Local Journalism when the Journalists Leave Town

“One more nail in the coffin”, said local politician Christer Wiström when the decision became public to shut down the local editorial office in the municipality of Ydre in Småland, Sweden, in November 2014 (Henningsson, 2014). This quote illustrates the perceived importance of an editorial presence for the infrastructure, social fabric and, indeed, the very existence of a municipality. While the quotation is attributed to a local Swedish politician, the issue affects places all around the Western hemisphere as editorial offices shut down. Notwithstanding the closures, these towns are firmly anchored in the same domestic, linguistic and cultural contexts, and their citizens have the same information needs regarding polity and community matters. However, as news organisations move out, it will take more resources for journalists to cover these municipalities in situ. At the same time, new forms of (hyper)local journalism have emerged, raising

hopes and questions about how they fit into a wider news ecology and what they can contribute to their communities. Thus, there are two simultaneous developments in local journalism: legacy journalism is moving out of communities and new forms of journalism are emerging. Sometimes the new forms are referred to as “citizen journalism” in previous research and at other times as “hyperlocal journalism” as they are very similar in character – that is, they are web-based alternatives to legacy media and include citizen participation (cf. Nip, 2006). For clarity, the term “hyperlocal media” will be used throughout this paper.

In light of these parallel developments, scholars ask whether hyperlocal media can fill the gap left by legacy media (Kurpius et al., 2010; Metzgar et al., 2011; van Kerkhoven & Bakker, 2014). An important step towards answering that question is to identify the space that hyperlocal journalism needs to fill to compensate for the absence of legacy journalism. Against this background, this study asks: what happens to the news coverage of a municipality when there is no permanent presence of legacy journalists?

An underlying assumption underpinning both the discussion about retreating legacy journalism and the role that hyperlocal media can fill is the key significance of physical presence. Although rarely articulated in previous research, this is an implicit theory of – let’s tentatively call it – the geography of news, which stipulates that the physical presence of journalists affects news coverage. A broader concept could be the geography of information to include hyperlocal outlets that do not necessarily view themselves as a news organisation. Presence is not understood as simply being on the scene occasionally or when something dramatic happens, but a presence that has longevity in order for journalists to become familiar with the people, the community and their issues. For the purposes of this study and drawing on existing theories from the field of geography, we propose that it is useful to apply the concept of “distance decay” as a theoretical framework when studying the importance of proximity in news in general and the presence of journalists in particular. In short, the theoretical proposition of distance decay suggests that the bigger the space between two places, the greater the “friction of distance” (Abernathy, 2017) will be, since any movement in space will consume resources. Accordingly, under the rule of distance decay (Kent et al., 2006), as things (people, cities, etc.) become further apart their mutual attraction will decrease and their relationships will depreciate. In order to counter distance decay, resources have to be invested to maintain mutual attraction and functioning relationships. In the context of journalism, both the proposition that journalists moving out changes the amount and quality of coverage and that non-legacy hyperlocal actors can contribute to the coverage through their geographical attachment are essentially distance decay arguments.

The connection between geography and news production has previously been highlighted within journalism studies as a criterion of news judgement. Pioneers Galtung and Holmboe Ruge (1965) noted that, geographically and culturally, distant nations and events face a higher risk of being filtered out (although geography is not a factor in Harcup and O’Neill’s [2016] widely referred to criteria of news selection). Thus, in journalism studies, distance decay is more observable the further away the “space of the event” (i.e. where the event is taking place) is from the “space of journalistic production” (i.e. the newsroom). Typically, news judgement criteria can be considered indicators of which news properties have sufficient attraction to counter the friction of distance.

Previous research on the role of geography in news has focused on foreign news (Galtung & Holmboe Ruge, 1965; Hanusch, 2014; Wilke et al., 2012). However, as indicated by other researchers (e.g. Napoli et al., 2017; Usher, 2015), proximity is not only a factor when news happens far away in a foreign land but may also have ramifications closer to home. Moreover, when investigating foreign news, several different forms of proximity are conflated – geographical, political, cultural and social – making it difficult to tease out the individual contributions to distance decay. Thus, by studying a context where the other forms of proximity lose relevance, this study attempts to add knowledge about the specific role that geographical distance plays in the shaping of local news. This, in turn, allows us to gain knowledge about the shape of the news coverage gap when legacy journalists leave and what holes are the most urgent for emergent hyperlocal media to plug.

