ARGUING INTO THE DIGITAL VOID? ON THE POSITION OF ONLINE DEBATES IN THE LOCAL PUBLIC SPHERES OF FOUR NORWEGIAN MUNICIPALITIES Abstract With the advent of the Internet, numerous online debate options have been created, giving citizens new arenas of political communication where space for expression is nearly unlimited. However, if online forums shall invigorate the public debate, the arguments published online must reach outside their initial setting. In this article, the position of newspaper-hosted online forums is studied and compared to the position of letters to the editor in the local public spheres of four Norwegian municipalities. The forums' visibility to the public and the degree to which they are paid attention to by other media and by local politicians are used as indicators of their position. Only one-tenth of the citizens regularly read the online forums, and when referred to in other media or in politics, they have an agenda-reinforcing rather than an agenda-expanding role. The greatest challenges to the online forums' position seem to be the audiences' perceptions of the quality of the discussion, along with the vast amount of contributions which makes them reader unfriendly. MARTE WINSVOLD Marte Winsvold is researcher in the Norwegian Institute for Urban and Regional Research, Oslo; e-mail: marte.winsvold@nibr.no. m Cfi n d a ro o o o rN £ I 0 £ Keywords: online debates, letters to the editor, public sphere, agenda setting, visibility II) G C (6 m With the advent of the Internet, numerous online debate options have been created, giving citizens new arenas of political communication where space for expression is nearly unlimited. All-inclusive participation in the public sphere has hence become a theoretical possibility. However, if online forums shall invigorate the public debate, the arguments published online must reach outside their initial setting. Few studies have been conducted about the actual reach of online debates. Does the public pay attention to them? Do they impact the political debate in other arenas? Do they reach the political agendas of elected bodies? This article's aim is to increase knowledge about the reach and impact of online debate as compared to the debate taking place in traditional media arenas of citizen participation. Such knowledge is important in understanding whether the public sphere is changing due to new arenas of communication made available by digital technology. Three aspects are singled out as especially important to the position of a specific arena in the public sphere: its visibility to the public, the degree to which content spills over to other public arenas, and the degree to which arguments from it are channelled into the decision-making system. This article compares the position of newspaper-hosted online forums in the local public spheres of four Norwegian municipalities to the position of letters to the editor in the same municipalities' local newspapers. The analysis is based on data collected in a population survey, a survey of politicians, and in interviews with local politicians and journalists in the four municipalities. What Determines the Position of an Arena in the Public Sphere? "The public sphere" is a contested concept, and scholars with different theoretical perspectives have focused on different features as being constitutive. Common to most perspectives however, is the presence of an audience (Arendt 1958; Habermas 1989; 1996; Thompson 1995; Hoff and Storgaard 1995). The "public" denotes that expressions are visible or audible to the members of a given community, and the public sphere manifests itself when something is spoken for everyone, or at least 2 for a significant number, to hear. When we know that others have heard the same £ argument or opinion as we have, we know that others relate to it and potentially act on it. Public expressions thereby become real in their consequences. For a communication arena to play a role in public debate, it therefore has to reach outside its initial setting and be visible to the general public. This visibility is an important feature of the public sphere. However, according to Hannah Arendt (1958), being visible to a certain amount of the population is not enough; the audience also has to be representative of the population. Arendt argues that an expression gains importance from the fact that everyone perceives and interprets it from their particular position. If an expression is perceived and interpreted only by a homogeneous segment of the population, it will be critically evaluated and judged only through the limited perspective of this specific group. A plurality of perspectives is necessary for an expression to be sufficiently critically examined. Following this vein of reasoning, for a communication arena to be truly public the audience should be of a certain size and it should be heterogeneous. Moreover, some scholars claim that for a communication arena to be public there ^ must be reference in other public sphere arenas to the arguments and ideas occur- C > ring in it. Critics of the early Habermas concept of a unitary public sphere argue that the public sphere consists of a complex network of various communication arenas or sub-spheres (Fraser 1992; Habermas 1996; Hoff and Storgaard 2005). For a communication arena to be public, it must be connected to other publicly visible communication arenas. The sub-spheres must be porous to one another, meaning that an idea, opinion or argument appearing in one sub-sphere may extend into other sub-spheres (Habermas 1996, 374). The constellation of sub-spheres will thus constitute a multi-faceted, but still coherent public sphere. If a sub-sphere is not connected to other sub-spheres, the stated arguments and points of view will not reach outside the specific setting where they occur, implying that they will be isolated from the wider public debate and hence must be regarded as private. In line with later Habermas, a second important feature of the public sphere therefore is that its communication arenas be somehow interconnected. Furthermore, Habermas (1989; 1996) holds that for a communication arena to be considered as part of the public sphere, ideas and arguments emerging from the discussion in it must be channelled into political decision-making institutions and potentially influence the agendas and priorities of the politicians. Thus, for a communication arena to be part of the public sphere, what is uttered there must reach the politicians. In this article, such channelling of ideas and arguments is considered as a third important feature that makes an arena part of the public sphere. In this article, the "position" of an arena in the public sphere refers to the arena's relative importance as compared to other public arenas of communication. The position of a specific arena is assessed by the extent