source ‘the Prime Minster said today …’). In order to explore the purpose of live news in different journalistic interventions, we drew on a previous study that classified the primary purpose of live interpretive news in four ways: (1) to supply the ‘latest’ news, (2) to offer interpretation to a story or issue, (3) to report from a location, or (4) to introduce the beginning of an edited package or summarize the end of it (Cushion and Thomas, 2013).

All the material was coded by two UK researchers, with regular team meetings to ensure all data were interpreted consistently and according to strict criteria. In order to test the reliability of coded data, we used Cohen’s kappa (κ), one of the most conservative intercoder reliability tests, since it takes into account the measurement of chance between coders, rather than simply measuring agreement. Cohen (1960) interprets kappa co-efficients in the following ways: <0 indicates less than chance agreement; 0.01–0.20, slight agreement; 0.21–0.40, fair agreement; 0.41–0.60, moderate agreement; 0.61–0.80, substantial agreement; and 0.81–0.99, almost perfect agreement. By re-coding roughly 10 per cent of the overall sample, our intercoder reliability test found substantial or almost perfect agreement across all variables.

Overall, the study asked to what extent have television news bulletins become shorter, less edited and scripted, and more live, improvised, and interpretive between 1991 and 2012? Or, put differently, to what degree have television news bulletins become mediatized over the last 20 years, adopting a rolling news logic in how they routinely report news?

Findings

As expected, we found that the length of items on BBC and ITV television news bulletins had shortened over time. In 1991/1992, the mean average was 2 minutes and 1 second; in 1999, 1 minute and 39 seconds; in 2004, 1 minute and 31 seconds; and in 2012, just 1 minute and 5 seconds. ITV, meanwhile, had a mean average of 1 minute and 33 seconds in 1991/1992, 1 minute and 40 seconds in 1999, 1 minute and 33 seconds in 2004, and just 59 seconds in 2012. In short, while the BBC cut the length of its news items by half, ITV had cut it by a third. H1, in this sense, is clearly confirmed since both broadcasters reduced the length of routine news items by a considerable degree, with the commercial broadcaster – supporting H7 – editing, on average, shorter items (less than a minute long in 2012).

However, in order to understand where and how bulletins have changed – and to answer H1 more comprehensively – we need to compare the average length of different journalistic interventions (see Table 1). While edited packages and the role of anchors have remained broadly the same length, live news items have shortened, most strikingly on the BBC, where the mean length of two-ways reduced from 1 minute 46 seconds in 1991/1992 to just 46 seconds in 2012. Needless to say, this places considerably greater restraint on the time reporters have to explain a story on-air. While the reduction in the average live news item is most pronounced on the BBC, it is Channel 5 news that has the shortest mean length (just 35 seconds, for example, in the use of two-ways).

OPEN IN VIEWER

Table 1.

Mean length of journalistic interventions on UK television news bulletins between 1991 and 2012.

	1991/1992		1999		2004		2012
	BBC	ITV	BBC	ITV	BBC	ITV	BBC	ITV	Channel 5
Edited
Reporter package	2 minutes, 15 seconds	1 minute, 42 seconds	1 minute, 59 seconds	2 minutes	2 minutes, 18 seconds	2 minutes, 12 seconds	2 minutes, 19 seconds	1 minute, 27 seconds	2 minutes, 3 seconds
Anchor	27 seconds	23 seconds	22 seconds	24 seconds	25 seconds	28 seconds	20 seconds	23 seconds	24 seconds
Live
Anchor/reporter two-way/live location	1 minute, 46 seconds	1 minute, 5 seconds	1 minute, 15 seconds	1 minute, 8 seconds	55 seconds	44 seconds	47 seconds	1 minute	35 seconds
Anchor discussion	–	–	1 minute, 4 seconds	1 minute, 35 seconds	–	1 minute, 20 seconds	57 seconds	52 seconds	38 seconds
Total N	199	136	223	142	233	205	325	285	292

But how have the form, structure, and style of television news bulletins changed overall in recent decades? In respect of the role of anchors, Table 2 shows their on-screen presence – on both broadcasters – has substantially increased as a proportion of all ‘air time’ over recent decades. On the BBC, anchors took up 2.8 per cent of all air time in 1991/1992, then rising gradually to 18.6 per cent in 2012. ITV anchors, by contrast, appeared in just a minuscule proportion of coverage in 1991/1992 (1.4%) and 2004 (0.6%), but their on-screen presence rose dramatically in 2012 to 19.9 per cent. Channel 5 featured anchors the most (21.2%). Thus, H2 is fully supported with the BBC and ITV substantially enhancing the presence of anchors during their evening bulletins between 1991 and 2012. Although the more commercially driven bulletins featured anchors more than the public service broadcaster, the difference was marginal providing lukewarm support for H7.

