Canadian Journal of Communication, Vol 33 (2008) 193-211
©2008 Canadian Journal of Communication Corporation

Mapping the Geography of Online News

Mike Gasher & Reisa Klein
Concordia University

Mike Gasher is Associate Professor and Director of the Department of Journalism at Concordia University, 7141 Sherbrooke Street W., Montréal, QC  H4B 1R6. Email: . Reisa Klein is a recent graduate of the MA in Media Studies program at Concordia University. Email: .

Abstract: This article is a product of the Geography of News Project, whose express purpose is to assess the Internet’s potential to foster an expanded, and more international, news geography. The article begins by discussing the relationship between news and geography, then describes and analyzes the results of a news-flow study of the websites of three international newspapers with extensive online editions: The Times of London, Libération of Paris, and Ha’aretz of Tel Aviv. It argues that while each of these newspapers adopts a distinct online strategy, the geographies they map out remain highly circumscribed, reinforcing significant distinctions in the newsworthiness of various parts of the world and suggesting that our understanding of news value must include the value of audiences for whom news is produced.

Keywords: Online journalism

Résumé : Cet article est le résultat du Projet de cartographie de la répartition géographique des nouvelles en ligne, dont la raison d’être est d’évaluer le potentiel d’Internet à assurer une couverture géographique plus internationale de l’actualité. Cet article commence par discuter du rapport entre nouvelles et géographie, puis décrit et analyse les résultats d’une étude de la diffusion des nouvelles sur les sites de trois quotidiens comportant des éditions en ligne d’envergure : le Times de Londres, Libération de Paris et Haaretz de Tel Aviv. L’article soutient que, bien que chacun de ces journaux adopte une stratégie distincte en ligne, les géographies qu’elles développent demeurent fort restreintes, renforçant des écarts significatifs dans l’importance journalistique de diverses régions du monde et suggérant que notre compréhension de la valeur d’une nouvelle dépend souvent de la valeur du public pour lequel on couvre l’actualité.

Mots clés : Journalisme en ligne

The Internet has been for at least a decade a source of excited speculation about the potential for transforming, even “revolutionizing,” journalism. This international network of computer networks has been posited as a medium that can be used to improve and facilitate news gathering, accelerate and expand news delivery, undermine corporate concentration, and allow greater citizen participation in news production. The time has come, however, for scholarly research to test the Internet’s vaunted technological potential against the evidence provided by online journalism’s application in particular economic, social, cultural, and political environments. For although the Internet reinvigorates the ideal of perfect communication, John Durham Peters reminds us that “communication as bridge means an abyss is somewhere near” (2000, p. 16).

This article is a product of the Geography of News Project, whose express purpose is to assess one specific aspect of the Internet’s potential: its promise to foster a new communications geography on a global scale and thereby serve as an agent of globalization (see Carey, 1998; Castells, 2001; Mattelart, 2000;).1 The project is committed to producing a series of intensive news-flow studies that will provide a precise measure of the territory covered by the online sites of daily newspapers. These studies are designed to reveal the extent to which the newspapers are taking advantage of the reach of the Internet’s architecture and new-found access to news audiences beyond their traditional coverage areas and established markets. Simply put, if a newspaper published online is “delivered” everywhere in the wired world, and is thus available to an expanded audience, will that paper’s news coverage expand accordingly and include more of its region, its country, the world? If so, which areas—which people, institutions, activities—will be included and which excluded in its new news geography?

The article begins by discussing the relationship between news and geography and considers the factors that govern the relationship between any one news organization and its coverage area. It then describes the project’s application of news-flow methodology to news sites on the World Wide Web. It subsequently presents the results of a 2004 news-flow study of three international newspapers with extensive online editions—The Times of London, Libération of Paris, and Ha’aretz of Tel Aviv—each of which, we argue, is well-positioned to attract international news audiences. It concludes with a discussion of how these three online dailies occupy the Internet’s geography and how this case study speaks to the larger question of online news geography.

Mapping the news

Every news organization draws its own map and subsequently operates, as both a business and a news medium, within clearly demarcated geographical bounds. This map is the product of a particular kind of triangulation, in which the geography the news organization occupies is a conceptual space delineated by three points: the news coverage it provides, the target audience to which it seeks to deliver news (and its corresponding delivery capacity), and the advertisers who buy access to that audience. As media economist Robert G. Picard explains it: “Media units operate in specific geographic markets and are inextricably linked to those markets by the product content and advertising services they provide within those markets” (1989, p. 19). Journalists enter the picture through their decisions about who comprises their audience and what that community considers newsworthy, whether that “community” occupies a neighbourhood, a city, or a nation. News coverage, therefore, shapes and is shaped by the territory the news organization stakes out.

But what happens when the geographic parameters are radically altered? For instance, what happens when a newspaper decides to establish a World Wide Web site and is no longer confined to the physical boundaries of its hard-copy circulation area? What happens when this online newspaper suddenly has the technological capacity to expand its geography in accord with the global architecture of the Internet?

If James Carey is correct in asserting that the Internet forges a “new media ecology,” and the global organization of media is displacing the national, then this “first instance of a global communication system” holds out great promise for the broader circulation of news on an international scale. Carey argues that “the Internet is at the center of the integration of a new media ecology which transforms the structural relations among older media such [as] print and broadcast and integrates them to a new center around the defining technologies of computer and satellite” (1998, p. 28). Manuel Castells echoes these remarks: “The Internet is a communication medium that allows, for the first time, the communication of many to many, in chosen time, on a global scale” (2001, p. 2).

