POLITICAL DISCUSSION AND NEWS USE IN THE CONTEMPORARY PUBLIC SPHERE: THE "ACCESSIBILITY" AND "TRAVERSABILITY" OF THE INTERNET Abstract This article is directed toward understanding the impact of Internet use on the accessibility of politically relevant online discourse (news and political discussion) and about the extent to which these forms of discourse are meaningfully and intimately connected. Through the use of nationally representative survey data, findings suggest: (1) when compared to offline counterparts, SES and political knowledge are equally, if not more relevant to frequent use of online news and engagement in online political discussion, suggesting that Internet use has contributed to a slightly less accessible public sphere; (2) when compared to offline counterparts, the relationship between online news and online discussion is the strongest, suggesting an especially intimate and important connection between the two forms of discourse. JENNIFER BRUNDIDGE Jennifer Brundidge is Assistant Professor at University of Texas, Austin; e-mail: jenniferbrundidge@gmail.com. (N CO n d a rN o o rN £ i 0 £ II) G C ß m KD In the last decade or so, the Internet has revolutionised the structures of the public sphere by creating a virtually unlimited number of news sites and forums of political discussion available on demand to citizens with Internet access, and by blurring and making more porous the boundaries between "news use" and "political discussion." Questions about what these changes portend for the state of democracy yield numerous and often complex answers. Yet by having the structural aspects of the public sphere brought so visibly into the forefront, the Internet may in fact help us to be more concrete about what the public sphere is, and how people engage with it. In this regard, Dahlgren (2001, 2005) points us in an interesting direction: Interaction [in the context of the public sphere] actually consists of two aspects. First, it has to do with the citizens' encounters with the media - the communicative processes of making sense, interpreting, and using the output. The second aspect of interaction is that between citizens themselves, which can include anything from two-person conversations to large meetings. To point to the interaction among citizens - whether or not it is formalised as deliberation - is to take a step into the social contexts of everyday life. Interaction has its sites and spaces, its discursive practices, its psychocultural aspects; in this sense, the public sphere has a very fluid, sprawling quality .... With the advent of the Net, civic interaction takes a major historical step by going online, and the sprawling character of the public sphere becomes all the more accentuated (Dahlgren 2005, 14). In this context, blurred and porous boundaries between "news" and "political discussion" help to create this fluid quality and an impressive "interspatiality," (Dahlgren 2001), which allow individuals to traverse seamlessly and with relative ease from one discursive space to the next. Indeed, the structural transformation of social boundaries - the increasing blurred and porous form they seem to be taking - is perhaps one of the most quint-y essential aspects of the contemporary public sphere, evoked frequently in Internet related scholarship (e.g. Bimber, Stohl and Flanagin 2005; Cammaerts and Van Audenhove 2005). Yet for the most part, vivid and compelling accounts of a struc- turally "fluid and sprawling" online public sphere with impressive "interspatiality" and weakened social boundaries have remained at an abstract theoretical level. Empirical research on online news and political discussion has instead tended to focus on particular outcomes of these processes, such as increased or decreased social capital and political participation (e.g. Bimber and Davis 2003; Shah et al. 2005), political learning (e.g. Eveland, Martin and Seo) or altered framing (e.g. ^ Zhou and Moy 2007) and selective exposure processes (e.g. Tewksbury and Althaus 2000). These studies do lend substantial credence to the importance of the mechanisms involved with online news and political discussion, but tell us little about individual-level interaction with the structural characteristics of online news and political discussion in and of itself, irrespective its more instrumental, pro-political-participation-purposes. In what was perhaps a rush by researchers to search for more instrumental "effects," some relatively basic questions about forums for online news and political discussion, as the essential "institutions" of the public sphere have been somewhat overlooked at the empirical level. 3 a o c > This may in part stem from the notion that critical and interpretive approaches are far better at capturing the more nuanced aspects of human experience. Quan- LO titative measures seem ill equipped to capture the abstract experiences of human ^ interactions within the "boundaries" of the public sphere. This notion is not unfounded. Yet an empirical, even quantitative approach could provide support for, if not perhaps a thorough test of the validity of critical theoretical claims. In this article, I therefore take an initial step toward empirically addressing the subject of individual-level interaction within the changing structural context of the "online" public sphere. I do this by first explicating two key concepts: accessibility and traversability. I argue that the structural boundaries of the Internet - the increasingly blurred and porous form they seem to be