OPEN IN VIEWER

Table 2.

Percentage of journalistic interventions on UK television news bulletins between 1991 and 2012 (by time).

	1991/1992		1999		2004		2012
	BBC	ITV	BBC	ITV	BBC	ITV	BBC	ITV	Channel 5
Edited
Reporter package	95	89.4	89.1	89.9	79.4	73.7	67.9	63.3	63.7
Anchor	2.8	1.4	3.5	4.1	6.8	0.6	18.6	19.9	21.2
Live
Anchor/reporter two-way/live location	2.2	9.3	7.1	5.4	13.8	17.6	10.8	14.1	14.9
Anchor discussion	–	–	0.3	0.6	–	8	2.7	2.8	0.2
Total N	199	136	223	142	233	205	325	285	292

While the visibility of anchors increased overall, the average length of news items remained roughly the same (see Table 1). We can thus conclude that news anchors today narrate far more news items per bulletins, but they remained relatively short (20−23 seconds, on average), quickening – as H1 confirmed – the pace of news reporting in television bulletins. This again places restrictions on the time afforded to anchors to explain or contextualize a story.

As expected, edited packages remained the most dominant journalistic intervention between 1991 and 2012. However, on both the BBC and ITV, the amount of time granted to reporter packages – as a proportion of the whole bulletin – declined gradually by over a quarter on both channels. This represents a significant shift in the character of news bulletins, which, according to Montgomery (2007), have long relied on edited packages to report what is happening in the world. By 2012, the BBC dedicated just over two-thirds of an average bulletin to edited packages (67.9%), with commercial channels using this convention marginally less (63.3−63.7%).

While the greater use of anchors explains part of this shift, an enhanced use of live news on both broadcasters also explains why edited packages featured less in reporting overall. Between 1991 and 2004, two-ways rose by more than six times on the BBC (from 2.2% to 13.8%) and almost doubled on ITV (from 9.3% to 17.6%). But while the use of two-ways dropped slightly on both channels in 2012, the greater use of studio discussions in 2012 meant overall live news was almost the same as in 2004. This confirms Patrona’s (2012) observation that the format of live studio discussion appears to be shaping television news bulletins (although not on the most commercialized broadcaster, Channel 5, where two-ways was the dominant live intervention). Overall, then, H3 is fully confirmed in that live news between 1991 and 2012 increased steadily, with correspondingly less edited news. While the public service broadcaster relied most on edited packages to report news, the differences with commercial broadcasters were not striking (again, offering lukewarm support for H7).

So far, our analysis has painted a fairly broad picture of news generally. But the review of the previous academic literature revealed that political news was singled out for being most susceptible to live reporting due to the fast-changing nature of contemporary politics (Barnett and Gaber, 2001). Our findings support this proposition, with the proportion of live political news far higher than all news (see Tables 2 and 3).

OPEN IN VIEWER

Table 3.

Percentage use of different journalistic interventions in political news items on UK television news bulletins between 1991 and 2012 (by time).

Politics news	BBC				ITV				Channel 5
	1991/1992	1999	2004	2012	1991/1992	1999	2004	2012	2012
Edited news
Reporter package	100	83.6	66.1	65.5	81.5	86.1	68.8	51.4	53.9
Combined anchor	–	2.9	5.2	15.4	–	2.1	–	19	15.4
Total edited news	100	–	–	80.9	81.5	–	68.8	70.4	69.3
Live news
Anchor/reporter two-way/live location	–	13.5	28.7	16.3	18.5	11.9	27	24	30.7
Anchor discussion	–	–	–	2.8	–	–	4.2	5.6	–
Total live news	–	13.5	28.7	19.1	18.5	11.9	31.2	29.6	30.7
Total N	19	30	19	88	8	17	28	53	57

We identified some notable changes in the use of live political news over recent decades. The BBC in 1991/1992, for example, had no live political news compared to 19.1 per cent on ITV. On both channels, this increased substantially into the millennium, peaking on the BBC (28.7%) and ITV (31.2%) in 2004. The public broadcaster reduced its live political coverage to 19.1 per cent in 2012, but ITV remained roughly the same (29.6%). Channel 5, by contrast, dedicated the most live coverage in 2012, all of which was through live two-ways (compared to some live studio discussions on the BBC and especially on ITV). Correspondingly, the proportion of time spent on edited packages in political news on both the BBC and ITV fell steadily from 1991 to 2012. Overall, then, we can conclude convincingly that – as H4 predicted – political news has become a disproportionately live topic compared to news more generally over time.