Technologically, the Internet expands the audience available to news organizations, but serving that audience with editorial and advertising content is much more complicated. If the Internet provides the “backbone” for this new geography, Carey (1998) acknowledges it will also require gradual and concurrent changes in physical, symbolic, and media ecologies, in the ways people live and imagine space. Stig Hjarvard states: “With the Internet, the problem of global reach becomes a question of language and content” (2001, p. 29, emphasis original).2

Although the Internet holds great technological potential to shift the points of the triangle described above, we must acknowledge that the circulation of news is not simply a technological question. If the Internet’s architecture, the flows of people, capital, goods, and services we associate with globalization, and the economic imperative for commercial news organizations to expand their markets all serve as centrifugal forces, they contend with the centripetal forces of community, language, conventional news values, the elements which constitute the digital divide, and the commercial imperative of selling news as a commodity (see Carey, 1998; Gasher, 2003).

The problem of uneven international information flows has been a source of friction between information haves and have-nots since the end of the Second World War (see Gerbner, Mowlana, & Nordenstrang, 1993; McPhail, 1987; Mowlana, 1997; Sreberny-Mohammadi, 1984). News-flow studies have consistently shown that news organizations are much more interested in some parts of the world than in others, resulting in what H. Denis Wu describes as a “discrepancy between the ‘real world’ and the ‘news world’” (2000, p. 110). When UNESCO sparked debate over a New World Information and Communication Order in the 1970s, it noted that almost 80 nations had acceded to independence since 1950, but imbalanced international flows of information and communication resulted in a persistent intellectual and cultural dependency whose effects “are as serious as those of political subjection or economic dependence” (ICSCP, 1980, pp. 34-35). This meant, in part, that developing countries were denied the kind of freedom of information advocated by the Universal Declaration of Human Rights. In practical terms, it also meant that information about serious problems such as arms proliferation, famine, poverty, illiteracy, racism, unemployment, environmental degradation, and disease was not reaching news audiences, including the people, institutions, and governments who might be called upon to address these issues.

A number of observers, from journalists (e.g., Manthorpe, 1998; Vasil, 2003) to scholars (e.g., Biltereyst, 2001; Hjarvard, 2001; Thussu, 2004), have noted the paradox between news organizations’ increasing ability to gather and distribute international news and their decreasing coverage of foreign affairs. Newspapers and television networks have closed foreign bureaus and reduced the space and time they give to international stories. Not only is coverage shrinking, but Daniel Biltereyst argues that “foreign items tend to be marginalized and, as much as possible, domesticated, personalized and ‘made relevant’” (2001, p. 48). This shrinking and partial news coverage of the world is especially problematic today because, as Castells argues, the media have become the central arena where politics takes place. “Outside the media sphere there is only political marginality” (1999b, p. 312).

Constructing the news space

News spaces are produced. If networks such as the Internet are open structures and thus able to expand without limits (see Castells, 1999a, p. 470), the online space any one news organization maps out is governed by both centrifugal forces encouraging expansion and centripetal forces inhibiting growth, bounding any one news organization’s news map in particular ways. Journalism, after Henri Lefebvre (1991), is a spatial practice; it creates a representational space—Wu’s “news world”—based on what journalists determine to be the lived and imagined space of members of their audience community. Lefebvre describes space as “social morphology” (1991, pp. 93-94).

The most obvious centrifugal force in the shaping of the online news space is the Internet’s global architecture and the ease of access to this network that news organizations enjoy. A second, extra-media force is globalization, understood here in the broad sense of intensified international flows of people, capital, goods, services, and symbols. As populations diversify due to immigration, as tourism makes us more and more familiar with distant parts of the word, as daily commerce connects us to goods and services produced abroad, and as our workplaces are increasingly linked to international networks, clear distinctions between local and foreign erode.

A third force is economic growth. Once a news organization has undertaken the costs of establishing a website and updating it regularly, it is in that organization’s economic interest to maximize its audience. The Internet gives news organizations access to online audiences well beyond their local markets, provided they offer both editorial and advertising content of interest to far-flung audiences (see Gasher, 2003; Hamilton, 2004; Picard, 1989).

At the same time, however, a number of factors inhibit online news expansion. Critiquing what he calls the “death of distance” argument, Vincent Mosco writes: “It is an increasingly popular myth that computer communication ends geography by completing a revolution in the process of transcending the spatial constraints that historically limited the movement of information” (2004, p. 85). Castells concedes that the Internet “redefines distance but does not cancel geography” (2001, p. 207). Not everyone, for example, benefits equally, or at all, from the intensified international contacts of globalization, because inherent within globalization are hierarchies of power in which some countries and some people enjoy far more mobility than others. People, capital, goods, services, and symbols do not flow evenly, but are tied to, and shaped by, relations of political, economic, social, even military, power (see Massey, 1991; Sassen, 1998).

In terms of online journalism specifically, this raises questions about conventional news values (what is newsworthy? to whom?), community affiliation, people’s access to computer networks, news production resources, and the commercial imperatives of the news industry.

Herbert Gans (2004), Benedict Anderson (1989), and John Hartley (1992) have all underlined the point that media—and the news media in particular—are implicated in the construction of community. Communities, like media audiences, do not simply precede the media through which they are represented, but media, in fact, play a part in constituting these communities. One function of journalism, Gans argues, is to bring the society and the nation into being by framing news events in a national context. Journalists, that is, “help impose unity on what is otherwise a congeries of individuals and groups acting inside a set of geographical and political boundaries” (2004, p. 298).

Anderson described the novel and the newspaper of the eighteenth century as new forms of imagining, which provided the technical means to produce in people a sense of “nation-ness,” an “imagined political community” (1989, p. 15). If novels created a “sociological landscape” (1989, pp. 35-36), Anderson described newspaper readership as a mass ceremony which “made it possible for rapidly growing numbers of people to think about themselves, and to relate themselves to others, in profoundly new ways” (1989, p. 40). New communities on new scales became imaginable.