to which the arena is visible to the general public, the extent to which it is referred in other media, and the extent to which it is considered in political decision making. The questions asked to evaluate the position of online debates in the public sphere are therefore: Are the online debates visible to the citizens? What proportion and what part of the population read the online debates, compared to those reading, watching and listening in other arenas of public debate? Are the online debates connected to other arenas of public debate? Do ideas or arguments raised in online forums make their way into other media? Are ideas or arguments from the online debates channelled into and considered in political decision making? Do politicians refer in internal party discussions or in council meetings to ideas or arguments occurring in these debates? In normative descriptions of the public sphere, certain qualities are often required of the discussion taking place therein, regarding for example the rationality of arguments, the reflexiveness of the dialogue or the representativeness of the participants (Habermas 1989; 1996; Dahlberg 2004; Janssen and Kies 2005). This article will not focus on such normative claims in assessing the position of the different arenas in the public sphere. Rather, its focus is on the reach of the content. However, the way in which normative expectations are met may be highly important in explaining the position of an arena in the public sphere. In discussing the findings, attention will therefore be drawn to the audiences' perceptions of how a public discussion ought to be conducted. Comparative Study in Four Norwegian Municipalities To explore the comparative position of online debates in the public sphere I will draw on data gathered in four Norwegian municipalities: Drammen, Troms0, Stavanger and F0rde. These municipalities were chosen because they all had local newspapers with debate forums attached to their online editions. The municipalities also had other arenas of public debate, such as local radio and television, with which the online debates might compete. It turned out, however, that letters to the editor in the local newspapers were the most important among the traditional arenas of public debate. The position of the online forums will therefore mainly be compared to the position of the letters. The municipalities varied in size from F0rde with about ten thousand inhabitants, to Drammen and Troms0 with nearly sixty thousand each, and Stavanger with a little over a hundred thousand inhabitants. The municipalities are situated in different parts of Norway: F0rde and Stavanger in the west, Troms0 in the north and Drammen in the east. They were not representative of the totality of Norwegian municipalities, but as they differed in size and geographical location, findings consistent across them may indicate tendencies that are more general. In terms of ICT infrastructure and availability, these municipalities, as well as the rest of Norway resembled other Scandinavian countries in that Internet access was high - 80 percent in 2005 according to TNS Gallup. The high Internet access could enhance the political importance of online debate forums. However, unlike the local press in many other European countries, the local press in Norway has had a remarkably strong and lasting position, which could hamper the position of new arenas of debate such as the online forums. So far, the local press has not seemed to lose readership to Internet publishing to any substantial degree, as has been the case with the national press (Hast 2005; Vaage 2008). The forums analysed were hosted by the online editions of the four leading local newspapers in the municipalities: Stavanger Aftenblad (www.Aftenbladet.no) in Stavanger, Drammens Tidende (www.dt.no) in Drammen, Nordlys (www.nordlys.no) in Troms0 and Firda (www.firda.no) in F0rde. The discussions were organised differently in the different forums. In dt.no andfirda.no, the discussions were pre-organ-ised in broad categories like "politics" and "culture," but the participants decided the specific subject matter for discussion. In nordlys.no, the web editor structured the debate according to questions decided on by the editorial board, whereas the discussion in aftenbladet.no was related to articles published in the newspaper's online version. In all forums, local political issues were discussed to an extensive 3 degree.3 Two of the forums were much used. In dt.no, there were approximately 100 new postings daily, in aftenbladet.no 200. In nordlys.no postings averaged only five daily, whereas infirda.no postings averaged only five weekly. There were at least two good reasons to choose the newspaper-hosted online forums over other types of local online forums, for the study. First, there were no ^ other local online forums in the four municipalities with any substantial activity or to which the politicians or the public paid particular attention. When choosing the newspaper-hosted online forums, I hence chose the online forums that probably had the most prominent position among local online forums in the respective local public spheres. Second, these forums were situated outside the formal governmental structure. Of the few studies conducted on the position of online forums, most have in fact focused on forums hosted by governmental bodies (see e.g. Coleman 2004; 0 fi C > Hoff and Storgaard 2005). However, the fact that they are government initiated violates the ideal that civil society must generate a dialogue that reaches political ^ elites without being sanctioned by them (Brandenburg 2006). Like government- initiated forums, newspaper-hosted forums do not truly belong to civil society, as media institutions also may exert censoring power. Still, their publicly stated aim 00 was that they should encourage free debate among citizens, and this locates them ^ closer to the ideal of a free and undistorted arena of civic public debate, as they are designed for the purpose of debate, not for the purpose of providing political authorities with input on specific and already established political issues. The public sphere position of online forums is compared to the position of letters to the editor in the printed versions of the newspapers. All four newspapers published letters to the editor in their printed versions and allocated approximately equal space to such letters. The letters in the papers had no clear organisational principle: Letters on the same subjects were sometimes presented together, but this was not always so. To assess the visibility of the two types of forums, a survey questionnaire was mailed in October 2005 to a randomly selected sample of 3,600 inhabitants in the four municipalities. The response rate was only 35 percent, varying slightly between the municipalities. Compared to data gathered