Our findings have shown live journalistic intervention changes the form and structure of news, delivering shorter news items with an increase in live exchanges between anchor and reporters. To explore the content of these live news interventions more closely, we thus evaluated its primary journalistic purpose. Our hypothesis suggested live news would become increasingly interpretive, especially in political reporting. However, since the proportion of live news was low both in 1991/1992 and in 1999, our findings should be treated cautiously. Nevertheless, Table 4 shows interpretation was the primary purpose of live news in 1991/1992 and 1999, but in 2004 live news – notably two-ways – were also used (by ITV especially) to a large extent to introduce/summarize an edited package. These tended to be shorter items, however, lasting just 10–15 seconds on occasions, since the reporter only introduces their package as opposed to offering lengthy interpretations. Introducing/summarizing an edited package in live news was used less frequently in 2012 on ITV and BBC, with interpretation the most dominant function for live news (although less so on Channel 5).

OPEN IN VIEWER

Table 4.

Percentage of journalistic interventions in live reporting on UK television news bulletins between 1991 and 2012 (by volume).

	BBC				ITV				Channel 5
	1991/1992	1999	2004	2012	1991/1992	1999	2004	2012	2012
Latest news	40	10.5	11.3	13.8	27.8	38.5	23.7	20.4	23.5
Interpretive	60	52.6	45.3	55.2	72.2	61.5	15.8	57.1	39.7
On location	–	5.3	5.7	10.3	–	–	–	2	–
General intro/summary	–	31.6	37.7	20.7	–	–	59.2	20.4	36.8
N	5	19	53	58	18	13	76	49	68

With the exception of 2004, we can thus conclude that the primary purpose of live news was to offer interpretation. Moreover, this was openly encouraged at times by the format of live two-ways. So, for example, in a two-way BBC item on 19 February 2004 about the anniversary of the Iraq war, John Simpson – an experienced world affairs editor – was simply asked by the anchor ‘So, John, what’s your judgment’?

However – to fully answer H5 – it would be difficult to conclude with certainty all live news had become more interpretive since live news did not meaningfully shape evening bulletins until after the new millennium. H5 also suggested political news would be increasingly interpretive over time. Although interpretation was the primary purpose of live political news in 1991/1992 and 1999, politics was only reported live just a handful of times. In 2004, the BBC reported seven live political items, all of which were interpretive (100%, a higher level of interpretation compared to other subjects). ITV meanwhile had 17 live political news items, 5 of which were interpretive (29.4%, again higher than other topics), the same number used to introduce/summarize an edited package. The remaining two brought the latest news. In 2012, the BBC ran 14 political items, ITV featured 17, and Channel 5 had 25, of which 11 were interpretive on the BBC (78.6%), 10 on ITV (58.8%), and 16 on Channel 5 (64%). Moreover, interpretive news tended to be the longest type of live news. In 2012, for example, 88.9 per cent of live political news was interpretive on BBC, 83.3 per cent on ITV, and 75 per cent on Channel 5.

In sum, interpretation was the primary purpose of live news in political items compared to other topics (crime, business, health, etc.). Our findings thus support H5 – that live political news tends to be the most interpretive topic reported and it has become more interpretive over the last decade or so. However, since our overall N in the 1990s was small, how far live news has changed remains questionable (in other words, live political news has become more interpretive because more political news was presented in a live format).

The final part of the study examines the use of external sources in different journalistic interventions (including actors who appear on-screen or as references made by journalists). We focus on the use of two dominant interventions – edited packages and two-ways – since both allow reporters the space and time to source within an item (as opposed to much shorter anchor-only items).