Gans and Anderson emphasized the bonds of community these media enabled, or their inclusiveness. Hartley focuses on the divisions they forge, or their exclusiveness. Newspapers, he argues, make certain kinds of “we claims,” based on very complicated understandings of proximity. He argues that the news media create domains of inclusion and exclusion, which he calls, respectively, “Wedom” and “Theydom.” These domains do not simply coincide with formal geographical and political boundaries, but are instead produced in the relationship between a news organization and the audience it interpellates. Hartley argues that “news is organized around strategies of inclusion and exclusion from ‘our’ community.” News, that is, “includes stories on a daily basis which enable everyone to recognize a larger unity or community than their own immediate contacts, and to identify with the news outlet as ‘our’ storyteller.” To alter the geography of news, then, is to revise an intimate relationship between the newspaper as “our” storyteller and its audience (1992, p. 207). Noting news organizations’ shrinking interest in international news, Biltereyst writes: “In a competitive media landscape with commercially-driven mass communication, many foreign news items are seen as too complex, while proximity to the audience grows as a news value” (2001, pp. 41-42).

If globalization is altering notions of community, challenging conventional notions of Wedom and Theydom, this process takes time. Guo-Ming Chen and William Starosta remind us that such group affiliations are, in part, negotiated via communication networks, and to forge strong affiliations requires the development of new “communication realities.” The “virtual environment of the media,” they write, is indeed creating a “public sphere” where global issues are being discussed and a new communication reality is being produced (2000, pp. 4-5).

The digital divide, of course, draws its own boundary lines between the Wedom of the wired and the Theydom of the wire-less, further configuring cyberspace. A number of critics (e.g., Mattelart, 2000; Mosco, 2004; van Dijk, 2005) have noted the demographic distinctions (based primarily on age, gender, race, language, and income and education levels) between those who have ready Internet access and those who do not. This most obviously circumscribes the potential audience for online news, but it also speaks to the question of what kinds of news will attract audiences and thus have commercial viability.

Most of the news organizations that have online potential, after all, are private commercial producers who sell news as a commodity and package news audiences for sale to advertisers. Such producers tend to be spatially concentrated, clustering in Western hub cities with advanced services in the media, entertainment, education, finance, health, and technology sectors (see Castells, 2001). Castells describes the “geography of networks” as a series of central hubs and subservient nodes, placing the major news organizations at those hubs and their “news teams” around the world at various information nodes (1999a; 2001).3 Because of its development within a capitalist economy, Castells argues that the Internet accentuates uneven development patterns. This means that some places, such as most of the African continent, have a “structural irrelevance” to the information network (1999a, pp. 135-136). He writes:

Under the new, dominant logic of the space of flows, areas that are non-valuable from the perspective of informational capitalism, and that do not have significant political interest for the powers that be, are bypassed by flows of wealth and information, and ultimately deprived of the basic technological infrastructure that allows us to communicate, innovate, produce, consume, and even live, in today’s world. (Castells, 1999c, p. 74)

Elsewhere, Castells concludes: “The geography of networks is a geography of both inclusion and exclusion, depending on the value attached by socially dominant interests to any given place” (2001, p. 238).

As a result of these contending push and pull forces, there is considerable uncertainty about whether the Internet will ameliorate or exacerbate what have historically been uneven international news flows, whether the lines between inclusion and exclusion will be eroded or reinforced, and whether certain kinds of news audiences will expand at the same time as others shrink or remain ignored.

Drawing the map

One way of determining whether or not daily newspapers’ online editions have taken advantage of the Internet’s geographical reach is to return to our first point of triangulation: to study the package of news content offered to attract audiences. More precisely, we wanted to see whether anything in that content package suggests the newspaper is remapping its target audience, reaching beyond its conventional audience borders. And, if so, in which particular directions? News-flow studies have provided a particularly effective method of measuring the quantity, the origin, and the type of news reported by specified news media, and they are particularly conducive to evaluating the geographic space occupied by news coverage.4

News-flow research is a form of content analysis that seeks to track the origin, content, and destination of news stories in selected media, to determine patterns in the international circulation of news items and document asymmetries in the ways different countries occupy the international-news agenda. News-flow research has been employed since the 1950s to analyze newspapers, wire services, magazines, and television newscasts; here it is applied to news sites on the World Wide Web. We have chosen to study daily newspapers, specifically, because of the number and variety of them available on the Web, because their news packages tend to be more extensive than those of either magazines or radio and TV broadcasters, and, most importantly, because by moving online, newspapers break the physical and geographical restraints inherent to the production and distribution of heavy, bulky hard-copy newspapers, gaining for themselves a new, global mobility.

The application of news-flow research to online news sites presents two challenges. The first is technical: how to capture the site for analysis, given the sheer size of some sites and given the continual updating of some forms of content on the site (e.g., breaking stories fed by wire services). The second is methodological: once captured, what is the coding protocol?

Conceding that the continual updating of some news items makes it difficult, if not impossible, to account for every single news item that appears on a given site on a given day, we chose instead to take a “snapshot” of our study sites at a specified uniform time on coding days. That is, beginning at our specified time (11:00 a.m. local time), we downloaded the complete website, then saved it to CD-ROM for coding. This is consistent with conventional news-flow studies that have been applied to particular editions of magazines and daily newspapers and particular TV newscasts. In other words, our snapshot of the website is analogous to a specific edition of a hard-copy newspaper (in that most newspapers publish several daily editions, which differ from earlier or subsequent editions) or a specific TV newscast (noon or 6:00 p.m. or 11:00 p.m.). Because we downloaded a composite week for each of the three websites (i.e., seven editions), and because we coded every section of the site (i.e., news, sports, business, etc.), we are confident that the large sample size—more than 3,500 items in this study—permits a fair assessment of these news packages.