A competing hypothesis to distance decay is that the absence of journalists in situ can be compensated for, or perhaps even bettered by, digital or other forms of communication, decreasing the role of and need for hyperlocal journalism altogether. However, despite the Internet’s potential to collapse distance and potentially counter the physical absence of journalists, this theory appears to be invalid because geography studies (Hampton & Wellman, 2001; Mok et al., 2010; Murnion & Healey, 1998) show that geographical distance persists online. Understood within the context of journalistic coverage of a certain geographical area, digital communication techniques cannot fully compensate for the lack of actual presence. Indeed, the journalists interviewed in Usher’s study (2015) witnessed distance decay (without explicitly referring to the term); they believed that moving the office of the Miami Herald from the town centre to a location 12 miles away had an impact on coverage because it took longer and became harder to know what was going on in the city and its institutions.

Hence, the lack of an editorial office in a municipality could decrease and otherwise affect its coverage compared with a similar municipality with an editorial office. This, in turn, suggests that hyperlocal media could provide a service to municipalities that would otherwise not be available.

Moreover, the absence of an editorial office will probably not affect all sides equally; neither will the gap increase proportionally or symmetrically, as some entities (e.g. authorities, organisations and corporations) have more resources than others to overcome distance decay by providing information subsidies for the news media.

The local perspective of the news media has sometimes been neglected in academia and in public debate (Friedland et al., 2012; Nielsen, 2015a), even though researchers and other regulatory institutions have presented many negative reports regarding closures of local newspapers and harsh staff lay-offs throughout the Western world in the last decade (Abernathy, 2014; Nygren & Althén, 2014; O’Donnell & Hutchinson, 2011; Pew Research Center, 2015). Democracy as we know it cannot function without journalists monitoring the elected politicians in power on all political levels – nationally, regionally and locally (Barnett, 2009; Barnett & Townend, 2014; Strömbäck, 2015; Truedson, 2015). Depending on how news coverage might be affected by an editorial absence – certain topics could become more frequent at the expense of others and reliance on official sources might increase – this effect on news performance (as detailed in the method section) will also have ramifications for citizens’ abilities to inform themselves about their municipality (de Vreese et al., 2017; Napoli et al., 2017).

Empirical research has, for example, shown that local politicians are more careful with public spending when there is an active local news media presence (Lund, 2010). In addition to holding the local government to account, research has also shown that citizens expect the local media to inform and educate, represent opinions and voices of the citizens, and campaign on matters of public interest (Firmstone & Coleman, 2015). From social and cultural perspectives, the news media shape our sense of belonging to a certain geographical place and help us to navigate the community and its social networks – to “orient” ourselves towards each other (Nielsen, 2015a). Some researchers suggest that local media play a role in the symbolic shaping of “a sense of place” (Hess & Waller, 2014) in today’s mediatised society. Others talk about the local news being a journal that can inspire through good examples or a local chronicle that creates the public memory of a community, a social glue (Costera Meijer, 2010; Hess & Waller, 2016; Nielsen, 2015a; Reader & Hatcher, 2012). The local identity of today’s small communities is challenged by a strong focus on the urbanised metropolitan lifestyle that characterises national or regional media outlets. Local media can therefore embody a kind of cultural counter-power as discussed by Castells (2010) and, for this reason, hyperlocal media could play an important role. Studies show that local proximity plays an important role in how citizens value local news. The local journalist has to be perceived as having social capital in local knowledge or a shared identity, which involves an “investment of time” (Hess & Waller, 2016). The “disconnected” news media could be one of the reasons for the growing mistrust that citizens in remote or small communities have for legacy media.

In addition to the changes in how societies produce and consume information in general and news in particular, a need for new ways to understand and study the media landscape has arisen. One pertinent and influential theoretical perspective among local and hyperlocal news researchers has been the news ecology model. The news ecology model, influenced by the biological ecological system, helps to illustrate and explain the complex system in which news is created and distributed. Using this model makes it easier to grasp the intricate flow of information between different media outlets, which can be both the traditional legacy media and other media sources, such as hyperlocal sites, blogs, citizen journalist websites and so forth. The model points to the importance of understanding news diffusion as a somewhat unpredictable and circular process in contrast to the old legacy media system. For instance, it is important to determine where a news story is first published and how it spreads and circulates throughout the system, how it is framed or altered, and how much of the content consists of original reporting and how much is simply aggregated from other news sources (Lund, 2010; Nielsen, 2015a).