As H6 predicted, Table 5 shows that in every year from 1991 to 2012, edited packages contain, on average, a higher ratio of sources per item than live two-ways. Live two-ways – particularly on the BBC – had a far lower ratio (approximately less than 1 per item). The amount of sources per item on edited news actually increased over recent decades and was higher on the commercial channels.

OPEN IN VIEWER

Table 5.

Combined average ratio of sources (on and off screen) by all news items on edited packages and live two-ways on UK television news bulletins between 1991 and 2012.

	BBC				ITV				Channel 5
	1991/1992	1999	2004	2012	1991/1992	1999	2004	2012	2012
Edited packages	3.3	2.2	2.8	4	3	3.8	4	4.2	5.4
Two-ways	1.4	0.6	0.2	0.9	1.2	1	0.9	1.3	0.6

A similar picture emerges of sources used in political edited packages and live two-way news items (see Tables 5 and 6). Once again, live interventions had a low ratio of sources and relied primarily on the journalist to report – even interpret (see Table 4) – political news for viewers. Edited political news, by contrast, contained a higher ratio of sources compared to all news. This was particularly the case on the commercial channels where – with the exception of ITV in 2012 – every edited news package contained, on average, five sources.

OPEN IN VIEWER

Table 6.

Combined average ratio of sources (on and off screen) by political news items on edited packages and live two-ways on UK television news bulletins between 1991 and 2012.

	BBC				ITV				Channel 5
	1991/1992	1999	2004	2012	1991/1992	1999	2004	2012	2012
Edited packages	4.2	3.9	3.1	4.1	5.2	5.8	5.2	4.2	5.4
Two-ways	–	1.2	–	0.9	1.5	1.3	0.9	1.6	0.7

In sum, the findings generally support H6, since television bulletins have increased the proportion of live news – notably in two-ways – meaning they have become less reliant on external sources. However, while the ratio of sources was similar for live two-ways on both the public and commercial bulletins, in edited news ITV and Channel 5 drew on a greater number, on average, per item, thus rejecting H7.

Discussion: toward a 24-hour news logic in fixed time programming?

Overall, our first six hypotheses have – to different degrees – largely been supported. To recap, we found news items had become shorter in television news bulletins (H1); the role of anchors reporting news had increased (H2); live news had steadily increased, with correspondingly less edited news (H3); live reporting was disproportionately about politics (H4); live news had become more interpretive including in live political news, although – as acknowledged – the sample size of political items was relatively small (H5); and, finally, live two-ways relied far less on external sources than edited news, notably in political reporting (H6). While we found news items were generally shorter on commercial rather than public service broadcasters, contrary to H7 we did not find any consistent or major differences over time between public and commercial broadcasters.

To interpret whether routine conventions in television news bulletins had changed over recent decades, we drew on the concepts of mediatization and journalistic interventionism. By treating routine conventions (edited packages, say, or two-ways) as journalistic interventions (Semetko et al., 1991) – editorial decisions that change how a story is conveyed – our findings broadly indicated that the media logic shaping bulletins has changed from reflecting the ‘day’s news’ to covering more ‘news as it happens’ (as news channels now claim). We found an increasing reliance on less pre-scripted material and greater emphasis on live reporting from journalists (rather than external sources) who were routinely asked to instantly react, update, and interpret news, particularly in the world of politics. Or, put differently, we identified a mediatization of television news bulletins, with broadcasters appearing to subscribe to a journalistic logic associated with the delivery of rolling news.

The analytical framework in this study is broadly similar to measures used to empirically test the degree to which political coverage has become mediatized by prioritizing a media over a political logic. In our view, this media logic is too broadly defined, encompassing different mediums, outlets, forms, and structures. Our approach to studying changing television news content, by contrast, has suggested a more nuanced framework to interpret mediatization, developing specific indicators that quantify changing formats, norms, and routines. In doing so, news media logic might be better understood as interpreting competing journalistic logics that operate to produce news differently over time and according to changing cultures and markets. Our approach thus expands upon the way news media logic could be conceptualized.