The coding protocol is the product of our experiences with a 2001 pilot study of the Montreal Gazette (Gasher & Gabriele, 2004), a 2003 study of three Canadian national newspapers (Gasher, 2007), and two months of pre-testing and trial intercoder reliability tests during the spring of 2004.5 For each story, the protocol required the coder to record the publication name; date; headline; source agency (e.g., Reuters); filing origin (where the story was filed from); subject origin (whether the same as the filing origin, different, or the same and different); countries cited (in the body of the story); word count; number of accompanying illustrations; and topic (based on a chart that assigned the sites’ own topic headings to the conventional sectioning of stories into news, sports, business, arts and entertainment, and lifestyle).

While most news-flow studies have confined their sampling to “news” items or stories contained only in the front section of newspapers, we chose to be as inclusive as possible, to avoid discriminating between different kinds of news stories and to allow us to detect news sites’ emphasis on particular topics, such as business or sports. Excluded were statistical summaries, event listings, editorial cartoons, and all advertising material.

Stories for this study were coded by two student research assistants and subjected to an ongoing intercoder reliability test.6 Once coded, data were entered into an SPSS (Statistical Package for the Social Sciences) database for subsequent analysis. For each site, the analysis was concerned with the following: where in the world its stories originated; which foreign origins predominated and which were largely excluded; which news agents tended to provide stories (wire services, staff, etc.); what topics foreign stories tended to address; and the relative prominence of stories from different origins (as measured by a combination of word count and number of illustrations).7

As one in a series of news-flow studies,8 we selected the sites of three brand-name newspapers—The Times, Libération, and Ha’aretz—from three countries that both generate a lot of news and attract considerable attention around the world: Great Britain, France, and Israel. The premise behind this selection is that if newspaper websites have the potential to expand their audiences through cyberspace, well-known, quality newspapers from countries that are recognizable news centres will have the greatest potential. These newspapers were chosen because they are well-known to international readers, they have stand-alone websites, and their articles provided the information (e.g., source and filing origin) our coding protocol demanded.9

We began downloading the sites on the random date of Wednesday, May 26, 2004. We then downloaded one edition of each site every eight days until we had a composite week.10 In sum, we coded 3,647 articles from the three sites.

The news world

At first glance, and based strictly on placelines, the news content on all three sites appeared to be predominantly local, in that the clear majority of stories on each site was filed from the home country. The Times site was by far the largest, with 2,164 articles over seven days (compared to 778 in Libération and 705 in Ha’aretz). Although The Times included articles filed from 40 places around the world, 92.7% of its articles carried U.K. placelines (see Table 1). The second most popular origin was the United States, accounting for just 1.1% of placelines (24 articles). Ha’aretz drew articles from just 26 places, and filings from Israel accounted for 89.4% of its articles. The United States, again, was the second most popular origin, with 3.7% of placelines (26 articles). Libération featured articles from 53 places (many more than The Times, twice as many as Ha’aretz) and just over two-thirds of its articles (68.4%) were filed from France. While the United States was the second most popular origin (44 articles, accounting for 5.6% of the total), Libération also featured 22 stories from Portugal, 20 from the United Kingdom, 15 from Germany, and 14 from Japan.

Overall, the three newspapers carried articles from 67 places in the world. However, less than 1% of their articles came from the continents of Africa (21 stories from 12 countries, 0.6% of the total) and South America (8 stories from 4 countries, 0.2% of the total) combined. Even large, politically powerful countries like Russia (13 articles) and China (8 articles) generated relatively few placelines.

Table 1

A different—and slightly less parochial—picture emerges, however, if we concede that placelines are not a reliable indicator of the subject of a story. A news story filed from Paris, for example, may not concern France at all. A news story filed from Berlin may concern Germany as well as France, Italy, and Belgium. Thus, we also coded stories for the “countries cited” in their texts to gain a more reliable account of which places in the world received news coverage.11 As a general rule, we found that placelines overstated coverage of the newspaper’s home country and understated foreign coverage (see Table 2).

In spite of its 2,006 placelines, for example, the United Kingdom was actually cited in Times articles 1,399 times, meaning that slightly less than two-thirds (64.6%) of Times articles concerned the newspaper’s home country in some way. A sampling of other countries revealed that they were cited in Times articles much more often than they produced placelines. The United States, for example, which had just 24 placelines, was actually cited in 468 Times articles, more than one-fifth (21.6%) of the total. Perhaps the best example is Iraq, which had become by 2004 one of the most dangerous places in the world for journalists. The Times carried 12 stories filed from Iraq, but Iraq was cited in 131 stories, an average of more than 18 stories per edition. South Africa had 4 placelines in The Times, but was cited in 59 articles. Brazil had no placelines, but was mentioned in 23 articles.

Table 2

Compared to its 630 Israel placelines, Ha’aretz articles cited Israel 597 times, meaning 84.7% of its articles talked about Israel in some way. The United States, with 26 placelines, was cited in more than one-quarter of Ha’aretz articles (183 times, or 26%). Ha’aretz carried just 6 stories filed from Iraq, but cited Iraq in 51 articles. Similarly, France was cited 449 times in Libération articles (57.7%), compared to its 532 placelines. The United States, with 44 placelines, was mentioned in more than one-quarter of Libération’s articles (196 times, or 25.2%). Again, Libération ran 10 stories with Iraq placelines, but 61 of its stories cited Iraq, an average of more than 8 stories per edition. Brazil, with 2 placelines, was cited 18 times, and South Africa, with 2 placelines, was mentioned 13 times.