In terms of methodology, researchers often use qualitative and quantitative methods, such as content analysis, surveys and interviews, to investigate distinctive local media systems’ relations to the news ecology model (Anderson, 2010; Anderson et al., 2015; Engan, 2015; Hess & Waller, 2014; Pew Research Center, 2015; Powers et al., 2015). A general conclusion that can be drawn from recent studies is that so-called keystone media play an important role on many levels in the world of news. Keystone media are the channels where most new news is produced (Nielsen, 2015b). Findings from different continents show that local newspapers are usually keystones in the local media ecosystem, implying that the absence of an editorial office also potentially impacts on coverage from hyperlocal media negatively ‒ meaning that, theoretically, hyperlocal media will have a tougher time producing coverage while, simultaneously, the need for hyperlocal media will be greater. The routine coverage of local municipality meetings or the close relationship between a local journalist and the community seems to be the starting point for a large proportion of news stories (Abernathy, 2014; Lacy et al., 2012; Nielsen, 2015b). A study in Denmark (Lund, 2010) shows that national media usually respond quickly and with force when scandals are discovered; however, the story is first broken in the local setting. In either case, full-time employees from the local newspaper are central to fulfilling the role of keystone media in the local setting. Recent debates in regular and social media suggest that routine coverage is being broken due to news organisations reducing costs or focusing on operational flexibility (Nygren & Althén, 2014). As journalists leave municipalities, questions arise about whether the hyperlocal news media will need and be able to take over the role of keystone media, creating news stories that can be circulated through the media ecology.

Empirical research on hyperlocal media shows that while these media are diverse, they also share some basic characteristics. Metzgar and colleagues define hyperlocal media as “geographically based, community-oriented, original-news-reporting organisations indigenous to the web and intended to fill the perceived gaps in coverage of an issue or region and to promote civic engagement” (Metzgar et al., 2011: 774). It should be noted that there are substantial overlaps between the notion of hyperlocal media and citizen journalism (and its precursor, civic/public journalism) as they share some traits: they are mostly web-based, have been started as alternatives to legacy media often with an emphasis on citizen participation and seek to drive civic engagement in some form (cf. Nip, 2006). The key difference for the purposes of this study is the focus on geography and original reporting in hyperlocal media. Due to their similar ambitions and, in some cases, problems regarding the different alternative forms of journalism in the news ecology, research dealing with both hyperlocal media and citizen journalism is taken into consideration.

While participants in hyperlocal media may be highly ambitious (Leckner et al., 2017) and researchers and journalists may have high hopes about their contributions, the outcomes fall short of the mark (Fico et al., 2013; Holt & Karlsson, 2015). The backdrop is that hyperlocal media, for example, struggle to secure long-term financial support and maintain operations, lack necessary journalistic skills, divide their time between hyperlocal journalism and a day job, and have a difficult time getting access to elite sources and institutions (Kurpius et al., 2010; Metzgar et al., 2011; Reich, 2008, van Kerkhoven & Bakker, 2014). Instead, their qualities lie in writing about things that are close to the heart and home, and being better at covering (sometimes promoting) local citizens, community groups, families, friends and colleagues in their reporting than legacy journalism (Goode, 2009; Leckner et al., 2017; Metzgar et al., 2011; Paulussen & D’heer, 2013; Turner, 2015; Williams et al., 2015). The conditions of hyperlocal media mean that expectations of them filling all of legacy journalism’s functions are unrealistic and, frankly, unjustified. Thus, using the same standard to measure the performance of legacy journalism and hyperlocal media will see hyperlocal media constantly coming up short (Ryfe & Mensing, 2010; Turner, 2015). However, despite their flaws and inadequate resources, hyperlocal media do bring something to the table – geographically anchored original news stories created by people with insights into the local conditions (Leckner et al., 2017; Metzgar et al., 2011; Williams et al., 2015). Arguably, hyperlocal media might be better suited to filling some gaps than others. This, in turn, points to the importance of charting the shape of the gap (if any) left by absent legacy media.