in an official national survey with a response rate of 69 percent (Saglie and Bjorklund 2005), the bias of the present sample does not however appear to be very large. Analysis of the data indicates that the most politically active part of the population is overrepresented in the survey. Furthermore, respondents under 30 years old are underrepresented, and thus the data are weighted for age in proportion to the age distribution in the population. The reported frequencies must in any case be interpreted with caution, as they may overestimate the overall local political activity. Interpreting results from multivariate analysis is less problematic since political activity and several other socio-demographic biases in the sample are then controlled. To assess the interconnectedness and the channelling function of the two arenas, interviews with 12 newspaper journalists and 23 local politicians were conducted in 2004 and 2005.1 As for the politicians, to correct the bias that might result from a limited sample, a survey questionnaire covering many of the same questions was sent to all (186) elected representatives in the four case municipalities in April 2005; 100 (54 percent) politicians responded. There is no indication of the nature of the bias in this sample, but these results must also be analysed with caution. As for the journalists, those responsible for covering local politics were interviewed, as were web editors and political editors. Newspaper journalists were chosen over radio and television journalists because the local newspapers attracted the largest audience - more than 90 percent of citizens read the local newspapers at least once weekly. In addition, the newspaper journalists had a more clearly local focus than the TV and radio journalists, who had a more regional focus. As discussed above, the limited representativity of the survey sample and of the interviewees reduces the generalisability of the study. Moreover, the study is carried out in four localised municipalities that are not representative of Norwegian municipalities and even less representative of local communities in other parts of Europe or the rest of the world. As the findings are contingent on the contexts of these specific online forums, they have hence limited validity in other contexts. To assess the generalisability of the findings, similar studies would have to be carried out in other contexts. Still, the study can hopefully tell us something more general about the mechanisms and processes at play in public spheres where new arenas emerge while old arenas are still present. Visibility - Do the Debates Reach the Public? To evaluate the position of online forums in the local public sphere we first need to establish their public visibility. According to the definition applied, the visibility of an arena consists of two aspects - the size of the audience and its heterogeneity. A survey among citizens in the four municipalities showed the extent to which the local population read what was published in the two forums. As figure 1 shows, online debates were not especially visible to citizens, compared to the debates taking place in the newspapers' letters to the editor. Figure 1: Proportion of Population in Four Municipalities Reading Online Debates and Letters to the Editor3 (N = 1275) 67 5 26 69 25 8 Online debates Letters to the ed tor 0 □ Never □ Month y or less often □ Daily or weekly_ Although nearly a third of citizens stated that they occasionally visited the y newspaper's online forum, less than a tenth did so weekly or more often. By comparison, two-thirds regularly read the letters to the editor in the newspaper's print edition. The letters were hence visible to a far larger part of the public than were the online forums. From the few other studies conducted on online debate audiences, audiences of approximately this size seem to be the norm: During the presidential campaign in 2008 in the U.S., for example, nearly twenty percent of the adult population had read someone else's commenting on the campaign on online newsgroups, websites or blogs (Smith 2009). As for the second aspect of visibility, however, analysis of the survey data showed that the online audience was more heterogeneous than was the letter audience. The online forum audience, (with some exceptions), was biased in the same way as the general Internet audience (Wellman et al. 2001; Norris 2004; Karakaya Polat 2005; Di Gennaro and Dutton 2006): Those reading the postings of the online forums were younger, more politically active, more male, and had higher income than the average citizen. However, the letter audience was in some respects even more biased than the online audience. The results, displayed in table 1, from a linear regression analysis, show the controlled effects of individual characteristics ^ on online and offline readership. 0 fi ¥ I C > Table 1: Readership of the Newspaper's Online Forum and Reader's Letters in the Newspaper's Print Edition4 (Bivariate Correlation and Linear Regression lo Coefficients) ^ Online Forums Letters to the Editor Pearson Correlation Beta (Standardized coefficient) Pearson Correlation Beta (Standardized coefficient) Political activity index 0,228* 0,233* 0,253* 0,194* Gender 0,141* 0,127* 0,096* -0,038 Income 0,107* 0,067* 0,109* 0,187* Age -0,001 -0,068* 0,396* 0,409* Education -0,012 -0,068* -0,064* -0,053 Tromsr (dummy) -0,46 -0,106* 0,048 0,017 Frrde (dummy) -0,14 -0,080* -0,020 -0,035 Stavanger (dummy) 0,003 -0,070 -0,076* -0,090* Drammen (reference) Adjusted R square 0,078 0,234 N 1036 1077 * Significant at a 0,05 level Clearly correlated to both online and print readership was the overall political activity of the respondents. An activity index was constructed based on eight questions concerning offline political activity. The analysis showed that the most politically active citizens were most inclined to read the online debates. That the online forums were most visible to the most politically active meant that the segment of the population absorbing arguments and ideas from this arena was the same that participated in other political arenas in the municipalities and thereby could take online arguments and ideas with them to other arenas, enhancing their reach. This may imply that the position of the online arenas was somewhat stronger than the proportion reading them indicated. The same goes for the letters, as the politically active were most inclined to read them as well. The effect of political activity was strong for both forums and contributed to making both audiences less heterogeneous than the ideal demands. The emergence of online forums hence seems to have reinforced the tendency of this group's overrepresentation in the public sphere. This finding mirrors findings from other studies, showing that those who are the most politically active offline also are the