Overall, the three sites mentioned 180 places in their coverage. Compared to their 12 African placelines, the three sites carried 388 stories that cited Africa, mentioning 38 different African countries at least once. That is, one in 10 stories (10.6%) mentioned an African country, at least in passing. Compared to their 8 South American placelines, the three newspapers cited 12 of the 13 South American countries in 131 stories (3.6% of the total), excluding only Suriname.

News was the principal topic in all three newspapers (43.2% of articles in The Times, 62.7% in Ha’aretz, and 45.8% in Libération), but each newspaper had a different second topic (see Table 3). For The Times it was sports (18.1% of articles), for Ha’aretz it was arts and entertainment (13.8%), and for Libération it was business (18%). In the three newspapers combined, news accounted for almost half of all articles (47.5%), followed by business (14.4%), sports (13.6%), arts and entertainment (11.9%), lifestyle (3.1%), and “other” (9.4%).12

News items in The Times came from 37 places, even if 87.7% originated in the U.K. While sports articles in The Times came from 12 places, still 93.6% of sports stories were filed from the U.K. All but one of The Times arts and entertainment stories (203 of 204) had U.K. placelines. News items in Ha’aretz came from 23 places, even if 85.1% of all news items were filed from Israel. All but one of the business (88 of 89) and all but two of the arts and entertainment (95 of 97) articles in Ha’aretz were filed from Israel. True to form, Libération carried news items from 40 places, with just over two-thirds (68.3%) featuring French placelines. Slightly less than two-thirds (60.7%) of the business items in Libération were filed from France.

table 3

A closer look at these figures indicates that business and sports often rival news as topics of foreign coverage. If, for example, news was the predominant topic of Times articles both filed from the United States (14 of 24 placelines) and citing the United States (232 of 468 citations), sports articles predominated in coverage of South Africa (31 of 59 citations, compared to 13 news articles) and Brazil (13 of 23 citations, compared to just one news story). In Ha’aretz, 2 of the 6 articles that cited South Africa were business stories, and 3 of the 6 from Brazil concerned business. In Libération, business (6 articles) and sports (5 articles) outnumbered news (3 articles) from Brazil, and business articles (16) outnumbered news items (12) from China. Because France was participating in the Euro 2004 soccer tournament in Portugal, 21 of the 22 Libération articles filed from Portugal concerned sports.

Both The Times and Ha’aretz relied heavily on staff reporting to fill their content needs (see Table 4). Staff reporters supplied 81.2% of articles in The Times, with exclusive use of wire services reduced to 1.9% of stories. In Ha’aretz, staff accounted for 72.4% of articles, with wire services providing 8.1%. In Libération, just over half (51.4%) of articles were exclusively staff produced, with the wire services accounting for 30.7%. Libération demonstrated a particularly heavy reliance on Reuters, which supplied 29.6% of all stories on the site.

Libération was especially dependent upon Reuters for international news; while Libération staffers produced stories from 25 places, almost half of the foreign placelines (47.6%) carried Reuters bylines. Times staffers produced articles from 31 places, but by far most staff stories in The Times (94%) carried U.K. placelines. Ha’aretz was the most parochial, with staff filings from 8 places and 97.8% of staff articles filed from Israel.

Table 4

Finally, we coded for word count and number of illustrations to capture some sense of the relative prominence of articles citing different countries. While Times articles overall tended to be short (one-half of the total contained fewer than 400 words), Times articles in which the United States was cited tended to be long: 223 of the 468 articles (47.6%) citing the United States contained at least 600 words. Similarly, almost half of the Times articles in which South Africa (27 of 59) and Brazil (11 of 23) were cited contained at least 800 words, granting those places somewhat more prominent coverage than their sheer number of citations would suggest. Ha’aretz generally featured longer articles than The Times, with half of its articles containing at least 600 words. More than half of the stories that cited the United States (51.9%), the United Kingdom (63.1%), France (60%), Russia (55.2%), and China (58.8%) had more than 800 words. Ha’aretz articles citing Brazil, on the other hand, tended to be short (3 of the 6 contained fewer than 400 words).

Who, what, where?

These results are based on a study of a one-week sample of just three newspaper sites on the World Wide Web. They do, nonetheless, begin to sketch out the online strategies these newspapers have adopted, they reveal some important distinctions between them, and they point to some more general factors governing online news geography.

With respect to the geographic origins of news stories, for example, the results point, first, to the distinct possibility of producing a newspaper that aspires to be international in scope. Libération appeared to be a far more extroverted news site than either The Times or Ha’aretz, by providing stories from 53 places around the world over a seven-day sample—even if the Paris-based newspaper relied on wire service material for most of those foreign dispatches and the paper’s coverage remained highly circumscribed.

At the same time, the study points strongly to the need to look beyond placelines in determining the extent of foreign news coverage these sites provide. A placeline may tell us where the reporter is stationed, but it is not a reliable indicator of the subject of that reporter’s story. Here, the “countries cited” variable is much more meaningful as an indicator of how these newspapers cover the world. If, as mentioned above, the “filing origin” variable overstated the number of articles that concerned the home country and understated international-news coverage, the “countries cited” variable nonetheless revealed a picture of parochialism and significant disparity between regions of the world in international coverage. After the home country and the United States on each site, citations drop off precipitously. The news world depicted on these sites strongly emphasized places with economic and political power; after the home countries, the top 10 places cited overall were the United States, Germany, Iraq, the European Union, Italy, Ireland, Australia, Spain, the United Nations, and Russia.