As is apparent in the review above, local news is important for communities in various ways. As editorial offices are shut down, questions arise about how this affects local news coverage, what gaps are left and how hyperlocal media can fill these. Against this background, the purpose of this study is to investigate, in a broad sense, the effect that the absence of an editorial office has on how a municipality is covered. This purpose is explored in the following two research questions:

The first issue pertains to the extent of news coverage. RQ1 thus asks: Does the amount of news coverage differ between municipalities that have/do not have an editorial office?

After that has been established, the study seeks to uncover how, if at all, the coverage differs by asking RQ2: How does the news performance differ between municipalities that have/do not have an editorial office?

As pointed out by Napoli and colleagues (2017), there is no developed method in journalism studies to investigate communities because journalism research is more commonly occupied with analysing “the individual media outlet or platform type, or the individual news story, rather than the community as a whole” (Napoli et al., 2017: 376). Following a similar approach to that of Napoli and colleagues (2017), we opted to use municipalities as a starting point. Municipalities are the smallest political unit in Sweden, providing a large proportion of public services and with a high level of autonomy. In order to investigate what the absence of an editorial office means for the coverage of a municipality, we took a number of steps.

The first step was to review a study carried out by Tenor (2016) that mapped the presence of at least one editorial office from legacy media organisations in all Swedish municipalities. From this study, we created a list of municipalities that did not, at the time, have any editorial office.

The second step was to use data from the Swedish Association of Local Authorities and Regions in order to acquire the number of inhabitants and what types of municipalities these were (e.g. commuting, tourism, rural, etc.).

The third step was to find municipalities of a similar size and character in order to have functionally equivalent points of comparison. We were then able to match six municipalities without editorial offices (Munkfors, Ljusnarsberg, Överkalix, Hofors, Ovanåker and Ragunda) with six municipalities that had an editorial office (Malå, Norsjö, Övertorneå, Gnosjö, Tranemo and Vindeln). These are in different parts of Sweden and are covered by various media outlets, but belong to the same type of municipality (manufacturing and rural).

The fourth step was to collect relevant news content. For that purpose, the Retriever database – the largest archive of news articles in the Nordic countries – was utilised. A sample from seven randomised days from the period 2014–2015 was created and the Retriever database was then searched for each of these days using the name of the municipality. This procedure created a sample of 606 news items, published in various outlets. It should be noted that this sample only contained legacy media news items. In total, there were 112 different outlets of which the majority were local newspapers. Regional newspapers, regional public service radio and regional public service television also appeared. In a few cases the municipalities were mentioned in national newspapers and magazines. The analysis did not separate printed from online publications and the same content that showed up as both printed and digital editions were counted as one item.

The fifth and final step was to determine content variables that are suitable for the evaluation of news performance and, ultimately, detail the possible news gap that a lack of an editorial office might create. Since much of the debate about hyperlocal media gravitates around the issue of whether they can fill the gap left by legacy media, a suitable starting point is to utilise measures frequently used in research taking stock of legacy media. For that purpose, the news items were coded according to recommendations from, in particular, Reinemann and colleagues suggested code sheet (2012) on how to measure news performance (see also Aalberg & Curran, 2012; de Vreese et al., 2017). We followed Reinemann and colleagues (2012) recommendation to study thematic, focus and style dimensions in assessing news performance. In addition, some other frequently used measurements for news performance – tonality (Messner & South, 2011), news topics (Sjøvaag et al., 2012), original reporting (Kovach & Rosenstiehl, 2001) and sourcing (Bennett, 2009) – were employed. This enables us first to establish the character of the potential gap and second to analyse how well suited hyperlocal media are to filling different facets of the potential gap. The full list of variables is shown in Tables 1 and 2.

A Cohen’s kappa intercoder reliability test (n = 71, corresponding to 13 per cent of the corpus) yielded satisfactory results as the average was .86 and the score for the different variables ranged from 0.72 to 1.