In addition, our study does not entirely confirm to how media logic is broadly understood. According to Esser (2013), there are three constituents to news media logic: professionalization, commercialization, and, less significant, technological changes. Dealing with the latter first, we would suggest technology is a key constituent in our interpretation of mediatization. A key finding of our study – that live news has become more pervasive over recent decades – was clearly related to improvements in satellite technology. Live news in the 1990s, for example, largely featured still pictures of journalists talking through a phone connection or link-ups in other studios (typically Westminster). By the turn of the millennium, reporters were frequently introducing their edited packages live on location – allowing journalists to visually convey the actuality of their reporting (e.g. outside No. 10 Downing Street). Close to a decade on, live news no longer appears a ‘novelty’ since it has remained widely used making it hard to reach a technologically deterministic conclusion about its use today. Rather, increasing the amount of live two-ways – or, indeed, using anchors more to narrate shorter items more often – appears to be conscious editorial decisions now embedded in the practices – or logic – of fixed time bulletins.

This suggests a change in the professional aspects of television news bulletins – a second constituent of news media logic (e.g. Esser, 2013) – with different norms and routines used to quicken the pace and immediacy of news delivery. In political news, for example, the enhanced use of live two-ways privileges the voices of journalists over external sources in edited packages in order to bring instant interpretation. Similarly, anchors have a more visible on-screen presence today, presenting far more news stories than in 1990s at the same rapid pace. Taken together, both editorial changes appear more consistent with a rolling news rather than fixed time logic. The professional aspects of news bulletins, in this respect, no longer appear as distinctive as they were before the growth of 24/7 media. Barnett et al.’s (2012) explanation for the growth of live two-way news reinforces our argument that 24-hour news values have shaped UK bulletins. They suggest that, for example, the immediacy of live two-ways – delivering the latest updates to a story – and reporters being seen at key locations such as No. 10 Downing Street are reasons why broadcasters have increased live coverage (Barnett et al., 2012).

Where our findings challenge a third constituent of news media logic (following Esser, 2013) is in the proposition that a greater degree of mediatization should be broadly linked to the influence of commercialized media content. We would argue the closely regulated culture of UK broadcasting helps explain the mediatization of evening bulletins. So, for example, the public service broadcaster, on several measures, mediatized content to the same degree over time as the main commercial broadcaster, ITV. While it could be argued that the BBC has become more market-driven, in our view the different types of journalistic interventions used to represent mediatization – of enhancing live or interpretive news – do not necessarily reflect a commercial logic. Rather, we would argue that the form, structure, and style of news need to be more closely assessed in order to make judgments about whether journalism has succumbed to any commercial pressure. In respect of selecting live two-ways over editorial packages, for instance, it could – notably in politics – be designed to better inform viewers rather than relying on political sources (where, of course, impartiality laws typically require an item to be balanced with competing political parties). From the 1990s into the new millennium, political parties became notoriously savvy at evading questions or sticking to party scripts, making journalistic interpretation arguably the most enlightening way of scrutinizing a politician (McNair, 2000).

At the same time, however, since the live two-way is a more improvised form of journalism – broadcast live, after all, without the same degree of editorial control and regulation – misjudgments can and have been made, a case in point being the BBC’s Andrew Gilligan live two-way in 2003 that led to the Hutton Inquiry (Montgomery, 2006). More generally, too, the journalistic value of live reporting could be questioned, with little time to contextualize or explain a story. Our study found the availability of time in live two-way news had been reduced over recent decades (lasting, on average, between 35 seconds and 1 minute in 2012). From a journalist’s perspective, this makes it a challenging prospect to convey the competing policy positions of different parties – a point evidenced by the paucity of sources drawn upon in live two-ways compared to edited packages. This raises important questions – and merits more sustained empirical attention – about the changing form, structure, and style of television journalism and the time-pressured environment in which journalists can adequately inform viewers about what is happening in the world. Nevertheless, while changes in technology and professional practices have undoubtedly changed the underlying logic of contemporary television news bulletins, how far this process of mediatization is entirely motivated by commercial forces needs to be carefully interpreted rather than assumed.

Moreover, we would argue that in order to develop a nuanced understanding about the mediatization of news, it is first necessary to theorize the characteristics of a specific programme/genre under analysis closely and then consider the competing logics that might have influenced the nature of its output over time. In doing so, longitudinal empirical analysis – developing relevant content indicators – can then be designed to quantitatively test the degree to which the form, structure, and style have adopted or resisted competing logics at work in its production, for our study has identified a clear adaptation of 24-hour news values over recent decades in UK evening television news bulletins. In our view, this prevailing logic should prompt further research about its journalistic value and, moreover, its impact on viewers’ understanding and engagement with fixed time news in the future.