This is not to suggest that placeline or filing origin is irrelevant. The predominance of domestic placelines may very well speak to the persistence of physical proximity as a news value. It may also speak to news organizations’ tendency to domesticate or make relevant foreign news items (see Biltereyst, 2001), and may thereby speak to the question of the sites’ target audiences. This is where qualitative textual analysis can complement our quantitative analysis. To what extent are foreign stories given a local angle or domesticated? How are such stories presented to audiences as newsworthy?13

Closely related to this point is the “source” variable. All three sites featured a predominance of staff-produced articles, ranging from just over half in Libération to more than 80% in The Times. If Ha’aretz and Libération relied on the international wire services for most of their foreign stories, Times staff accounted for better than two-thirds (67.3%) of its foreign dispatches. This is a question of journalistic voice, of the extent to which a newspaper wishes to put its own particular stamp on its news coverage, to assert its own voice and style for an audience accustomed to that voice. It may also be a question of audience address. If it can be argued that staff stories are tailored for a particular and identifiable news audience, wire stories are intended for far more universal reception, and may thus be more digestible or comprehensible to more audiences. Normally, sites with a high percentage of staff content are praised for their originality and sites with a predominance of wire copy are criticized for serving simply as repeater stations. Here, however, wire copy, because of the accessible style in which it is written, may signal a news site’s intention to be more extroverted, more inclusive of non-national readers. This may suggest, therefore, that The Times, with its high degree of staff-produced content, is targeting a domestic audience and that Libération, with its high degree of wire content, is less confined to a local readership.

One clue to the actual constitution of the sites’ audience is the origin of letters to the editor. Both The Times—230 letters from readers in six countries, 214 of them (over 93%) from the U.K.—and Libération—12 letters, all from French readers—would seem to have predominantly domestic readerships. Ha’aretz, on the other hand, attracted 37 letters from seven countries, including 11 from the United States, suggesting a much more international audience. Because its home turf and its coverage specialty is the Middle East, Ha’aretz would seem to be well positioned to attract international readers precisely because of its concentration on local news, pulling readers into its site rather than pushing a wide-ranging international news package onto far-flung audiences.

Topic, of course, is another key element in the content package newspapers provide to audiences, and there is nothing to prevent a newspaper from abandoning the something-for-everyone approach of its hard-copy edition and specializing in a particular topic area online. For example, a previous study that compared the hard-copy and online editions of the Montreal Gazette (Gasher & Gabriele, 2004) found that while sports articles accounted for less than 20% of content in the Gazette’s hard-copy edition, more than half of its online edition (50.8%) comprised sports articles from around the world. In a study of three Canadian national newspapers (Gasher, 2007), sports was the clear No. 2 topic to news among international filings. In a study of newspaper, television, and radio in 10 countries, Shoemaker & Cohen (2006) also noted a predominance of sports news.

This suggests that some topics—certain kinds of news such as natural disasters, sports, business, and arts and entertainment stories—have more international mobility than others and thus more potential to generate audiences that are not border-bound. Sports fans in North America may be interested in soccer or motor racing, regardless of where events are taking place. Similarly, Hollywood film and television fans can be found all over the world. A closer look at the stories within these topic areas could prove fruitful in this regard. Similarly, it can be argued that some topic areas are gendered and, again, a closer textual analysis of stories within particular topic areas—e.g., sports, business—could be revealing in terms of their intended audiences.

The results here suggest a strong emphasis that all three sites emphasize news stories. News was the predominant topic for all three newspapers, and no secondary topic rivalled this emphasis. The only exception to this was, as mentioned above, an occasional prevalence for secondary topics from specific countries: sports articles from South Africa and Brazil in The Times, business articles from China in Libération. This would, however, suggest event-specific news judgments rather than any larger tendency.

A highly circumscribed news world

The findings here suggest strongly that the geography these three newspaper websites are mapping out, if it is to extend beyond conventional boundaries at all, will remain highly circumscribed. The results also suggest that these newspapers will map their coverage areas and designate their audiences each in their own way; the formula for Ha’aretz will not be the same as the formula for The Times.

What will be much more interesting to determine, as we conduct further studies and as we complement the quantitative news-flow studies with qualitative textual analyses, is precisely how these newspapers are establishing the contours of their online geography, and based on what criteria.

The question of news value, for example—the question of what makes a given news event of interest to a particular audience—will remain central. If, on a macro scale, globalization theorists can comfortably assert a tendency toward a more integrated world, how does that inform the more intimate, micro-scale relationship between a reader and his or her newspaper? If one of those tried and true news values is proximity—that is, geographical, cultural or emotional closeness—what will make readers feel close to far-away stories? Clearly, there are a number of possibilities. The Times approach seems to emphasize staff-generated content for a conventional, British audience. Ha’aretz seems to be exploiting its location and expertise to attract audiences already interested in Middle East politics.

Equally central is the question of community, and the myriad ways in which people constitute and imagine community in our time; we recognize not merely communities based on shared locale and citizenship, but those based on shared interests, language, religion, ethnicity, race, and sexual orientation. If, customarily, newspapers speak to and for a particular situated community, they face, again, a number of choices online. They could, simply, reinforce that conventional community relationship, as The Times seems to be doing. They could specialize in a particular topic area, as Ha’aretz seems to be doing with its Middle East coverage, and other papers could do by focusing on business news or sports news. They could, alternatively, try to be all things to all people everywhere, as Libération hints at attempting.

Historically, newspapers have adopted unique means by which to align the three points of the triangle described at the outset of this paper: the content package, the advertising package, and the target audience. Adjusting one of these points necessitates adjusting all three. This is to a great extent an economic relationship. As long as the public service of journalism is in the private hands of profit-seeking commercial news organizations, these organizations need to be able to develop a viable online formula for selling audiences to advertisers.