RQ1 asks if the amount of news coverage differs between municipalities that have/do not have an editorial office. At first glance, the results show that there is very little difference: the coverage of municipalities without editorial offices is 94 per cent of the other municipalities (Table 1). While collecting the data for RQ1 and comparing the amount of coverage between municipalities with and without editorial offices, we discovered that some of the news items were syndicated (e.g. identical or nearly identical to news items published in two or more different news outlets). If syndication is taken into consideration, then the impression is rather different. As Table 1 shows, non-syndicated news items (i.e. news designated for just one outlet) are more common in municipalities that have an editorial presence (p < .001). Hence, the absence of an editorial office seems to result in a dilution effect, producing news stories that are feasible in several news outlets (and possibly municipalities). Or, put differently, the absence of an editorial office in a municipality seems to result in less news aimed specifically at that municipality. Thus, the answer to RQ1, taking syndication into consideration, is that the absence of an editorial office in a municipality implies less coverage than in similar municipalities with an editorial office. There is indeed a news gap when journalists leave town. The next step is to characterise the shape of that gap.

With regard to RQ2, which concerns investigating news performance, we can see that the observation made in relation to RQ1 is confirmed by the share of original news items (e.g. news items attributed to a named journalist or media outlet) where there is a significant difference (p < .05), showing that municipalities without an editorial office are less covered by original stories.

As regards what news topics are covered, there are only significant differences in two areas: crime and community news. Municipalities without an editorial office are more likely (almost twice as likely compared with other municipalities) to be covered in crime news. Conversely, they are less likely to be covered in community news. Looking further into these differences, Table 1 shows that municipalities without an editorial office are more likely to feature decision-making authorities, have episodic coverage and have less expressive news than that in other municipalities. These three observations, together with the higher share of crime news, point to a possibility that news about municipalities without an editorial office emanates, to a greater extent, from institutional sources. We base this on the rationale that crime news is unlikely to come from anywhere other than the police (especially when there are no journalists present). Organisations issuing press releases (or pursuing similar activities) are likely to focus on themselves (e.g. decisionmaking authorities) and activities related to themselves (e.g. episodic) and, in order to get past editorial gatekeeping, try to keep their press releases neutral. In regard to all other variables in Table 1, there were no significant differences between municipalities with/without editorial offices.

In terms of the issue of sources mentioned in news items, Table 2 gives an overview.

The results in Table 2 illustrate that there are significant differences in four kinds of sources. National government elected officials, non-affiliated associations and organisations (for instance Statistics Sweden, a state agency but run by a director general rather than directly run by a political body), and media analysts are the three most common in municipalities without editorial offices. Local government official employees (i.e. not a politically appointed position) are less common as sources in municipalities without an editorial office. It should also be noted that the difference in using ordinary citizens as a source was close to significant. Interpreting the results in Table 2, it seems that sources that are not locally anchored are more common in news stories in municipalities without an editorial office. This too is, in our view, an indication of an increased role of various institutional actors in the sense that the news item might originate from press releases and not from actual interviews.

To explicitly answer RQ2, the results suggest that there are some differences between how news organisations perform, depending on whether there is an editorial office present in the municipality or not. Moreover, this difference points to a weaker news performance (e.g. fewer stories, less original reporting, more dependence on institutional sources) and the disadvantage of those municipalities where there is no editorial office.

Before moving on to the discussion, it should be noted that this study has some limitations. While the method used is sufficient to give an initial sense of the issues discussed, caution must be advised. Distance decay is treated as a binary – there either is an editorial office or not; a more fine-grained measurement of distance was not applied. The selection of cases was conditioned by which municipalities were similar in two dimensions (size and type), but differed in the presence/absence of an editorial office. Other dimensions can affect the coverage of a municipality; these have not been accounted for in this study.

The results indicate that there is a distance decay effect, that is, geographical distance matters in how municipalities are covered. Moreover, distance decay does not affect local news coverage uniformly because stories containing news properties associated with resourceful actors counter distance decay better than those news properties without. The lack of an editorial office creates a gap that affects the amount of original coverage, what areas are covered and how. Thus, the findings from this study support the suggestions of previous research on hyperlocal media (e.g. Kurpius et al., 2010; Metzgar et al., 2011; van Kerkhoven & Bakker, 2014) that the absence of legacy journalism creates a gap in coverage that hyperlocal media, potentially, can fill. Additionally, the study was able to detail some of the contours of the “content void” following the absence of an editorial office. Phrased as a theoretical proposition, the study suggests that: the absence of regularly physically present journalists in a community leads to less original news and institutional actors exercising more influence on the coverage at the expense of community news.