most politically active online (Smith 2009; Torpe 2005). Gender correlated with online readership as expected, but the gender gap was larger than expected: Whereas thirteen percent of men stated that they visited the online forums regularly, only four percent of women did so. Although the gender gap in overall Internet use has decreased in Scandinavia during the last ten years (Tobiasson 2005), these forums still seemed to attract a predominantly male audience. Assuming male and female readers have different perspectives, the discussions in the online forums were mostly interpreted from a male point of view. There was also a statistically significant gap between male and female print readership, but the difference here was far less. In other words, the advent of online forums had made the public sphere audience more lopsidedly male. Income positively correlated with both types of readership, indicating that high income spurred readership. However, income meant comparatively less to online than to offline readership. Age did not correlate with online debate readership in the same way as it has been shown to correlate with general Internet use: Although the beta coefficient of the regression analysis shows that younger people were somewhat more inclined than elder people to visit online debates, the effect of age was weak, and at the bivariate level it was not even statistically significant. This finding is contrary to the findings from the earlier mentioned survey conducted among American adults (Smith 2009), which find that those under 30 years old are significantly overrep-resented among the online debate audience. Concerning readers of the letters to the editor, the effect of age was striking, but with an opposite sign: Elder people were much more inclined to read letters than were younger people, and age was the variable that most strongly predicted letter readership. Earlier research has shown that the Internet audience has higher education than the average citizen (Norris 2004; Karakaya Polat 2005; Di Gennaro and Dut-ton 2006), but this did neither apply to the online nor to the letter audience: Those with lower education were more inclined to read the online forums as well as the letters, than were those with higher education. Research on TV audiences shows that the same applies to TV debates; their audience is less educated than audiences of political TV programs with other formats (Brenna 2007). It hence seems like the debate format appeals to the less educated segments of the population, both offline and online. According to the definition of visibility applied in this study, two aspects affect the visibility of an arena in the public sphere, namely the size and the composition of the audience. Regarding the first aspect, demanding that the content of the arena be read by a significant proportion of the concerned population, the letters were clearly more visible than the online forums. As for the second aspect this demands that the audience be heterogeneous, which means that background characteristics should as little as possible affect the tendency to read the different forums. The size of the adjusted R2 tells how much of the variance in readership can be explained by the background variables included in the regression model. From the adjusted R2 of the two models, we see that the totality of the applied background variables explains less of the online than the offline readership. This means that the online audience was less biased and more heterogeneous than the offline audience regarding the control variables, and that the online audience to a greater extent, were visible in £ the sense "heterogeneous." The two forums' positions regarding visibility were hence ambiguous. The letters were visible to more of the population, whereas the ^ online forums were visible to a more heterogeneous sample of the population. Interconnectedness - Do the Debates Reach other Media? Q If the online forums are to be connected to other public sphere arenas, ideas or arguments occurring in them need to be referred to in those arenas. To assess the interconnectedness of online forums with other communication spaces, twelve 3 a c > journalists in the newspapers hosting the online forums were asked whether they picked up ideas from the online debates or referred to arguments from them in ^ their newspaper articles. Although their attitudes towards the online forums var- ied from positive to outright hostile, their referring practices scarcely varied: All of them visited the forums regularly, but they claimed that they rarely or never referred to arguments or ideas appearing there. When they did, it was mostly as "curiosity" (Journalist 12) or "to show that a specific subject matter attracted the public's attention" (J. 6). None of the journalists thought that the forums gave them new ideas or introduced new subjects of which they had not previously thought. Rather, the journalists would sporadically monitor the contributions to confirm that they themselves had indeed captured all relevant arguments and views, but they rarely thought that they learned something new from this monitoring. Still, the attitudes towards the forums differed considerably across the newspapers. The journalists in Drammens Tidende, for example, had low esteem for the online forums, and among other things described them as "a whining post with low credibility" (J. 1). The attitudes of the journalists in Firda were the most positive. "The online debate is interesting as a source of new ideas, and what we see now is only the beginning," a Firda journalist stated (J. 6). However, as we saw, they all claimed that they rarely or never referred to arguments or ideas appearing online; furthermore, all journalists regarded the online debate as of little importance when it came to agenda setting: "The agenda is set by politicians, pressure groups and media. The role of the online debate is merely to get the temperature up" (J. 10), one journalist said. Journalists from all newspapers agreed that although arguments from the online forums were occasionally channelled into the newspapers, new subjects appearing online that widened the media's agenda never were. All journalists pointed to the letters to the editor when asked how citizens normally contributed to the media's agenda. They all reported that they picked up ideas from the letters and published articles on the subjects - either by responding to the letters or by interviewing the letter writers for ordinary newspaper articles. The journalists gave several reasons why they chose to refer to letters and not to online forums. Many of them attributed their lack of attention to online forums to the quality of the forums' content. The online content was for example described as "fast and unfounded nonsense" (J. 4). Their attitude towards the quality was however