As a commercial product, not all news is of equal value. The same may be said for news audiences. If audience members for online news abandon their common ties to place—i.e., any one political community, whether city or nation—we may see news producers seek to maximize revenues by peddling news packages that generate audiences with the particular demographic characteristics most attractive to advertisers (e.g., age, income, education, place of residence). Focusing on business or sports news, for example, could allow them to reach targeted demographics. This would add a whole new dimension to the notion of “news value,” reinforcing the newsworthiness of certain kinds of news items and the commercial distinctions between news audiences—or audience markets. This hypothesis remains to be tested by subsequent research.

But the relationship between news organizations and their audiences is also cultural. If we accept Hartley’s argument that newspapers work as communications vehicles by making “we” claims and serving as storytellers to their community, the question becomes: who constitutes that community? News sites on the World Wide Web are in the process of redefining and reconstructing their readership base, and its characteristics—geographical, demographic—may remain in flux for some time yet. Still, Hartley’s argument underlines a critical point; newspapers don’t serve “the public” in any inclusive sense, but instead they serve specific readership segments, providing news (and advertising) that speaks to those readers. The Internet may allow newspapers to expand their readerships online, but it does not erase the boundaries—cultural, political, economic, historical—that divide people, nor will it put an end to the exclusive forms of address newspapers adopt in constituting specific audiences.

Research to date has made clear that news flows are determined less by ideals of equity, fairness, and balance in international-news exchanges than by the political, economic, and military power of the nations who make news. We suggest that another component of international-news value that needs to be considered is the commercial value of the audiences who consume news, the audiences for whom international news is produced. Definitions of news value, in other words, cannot be reduced to a series of strictly professional criteria of news judgment, or reduced to the inherent properties of the news event itself. Instead, we recommend expanding the notion of news value to include the audience in terms of the criteria by which journalists distinguish between members of Wedom and Theydom, and the commercial value of particular groups of news consumers, wherever they may be. We believe that only when news organizations find their economic formulae and constitute their audiences will we able to draw the precise contours of the geography of online news.


1. This project, based in the Department of Journalism at Concordia University in Montréal, has received funding from Concordia University (2000-2003), le Fonds québécois de recherche sur la société et la culture (2002-2005) and the Social Sciences and Humanities Research Council of Canada (2004-2007).

2. At least one major news organization has attempted such an expansion online and demonstrated its feasibility. The commercial-free BBC World Service (, which was already sending short-wave radio broadcasts around the world, has used its brand recognition to seek diasporic audiences online by developing websites in 43 languages. Its sites report visits from 89 countries in a normal news day (Westcott & Mukherjee, 2004). The BBC World Service tailors its news content to these diverse audiences by having different crews produce original stories for its sites.

3. Castells (2001) has his own triangulation of Internet geography, consisting of its “technical geography” or infrastructure, the spatial distribution of its users, and the “economic geography” of online content producers (p. 208).

4. News-flow studies, of course, address only one part of the content picture, and provide simply a starting point for this project. In future stages of the project, we propose to study news content through textual analysis of specified international stories to determine who they are written for; to conduct interviews with editors of online newspapers to determine their editorial goals; and to assess online advertising content to determine its geographical parameters (i.e., local, regional, national, international) and consider its target market.

5. The authors are particularly indebted to Linda Charbonneau, Michael Craig, and Andreea Mandache for their work in refining the methodology. The coding manual is available upon request.

6. The formula for Scott’s Pi, explained in Riffe, Lacy, & Fico (1998), is a modified Chi-Square that takes into account both chance agreement and the strength of the operationalization of each variable. Pi = %OA – %EA / 1 – %EA where OA = observed agreement and EA = expected agreement. The two coders had intercoder reliability scores of greater than 80% for every value. The authors would like to thank research assistants Anne Marie Reynaud and Noah Boudreau for their coding work.

7. The prominence variable is a combination of the number of illustrations accompanying an article and the number of words in the article. We gave each illustration (photograph, map, chart) a value of 100 words. The articles were then assigned one of five categories: brief (up to 199 words), short (200 to 399 words), medium (400 to 599 words), long (600 to 799 words), or very long (800 words or more).

8. Besides the Canadian newspaper study in 2003, we conducted a study of three U.S. newspapers (the New York Times, Chicago Tribune, and Los Angeles Times) in 2004.

9. In a study of British news websites, Neil Thurman (2007) found that Americans made up an average of 36% of the audience, and up to 39% of readers came from countries other than the United States. A majority of The Times’ online readers came from outside the United Kingdom, 41% from the U.S. (p. 291).

10. We downloaded seven issues of The Times between Thursday, May 27, and Wednesday, July 14. We downloaded seven issues of Libération between Wednesday, May 26, and Monday, August 2. (Due to technical problems on the downloading days of Sunday, June 27, and Monday, July 5, we substituted Sunday, July 25, and Monday, August 2.) We downloaded seven issues of Ha’aretz between Friday, May 28, and Thursday, July 15.

11. Countries were counted as “cited” in an article whenever they were explicitly mentioned. This included both passing references and more substantial discussions. Besides countries, we coded as well for supranational governing institutions—the United Nations, the European Union, and the Vatican—which newspapers customarily assign placelines.

12. Stories were categorized as “other” if they were included in sections that either did not fit one of the conventional topic categories or were a mix of topics. This included stories on automobiles, real estate, and technology.

13. Coverage by Canadian news media of the tsunami that hit Asia in December 2004 prompted considerable criticism because of their focus on the Canadian angle: the few Canadian victims and the inadequacy of Canadian relief efforts (e.g., Blouin-Gagné, 2005).