taking - are not increasing the accessibility of the public sphere, but are increasing its traversability. To examine this proposition, I generate several hypotheses related to accessibility and travers-ability and test these using nationally representative survey data. Conceptualising Accessibility and Traversability Before more explicitly relating the concepts of accessibility and traversability to the online public sphere, they may usefully be traced back to public sphere theory. Indeed, in many ways, the issues raised by the institutions of the online public sphere are old ones because the role of news media and forums of political discussion in the operation of democracy are classic concerns of public sphere theory. Most fundamentally, news media and forums of political discussion are theorised to work in tandem toward the proper formation of deliberative public opinion, with one activity solidifying the other. Indeed, "publics," according to Habermas (1991) and also Dewey (1954), exist as discursive processes: By "public sphere" we mean first of all a domain of our social life in which such a thing as public opinion can be formed. Access to the public sphere is open to all citizens. A portion of the public sphere is constituted in every conversation in which private persons come together to form a public. When the public is large, this kind of communication requires certain means of dissemination and influence; today, newspapers and periodicals, radio and television are the media of the public sphere (Habermas 1991, 398; emphasis added). Other prominent theorists make similar claims (e.g. Tocqueville 1840; Bryce 1888; Tarde 1899). Bryce (1888, 4) illustrates, albeit in a highly gendered fashion, how the deliberative process takes shape: "A business man reads in his newspaper at break-fast on the events of the preceding day. He goes down to his office in the train, talks there to two or three acquaintances, and perceives that they agree or do not agree with his own still faint impressions. ... Then debate and controversy begin." More recently, Page (1996) makes the claim that "public deliberation" can take place only with the assistance of "professional communicators," while Anderson and his colleagues suggest that, "News is what people talk about, and news makes people talk" (Anderson, Dardenne and Killenberg 1994, 37). Implicit (and in some cases explicit) in the normative arguments of these and other theorists are at least two basic claims: (1) Spaces of news and political discussion can and should be accessible to all citizens, so that "debate and controversy" can in fact begin in any meaningful or widespread way. (2) People can and should be able to traverse with relative ease from news use to political discussion - from reading the newspaper, for instance, to discussing what was read in the newspaper on a train. In short, the accessibility and what might be referred to as the traversability of the news media and forums of political discussion are normatively desirable characteristics of the public sphere. Accessibility, in this sense, involves the degree to which the structure of the public sphere may be easily penetrated. Assuming the essential existence of the news media and forums of political discussion, constructs related to accessibility would seem to be primarily governed by the porousness of the boundaries between the private and the public sphere. Traversability, on the other hand, becomes relevant only after people have accessed the public sphere and is governed by the nature of the boundaries between the news media and forums of political discussion. In this case, blurred and porous boundaries between the categories of news and discussion make travel between these discursive spaces respectively seamless and easy. This is because these boundary characteristics translate into news media and forums of political discussion that are tightly connected in time and space, allowing citizens to discuss news with others near or at the same time that they receive it. The classic public sphere seems to have been, in its way, relatively traversable, grounded as public deliberation was in world of Arts and Letters. The salons and coffee houses at this time were, according to Habermas (1989, 32-33), "centers of criticism - literary at first, then also political" and were places where "literature had to legitimate itself." In this sense, the boundaries between the discursive spaces of the classic public sphere were both blurred and porous. They were blurred because literary works were at least metaphorically speaking, both "written" and "discussed" in the coffee houses and salons - literature was discussion and discussion was literature. The boundaries between literature and discussion were porous to the extent that the very purpose of the salons and coffee houses was the discussion of literature. It was obviously then quite easy and indeed expected to broach topics y political and otherwise that were related in someway to literature. Collectively the structural boundaries of the classic public sphere seem to have provided direct mechanisms for connecting literature (including news) and discussion in time and space, allowing participants to traverse seamlessly and with ease from one form of discourse to the other. This in fact, is at the very core of the idea of the classic public sphere. On the other hand, the classic public sphere was not nearly as accessible as it was traversable, at least by contemporary standards. Though there was an active emphasis