ambiguous. On one hand, the superficial or flimsy style was cited as a reason for not paying attention to the contributions. On the other hand, they stressed the importance of such arenas' having open access and low demands as to the form and style of arguments. In this line of reasoning, the debates were described as for example "to the point, inclusive and alive" (J. 7), and they were deemed as necessary democratic channels. The journalists moreover considered the online participants as slightly more representative of the population than writers of letters to the editor; as one journalist said, "My impression is that those who write letters to the editor are well-off men, while the variation is larger among those who write online concerning income, education and social status. You find both the school drop-outs and the masters students" (J. 1).Online discussions were considered as "a call from the heart of the people" (J. 7). That the online contributors were perceived as representative did not, however, make the journalists pay more attention to these forums. Moreover, that the forums lacked editing was perceived as a problem by the print journalists. They argued that the lack of editing made the forums hard to read and hard to extract the essence from. In aftenbladet.no, for example, there could be as many as 200 new postings daily. Not all were related to the subject under which heading they were posted, a lot of them veered off the subject, and they were not always responses to the previous postings (Skogerb0 and Winsvold 2008). In sum, the online forums were considered reader unfriendly, and this gave them less impact with the print journalists as a source of arguments or new ideas. From the perspective of the journalists it was considered both time and frustration saving to interview a sample of citizens about their opinions instead of struggling through the vast amount of online postings, many of which were seen as nonsense. The online forums were neither valued as sources, as were the letters, nor acknowledged as independent arenas in the public sphere. Overall, the forums were considered as tolerable, but unnecessary noise, to which the journalists had to pay some attention because their own newspaper hosted them, but with which they could have done without. The online forums were not entirely unconnected to the journalist-driven debate that took place in the newspaper, but the letters were far more referred to and therefore far more connected to the media debate. The dominant actors in the media world, in charge of the media agenda, had hence not altered their practice much due to this new arena. The few studies that exist on digital media's agenda-setting effect mostly point in the same direction. In their study of the agenda-building role of social media Lariscy et al (2009) found very little use of social media among American business journalists. Lee (2007) found in his study from the U.S. presidential election of 2004, that the blog agenda was similar to that of mainstream media, whereas Seweetser et al (2008) found in their study of the same election that the correlation between the blog and the media agenda occurred because the media transferred their agenda to the blogs. However, from their study of the interplay between online postings and traditional offline media coverage in China, Zhou and Moy (2007), found the opposite: In China online postings played an important role in transforming local events into national media issues. Moreover, online posting exerted a significant y frame-building impact on offline media in the early stages of coverage. The varying impact of online communication arenas on the mass media agenda indicates that their importance is highly contingent on the context. Channelling - Do the Debates Reach the Political Agenda? ft 0 If arguments or ideas from online forums shall be channelled into the political agenda, politicians need to read and refer to them in arenas where political agendas are decided. The politician survey showed that half of the politicians (50 percent) had visited the online forums and read contributions posted there, although fewer (37 percent) did so regularly. Still, a higher percentage of politicians than of ordinary citizens read online debates. The interviews revealed, however, that the ways in which politicians monitored the forums did not necessarily give this arena an independent agenda-setting role. Some politicians selectively and purposefully read contributions in the online forums that were related to issues they were responsible for in the Municipal Council. They thus systematically gathered information about a few specific topics of concern to them. Especially those responsible for politics on children and youth used the online forums because they believed them to be the ^ "youths' channel" (Politician 18). Other politicians reported reading online contribu- C > tions on subjects that were substantially covered in the media. Used by politicians in these ways, online forums did not contribute to setting new issues on the agenda, C^ as politicians searched out information on issues that were already on it. ^ The forums' policy regarding anonymity seemed to impact the way politicians used them and hence their position in the local public sphere. Several politicians said that they ignored anonymous postings, and conversely, signed articles attracted politicians' attention. In aftenbladet.no in Stavanger, where participants had to give their real names, several politicians stated that they studied the forum to "see who is writing" (P. 19). That it was interesting for them to see who was partaking in the online forums indicated furthermore that these politicians did not feel sure about these forums' position in the local political debate; if important stakeholders participated, they had to be paid attention to, if not, they could be ignored. Therefore, for these Stavanger politicians the online forums' position was not clearly defined, but was rather fluid and dependent on how they were used and by whom. It would be a stronger indication of channelling if the postings were explicitly referred to in agenda-setting forums than if they were merely read. According to the politician survey, one-third of politicians (34 percent) stated that they had used arguments from the online debates in political discussions, although only 17 percent had done so more than once. Interviews showed that the nature of this usage did not widen the political agenda, but rather confirmed it. Most politicians that referred to online debates did so concerning issues that were already on the political agenda and that were to be discussed in Municipal Council meetings. They used arguments or opinions occurring in the forums to support their own opinions, to prove that they were "backed by ordinary people" (P. 21) or to show what "most people mean" (P. 27). Moreover, politicians used the online debates as a kind of polling instrument, to measure the importance of issues on the political agenda. The online debates gave an indication of how important an issue was to citizens: "On the net, it is measurable how many are interested in a specific subject. You can very quickly gauge whether an issue is deemed uninteresting or fabulously interesting" (P. 16), one politician stated. Only two of the 23 interviewed politicians reported that they had actually picked up new issues from the forums, which they then had discussed in their political party and thereafter brought to the Council agenda. All politicians named the newspaper as the most important public arena in influencing the political agenda. They considered both editorial coverage and letters to the editor to be important mainly because of their high visibility, but felt editorial coverage was more important than letters to the editor. Accordingly, the online debates were regarded as unimportant because of their low visibility. Still, a certain agenda-setting value was ascribed to the online forums, mainly because "such things spread [...] and the press may cover it" (P. 20). The press was even here indirectly important as it guaranteed the dissemination of opinions. The attitudes of the interviewed politicians concerning the newspapers' greater importance were reflected in the survey: Only 4 percent of politicians fully agreed that it was important for them to keep an eye on the online debates, compared to 38 percent who agreed it was important for them to keep an eye on the letters to the editor. Although they referred to the online debates as reflecting the opinions of ordinary citizens, when asked directly, most politicians did not consider such opinions to be representative of the majority's opinions, and many thereby justified ignoring the arguments and issues occurring in the online debates. The politicians believed that few people participated, and many also thought that a special kind of people expressed themselves in these forums: "It is the cantankerous or quarrelsome who like to vent their opinions" (P. 33); "people with strong and simple answers to difficult questions" (P. 14); "with much time to spend" (P. 20). Moreover, several politicians thought that people's online utterances were not representative of their true attitudes. "Only when you are discontented, furious, or against something, you go online to demonstrate your opinion," one politician said (P. 15). These politicians held, hence, that they did not need to consider what occurred in these forums because people did not necessarily mean what they said. They were just venting frustration, and when they were done, they resumed their real and more moderate opinions. These politicians regarded the opinions stated online as invalid because of the argumentation style. Related to this, the participants' motivation was also questioned and deemed as insincere and even destructive: "The participants enter some predefined roles. Some shall always provoke, some shall always be mad" (P. 33), one politician said. Others thought the contributors participated most for fun without expecting to be taken seriously. The attitude of many politicians was that these were illegitimate motivations, and that the participants should have endeavoured instead to discuss issues of concern in a constructive, serious manner. Politicians, like journalists, perceived the fact that everyone could express themselves in these forums as reason for both paying and not paying attention to them. The online debates were regarded as reflecting the people's voice, but simultaneously much content was perceived as unfounded opinions, lacking the filtering and refining of the mass media's editing process. So, for both journalists and politicians, the open access, which should ensure plurality in the public debate by allowing more people to take part, simultaneously decreased the forums' status. Some politicians even explicitly mentioned the perceived lower status of the online contributors to explain why, for example, letters were taken more seriously: "[In the letters] people with a little more weight are published" (P. 31).The restricted y access to the newspaper, favouring those with a respected position in the local community, hence elevated the letters' status. Politicians regarded the quality of the online arguments as generally low but, like the journalists, they did not necessarily look upon this as negative. They described the style as direct, to the point, a discussion with "strong, sharply pointed and concise expressions" (P. 25). The spontaneous form and fast response were also regarded as positive. At the same time, the contributions' low quality was cited as a reason for not paying attention to opinions expressed in the online debates: "It is full of personal attacks, ill-founded and unserious. It is too easy to participate. People just say "no," but you don't know why" (P. 14), one politician said. The discussion was frequently compared to the discussion that takes place in toilet stall graffiti. So, like the journalists, the politicians read the online forums mainly to check i4- 4- U 7 U r^rl ^ I I vaI ATT^MI- Af I- ^ -i ri i n rv 4-/-\ irnmr ^IfA^i^T? AM C > if they had captured all relevant arguments pertaining to issues already on the political agenda. When they referred to the online forums they did so, for the most part, to show that they had citizens' support. Only to a very modest extent new arguments and ideas from the online forums reached the political agenda; mostly the forums were used by politicians in an agenda-confirming rather than Ij^ in an agenda-expanding way. As a source of public opinion, letters to the editor were deemed much more important, primarily due to their larger visibility, but also due to their being edited, filtered and therefore seemingly quality-assured by the newspaper. Only to a minor degree therefore, can the advent of online forums be said to have changed the public sphere regarding citizens' access to the political agenda, in these municipalities. Still, politicians were aware of the forums, they monitored them and were ready to pay more attention to them should it prove necessary. This result is in line with what Torpe (2005) found when assessing the general impact of a Danish municipal online forum. Moreover, the politicians' surveying practice showed that they acknowledged that the online forums were, although perhaps not important arenas in the public sphere, at least independent sources of public opinion. The channelling of online content was direct, not indirect via the media, as the politicians surveyed these forums to a greater extent than the media referred to them. To some extent, therefore, this arena represented a new channel through which public opinion