Anderson, Benedict. (1989). Imagined communities: Reflections on the origin and spread of nationalism. London & New York: Verso.

Biltereyst, Daniel. (2001). Global news research and complex citizenship: Towards an agenda for research on foreign/international news and audiences. In S. Hjarvard (Ed.), News in a globalized society (pp. 41-62). Göteborg: Nordicom.

Blouin-Gagné, Isabelle. (2005, June 4). Tsunami riders: ‘Extreme’ tourism, disasters and the media. Paper presented at the annual conference of the Canadian Communication Association, London, ON.

Carey, James W. (1998). The Internet and the end of the national communication system: Uncertain predictions of an uncertain future. Journalism and Mass Communication Quarterly, 75(1), 28-34.

Castells, Manuel. (1999a). The rise of the network society. Oxford: Blackwell Publishers.

Castells, Manuel. (1999b). The power of identity. Oxford: Blackwell Publishers.

Castells, Manuel. (1999c). End of millennium. Oxford: Blackwell Publishers.

Castells, Manuel. (2001). The Internet galaxy: Reflections on the Internet, business, and society. Oxford: Oxford University Press.

Chen, Guo-Ming, & Starosta, William J. (2000). Communication and global society: An introduction. In G. Chen & W. J. Starosta (Eds.), Communication and global society (pp. 1-16). New York, NY: Peter Lang Publishing.

Durham Peters, John. (2000). Speaking into the air: A history of the idea of communication. Chicago, IL: University of Chicago Press.

Gans, Herbert J. (2004). Deciding what’s news: A study of CBS Evening News, NBC Nightly News, Newsweek, and Time. Evanston, IL: Northwestern University Press. (Original work published 1979)

Gasher, Mike. (2003). Paper routes: The geography of news in digital times. In G. Tremblay (Ed.), Panam: Cultural industries and dialogue between civilizations in the Americas (pp. 470-485). Québec, QC: Les Presses du l’Université Laval.

Gasher, Mike. (2007). The view from here: A news-flow study of the on-line editions of Canada’s national newspapers. Journalism Studies, 8(2), 299-319.

Gasher, Mike, & Gabriele, Sandra. (2004). Increasing circulation? A comparative news-flow study of the Montreal Gazette’s hard-copy and on-line editions. Journalism Studies, 5(3), 311-323.

Gerbner, G., Mowlana, H., & Nordenstrang, K. (Eds). (1993). The global media debate: Its rise, fall, and renewal. Norwood, NJ: Ablex Publishing.

Hamilton, James. T. (2004). All the news that’s fit to sell: How the market transforms information into news. Princeton, NJ: Princeton University Press.

Hartley, John. (1992). The politics of pictures: The creation of the public in the age of popular media. London & New York: Routledge.

Hjarvard, Stig. (2001). News media and the globalization of the public sphere. In S. Hjarvard (Ed.), News in a globalized society (pp. 17-39). Göteborg: Nordicom.

ICSCP [International Commission for the Study of Communication Problems]. (1980). Many voices, one world: Report of the International Commission for the Study of Communication Problems. London, New York, & Paris: Kogan Page, Unipub, UNESCO.

Lefebvre, Henri. (1991). The production of space. Oxford: Basil Blackwell.

Manthorpe, John. (1998). The decline of foreign news in Canada’s media. In D. Logan (Ed.), Journalism in the new millennium (pp. 131-140). Vancouver, BC: Sing Tao School of Journalism.

Massey, Doreen. (1991, June). A global sense of place. Marxism Today, 24-29.

Mattelart, Armand. (2000). Networking the world: 1794-2000. Minneapolis, MN: University of Minnesota Press.

McPhail, Thomas. (1987) Electronic colonialism: The future of international broadcasting and communication. Newbury Park, CA: Sage Publications.

Mosco, Vincent. (2004). The digital sublime: Myth, power, and cyberspace. Cambridge, MA: MIT Press.

Mowlana, Hamid. (1997). Global information and world communication: New frontiers in international relations. London: Sage Publications.

Picard, Robert G. (1989). Media economics: Concepts and issues. Newbury Park, CA: Sage Publications.

Riffe, D., Lacy, S., & Fico, F. (1998). Analyzing media messages: Using quantitative content analysis in research. Mahwah, NJ: Lawrence Erlbaum Associates.

Sassen, Saskia. (1998). Globalization and its discontents: Essays on the new mobility of people and money. New York, NY: The New Press.

Shoemaker, Pamela J., & Cohen, Akiba A. (2006). News around the world: Content, practitioners, and the public. New York & London: Routledge.

Sreberny-Mohammadi, Annabelle. (1984). The ‘world of the news’ study. Journal of Communication, 34(1), 121-134.

Thurman, Neil. (2007). The globalization of journalism online: A transatlantic study of news websites and their international readers. Journalism, 8(3), 285-307.

Thussu, D. K. (2004). Media plenty and the poverty of news. In C. Paterson & A. Sreberny (Eds.), International news in the twenty-first century (pp. 47-61). Eastleigh, U.K.: John Libbey Publishing.

Van Dijk, J. A. G. M. (2005). The deepening divide: Inequality in the information society. Thousand Oaks, CA: Sage Publications.

Vasil, Adria. (2003). World domination. Ryerson Review of Journalism, 20(2), 56-61.

Westcott, C., & Mukherjee, J. (2004). New media at the BBC World Service. In C. Paterson & A. Sreberny (Eds.), International news in the twenty-first century (pp. 79-94). Eastleigh, U.K.: John Libbey Publishing.

Wu, H. Denis. (2000). Systemic determinants of international news coverage: A comparison of 38 countries. Journal of Communication, 50(2), 110-130.