There are at least three inferences that can be made from this. The first is theoretical, the second concerns the role of journalism when the ties to the communities it allegedly covers are weakened and the third relates to the role that hyperlocal media can reasonably play as the gap takes shape.

First, while most research dwelling on the role of proximity of news coverage considers foreign news, which by definition takes place in distant places, one could instead ask: how far is too far before news coverage is affected? Not so far, it seems, according to the results from this study, because physical distance between journalists and the municipality seems to impact on reporting already at the local level (the distance to the nearest editorial office in the municipalities without one varies roughly between 30 km and 100 km). In current research, editorial presence is arguably a tacit assumption, such as when Harcup and O’Neill acknowledge the effect of “practical considerations, such as the availability of resources and time” (Harcup & O’Neill, 2016: 23). The findings from the present study point towards a need for geographical distance (e.g. between the event and the editorial office) to be considered a dimension of news value in its own right, because the lack of an editorial office means that one of the gatekeeping forces works differently than it would otherwise. Shoemaker and Vos (2009) posit that gatekeeping starts with the event. Yet, as evidenced by the differences in covering community events (Table 1), when the first gate (in the form of journalistic presence) is absent, it has ramifications for which events are recognised and how stories are told.

This leads to the second inference about the consequences of editorial absence for media organisations and citizens alike. The absence of editorial offices has a tangential effect on coverage, because the news media performs stronger with regard to journalistic values and citizens’ expectations when an editorial office is present than in municipalities without one. This has ramifications for citizens’ abilities to keep up with both polity and community. An absence of a steady presence of local journalists seems to lead to fewer community news stories (and more crime stories). From the perspective of helping local news survive in the future, it could be asked whether legacy news media organisations are digging their own graves. Polity and community news are what the citizens claim to expect from their local media, according to earlier research (Costera Meijer, 2010; Hess & Waller, 2016; Nielsen, 2015b; Reader & Hatcher, 2012). Cutting back on news content that is expensive to produce, yet most appreciated by the public, might save money in the short term, but might be a strategic mistake in the long term. Since the use of social media has become a regular habit for many citizens, the information needs of local communities might be met through other platforms than the legacy media. This development might be hard to reverse, as long as Facebook and Instagram maintain their dominant role (and remain free of charge).

The gap seems to be largest in community news, which points to the third inference that can be made from the results: specifically, the role of hyperlocal media in a news ecology where legacy news media have less presence. As pointed out in the literature review, hyperlocal journalism faces some challenges due to a lack of resources, training, skills and interest, which impacts negatively on participants’ abilities to pursue news reporting that can meet the standards of mainstream journalism. Instead, hyperlocal media are better at keeping up with local affairs and raising concerns about issues that are close to them geographically, socially and emotionally, living up to their role of providing “geographically based, community-oriented, original-news-reporting” (Metzgar et al., 2011: 774). Thus, in view of this study’s results, it seems that the decline of community news in legacy journalism can be at least partially countered by actors involved in hyperlocal journalism. However, the increased presence/importance of institutional actors in the absence of journalists is more difficult to counteract. Another challenge facing hyperlocal news in the absence of editorial offices is to unearth community news itself, because local papers are the customary keystone media in the news ecology.

Our findings suggest that the gap the editorial office leaves is filled with syndicated news and news stemming from institutional actors. This can be interpreted as a sign of news organisations trying to produce the same amount of coverage with fewer resources. The need for more research is pressing, especially if the closing of editorial offices continues to increase syndication practices and reliance on information from institutional actors. If the results can be reproduced, how is a municipality (or others) affected when it is less viewed as a community and more as a locality from which crime reports emanate, represented by sources far away from its location? Here it might also be worth making comparisons with studies on minorities as there are similarities in the patterns of coverage ‒ low coverage, negative coverage and the voices of those covered not being heard enough (e.g. Bleich et al., 2015, Pietikäinen, 2003). Factoring in the role of an editorial office might also give valuable insights into the connections between shrinking local media and growing mistrust between the peripheries or rural areas and the metropolitan urban elites. This includes the circumstances in which community news is increasingly left to be covered by hyperlocal media on their (rural) publishing platforms, while being downgraded by legacy media on their (urban) publishing platforms. Ultimately, this is a question of understanding the functions and effects of distance decay in a world of shrinking local media and the consequences for local communities.