could reach the political agenda without being filtered by the media. Conclusion The position of online debates in the local public sphere proved to be modest compared to the position of long-established public arenas such as letters to the editor in newspapers. Online debates were visible to a far smaller proportion of the population than were the letters. Still, the online audience was more heterogeneous than the letter audience, as the online content reached a more diverse sample of the population. Online forums were rarely referred to in other media and arguments, and ideas appearing in them only occasionally entered the political agenda. When they were referred to, the references concerned mainly arguments about issues already on the media agenda or the politicians' agenda, rarely new ideas or perspectives. The referring practice hence had an agenda-confirming rather than an agenda-expanding function. Compared to the position of the letters to the editor which was the main other arena where citizens could participate in the local public debate, the position of online forums hence seemed to be of less importance. All in all, in the four municipalities studied, the public sphere had not changed considerably due to these new outlets of expression. Although they were to some extent read by the citizens, occasionally referred to by the newspapers and laxly monitored by the politicians, all three audience groups still paid considerably more attention to the traditional media arenas which they were used to. The novelty of online debates is one obvious explanation for their lacking impact. Yet another explanation of the online forums' modest position may be the way in which normative claims of public dialogue were met in the debates. Politicians and journalists seemed to expect that the debate should meet certain minimum standards regarding quality, pluralism and sincerity. These expectations guided both the interpretation of the content and the extent to which the online debates were actually read and arguments expressed in them actually referred to or taken into account. Both the perceived low quality of arguments and the presumed bias of the participants were repeatedly cited by politicians and journalists as reasons for not paying attention to the online debates. Moreover, the participant's motivations were deemed to be dubious. A widespread belief among politicians and journalists was that the participants did not have serious intentions, but rather participated for fun or because they liked a good fight. Regardless of where it was published, both journalists and politicians seemed to have an ambiguous attitude towards content produced by ordinary citizens. On one hand they depended on it, as the public debate should appear to be representative of the citizens' concerns and of their different perspectives. In this vein of reasoning they used both letters and online forums as information sources. On the other hand, neither journalists nor politicians acknowledged that citizens had an agenda-setting role in the public debate, at least not a prominent one. For them, agenda setting was something that came out of the editing or filtering process which they undertook; the citizens' role was only to provide the raw material of opinions and concerns. The online forums' greatest challenger was therefore not one of the traditional media arenas, but rather the way political agenda setting was conceived of by those with agenda-setting power. Another major challenge to the position of online communication arenas was the huge volume of contributions resulting from their inclusiveness. In the pre-Internet era, the public debate took place in very few media arenas, to which only a minority of citizens had access as participants. A result of this restricted public sphere was that nearly all citizens could be audience to the same debates. In a situation with full access for everyone to speak, the volume of utterances becomes too large for such a shared space of communication, and the public sphere will inevitably be more fragmented. This shows that different participatory values are at odds in the expectations of digital communication in the public sphere. An all-inclusive open public sphere, as might be realised online, perhaps can only be realised at the expense of a shared public sphere, that is, a public sphere which is visible to a significant and heterogeneous sample of citizens. The potential for democratic gain made possible by the online public sphere, may therefore also involve a loss, as extremely wide participation precludes visibility. Such fragmentation of audi-y ences is less of a loss if the different public sphere fragments are interconnected and content spills over from one arena to another. The present analysis has shown, however, that this was only to a minor extent the case with the online forums. They represented therefore a movement towards a more fragmented public sphere. However, one may question whether the ideal of one shared public sphere, formulated with reference to an historic epoch where neither mass media, nor Internet existed, is still relevant in today's communication landscape. This study shows that as a description, the Habermasian ideal does not account for the modern public spheres of the case municipalities in a precise way. However, although the actual description of existing public spheres deviates from the ideal, the ideal itself may still be normatively relevant. In order to assess the normative relevance of this ideal, one has to consider the implications of a more fragmented public sphere. IA / U Tin I I -tat/-» I /^r n f^a tjj-n Tin I I t a t/~\ i f rn^VArl I'M i K11 ri^Kn-rA if1 VAt^m/^Arl Ktt C > What will we lose and what will we gain if a shared public sphere is replaced by loosely connected sub-sphere? What should be the function of the public sphere in modern society? Questions such as these should be addressed in order to revise the normative basis of the theory of the public sphere. This study gives some indications of the position of one particular form of online Ij^ communication in the public spheres of four local communities. Obviously, it does not tell the whole story about Internet and the public sphere. First, as described in the data section, the case communities have certain characteristics that affect n the way the online forums are used and perceived by the audience. Second, both in these municipalities and elsewhere, Internet adds to the public sphere through more than merely online debates. The huge volume of information and opinions published on governmental websites, in online papers, blogs etc., probably affect both opinion formation and agenda setting. This study gives an example of how a new digital communication arena has been received and fitted into the existing landscape of local communication arenas, and how the audience perceives and deals with the tension between the traditional and the new communication arenas. To establish knowledge about how online forums or other digital communication arenas affect the public sphere in other contexts, further studies are needed. Further studies are also needed to understand the reasons for the relative importance of different digital and non-digital communication arenas, in an increasingly rich and complex communication landscape. New arenas emerge and vanish with high speed, and what may, at one point, look like the future of political communication may soon turn out to be obsolete. Notes: 1. The percentage of postings addressing local political issues varied from 20 percent in dt.no to 68 percent in nordlys.no during the period January-April 2005. 2. The interviews were conducted in Drammens Tidende (DT) (October 2004; journalists 1 to 4); Firda (October 2005; journalists 5 - 6); Stavanger Aftenblad (SA) (April 2005; journalists 7 - 9); and ) Nordlys (December 2004; journalists 10 - 12). Seven politicians were interviewed in Stavanger (April 2005; politicians 13 - 19), four in Tromsr (December 2004; 20 - 23); five in Frrde (October 2005; 25 - 29), and six in Drammen (October 2004; 30 - 35). 3. The questions posed were: How often do you visit the online forum of your local newspaper? How often do you read the reader's letters of your local newspaper: daily, weekly, monthly, less often, never? 4. The municipalities were included as dummy variables, and the municipality of Drammen was excluded from the analysis, thereby representing the reference value. Age is a continuous variable. Education and income are ordinal variables, whereas gender is a dummy, coded man=0, woman=1. Political activity is a continuous index ranging from 0 to 8, where low value indicates low political activity. The variables in the index were based on the following yes/no questions: Have you during the last four years: 1) Voted in the municipal election? 2) Participated in a protest meeting or demonstration concerning local issues? 3) Contacted local politicians about a local issue? 4) Contacted the municipal administration about a local issue? 5) Contacted the media about a local issue? 6) Signed a petition about a local issue? 7) Called a radio or television program about a local issue? 8) Written a letter to the editor? References: Arendt, Hannah. 1958. The Human Condition. Chicago: University of Chicago Press. Brandenburg, Heinz. 2006. Pathologies of the Virtual Public Sphere. In S. Oates, D. Owen and R. K. Gibson (eds), The Internet and Politics: Citizens, Voters and Activists, 207-222. New York: Routledge. Brenna, Nils. 2007. Interview with senior researcher Nils Brenna at the research unit at NRK (the Norwegian Broadcasting System) Time and place of interview??????? Coleman, Stephen. 2004. Connecting Parliament to the Public via the Internet: Two Case Studies of Online Consultations. Information, Communication and Society 7, 1, 1-22. Dahlberg, Lincoln. 2004. Net-Public Sphere Research: Beyond the "First Phase." Javnost - The Public 11, 1, 27-44. Dahlgren, Peter. 2005. The Internet, Public Spheres, and Political Communication: Dispersion and Deliberation. Political Communication 22, 2, 147-162. Di Gennaro, Corinna and William Dutton. 2006. The Internet and the Public: Online and Offline Political Participation in the United Kingdom. Parliamentary Affairs 59, 2, 299-313. Eide, Elisabeth. 2001. Mannsbastioner med iboende treghet. Medier, kjonn og makt [Inert Male Bastions. Media, Gender and Power] In M. Eide (ed.), Til Dagsorden! Journalistikk, makt og demokrati. PAGE NUMBERS???? Oslo: Gyldendal Akademisk. Fraser, Nancy. 1992. Rethinking the Public Sphere: A Contribution to the Critique of Actually Existing Democracy. In C. Calhoun (ed.), Habermas and the Public Sphere, 109-142. London: MIT Press. Habermas, Jürgen. 1989. The Structural Transformation of the Public Sphere. Cambridge: Polity. Habermas, Jürgen. 1996. Between Facts and Norms. Cambridge: Polity Press. Hoff, Jens and Kresten Storgaard, eds. 2005. Informasionsteknologi og demokratiskinnovation - borgerdeltakelse, politisk kommunikation og offentlig styring [Information Technology and Democratic Innovation - Citizen Participation, Political Communication and Public Governance]. Kobenhavn: Forlaget Samfundslitteratur. Host, Sigurd. 2005. Det lokale avismonsteret: dekningsomräder, mangfold og konkurranse 1972-2002 [The Local Newspaper Pattern: Coverage, Variety and Competition]. Fredrikstad: Institutt for Journalistikk. Janssen, Davy and Raphael Kies. 2005. Online Forums and Deliberative Democracy. Acta Politica 40, 3, 317-335. Karakaya Polat, Rabia. 2005. The Internet and Political Participation: Exploring the Explanatory Links. European Journal of Communication 20, 4, 435-59. Keane, John. 1991. The Media and Democracy. Cambridge: Polity Press. Lariscy, Ruthann Weaver, Elizabeth Johnson Avery, Kaye D. Sweetsera, and Pauline Howes. 2009. An Examination of the Role of Online Social Media in Journalists' Source Mix. Public Relation Review 35, 3, 314-316. Lee, Jae Kook. 2007. The Effect of the Internet of the Homogeneity of the Media Agenda: A Test of the Fragmentation Thesis. Journalism and Mass Communication Quarterly 84, 4, 745-760. Norris, Pippa. 2004. Building Knowledge Societies: The Renewal of Democratic Practices in Knowledge Societies. UNESCO World Report. Saglie, Jo and Tor Bjorklund. 2005. Lokalvalg og lokalt folkestyre [Local Elections and Local Government]. Oslo: Gyldendal Akademisk. Skogerbo, Eli and Marte Winsvold. 2008. Nettet som debattarena [Internet as Arena of Debate]. In E. Gunn and E. Skogerbo (eds.), Digitale Dilemmaer. Nye Medieformer,Nye Utfordrinnger [Digital Dillemas. New Media Forms], 39-60. Oslo: Gyldendal Akademisk. Smith, Aaron. 2009. The Internet's Role in Campaign 2008. Pew Internet and American Life Project. Sweetser, Kaye D., Guy J. Golan; Wayne Wanta. 2008. Intermedia Agenda Setting in Television, Advertising, and Blogs During the 2004 Election. Mass Communication and Society 11, 2, 197216. Thompson, John B. 1995. The Media and Modernity: A Social Theory of the Media. London: Polity Press. Tobiasen, Mette. 2005. Digitale skillelinjer [Digital Divides]. In L. Torpe et al. (eds.), Demokrati pä nettet, 117-147. Aalborg: Aalborg Universitetsforlag. Torpe, Lars. 2005. Digital kommunikation og menings- og politikdannelse [Digital Communication and Opinion- and Policy-Making]. In L. Torpe et al. (eds.), Demokrati pä nettet. PAGE NUMBERS????? Aalborg: Aalborg Universitetsforlag. TNS Gallup. 2005. Wellman, Barry, Anabel Quan-Haaase, James Witte and Keith Hambpton. 2001. Does the Internet Increase, Decrease or Supplement Social Capital? American Behavioural Scientist 45, 3, 436-55. Zhou, Yuqiong and Patricia Moy. 2007. Parsing Framing Processes. The Interplay Between Online