placed on the idea of accessibility, grounded in Enlightenment ideals of equality, questions remain as to how that idea bore out in reality, even among white propertied men. And, as acknowledged by Habermas himself, women and people of lower socio-economic status, were not admitted. i^-n /-» 4-1-» A 1 <-»4-i -t /-vi 4-V\ "n^i-lii rr»1 rrr\ rt i iV>11 /i r»^ 1 c > One of the chief features of the "industrial age public sphere," however, was to democratise information thereby increasing accessibility (at least to information). The mass production and distribution of newspapers made politically relevant information widely available to the public and made large-scale democracies possible (e.g. de Tocqueville 1840/1945). Even the much-maligned television, seems ^ to have had, at least initially, a democratising effect on information. As witnessed by Blumler (1970): "[television] conveys impressions of the world of politics to individuals whose access to serious coverage of current affairs is otherwise quite limited" and could "promote the development of more effective patterns of citi- ^ zenship." In fact, early research findings indicated that voters, not excluding less informed ones, became more informed through their television use (Trenamanand and McQuail 1961; Blumler and McQuail 1968). At the same time, a classic line of thought in the social sciences has been in effect that the traversability of the public sphere in the western world was problematised during the industrial age through increased urbanisation and suburbanisation, which generally resulted in anomie, the erosion of community infrastructure and available forums for civic association (Durkheim 1952; Putnam 2000). Up to a point, Marx (1844) can be placed with this group, for his concern with the structural forces of capitalism and industrialisation in separating humans from labour, nature, and political community. The Internet provides a departure from the hampered tra-versability associated with the late 19th and the 20th century - a transition from an industrial to an information age. The concepts of accessibility and traversability help generate empirical questions about individual characteristics that might make the boundaries between people and online news and forums of political discussion more or less permeable or porous (accessibility), how intimately online news and political discussion are related (traversability), and how these relationships compare with traditional news media use and "face-to-face" forums of political discussion. From a theoretical perspective, we can further look to the knowledge gap hypothesis, rational choice theory, and theories of the public sphere, for clues about answers to these questions. Accessibility: How "Public" is the Online Public Sphere? At first glance, the contemporary public sphere would seem more accessible than ever before. There are now more available news sources and forums of political discussion than ever before, most of which are immediately accessible online, which is to say they are available at any time of day, to anyone with access to the Internet, from any location where the Internet is available, and for the most part without cost (Madden 2006). The contrast on each of these counts with print newspapers, television news, and radio can hardly be overstated. Moreover, the number of these resources and the number of people taking advantage of them are on the rise. Since the creation of the World Wide Web in the early 1990s, news sites and the users of them have rapidly multiplied. By the end of 2005, nearly 50 million people in the US obtained some of their news through the Internet on an average day (Horrigan 2006). A recent 2008 Pew report finds that the proportion of Americans who report regularly learning about the presidential campaign online has doubled since 2000 (9 percent) to 2008 (24 percent). Online political discussion has also been steadily growing in prevalence. Surveys report that almost a third of Internet users regularly engage with groups online, with nearly 10 percent reporting that they engaged in online discussions about the 2004 presidential election (Rainie, Horrigan and Cornfield 2005). Yet there are structural level and individual level mechanisms that may make access to these resources more or less likely. At the structural level, there are a number of economic and technological barriers that may make access to the online public sphere less likely. First and foremost, people may not have simple physical access to the Internet, let alone access to the more politically relevant aspects of it. Assuming sufficient physical access at the structural level, which is to say that there is a viable Internet connection and regular access to a computer to connect to it, the remaining relevant structural conditions lie at the level of the online public sphere itself. The structural level of the online public sphere obviously encompasses an array of phenomena, including media ownership, political economics, and legal frameworks, a thorough discussion of which is beyond the scope of this article. But perhaps most relevant to the changing structure of the news and political discussion are structural transformation of the boundaries around and between these discursive spaces - the increasing blurred and porous form they seem to be taking. In this way, the structural realm creates a kind of "political ecology," setting the boundaries that influence the navigation of information and discussion online (Dahlgren 2005). At this level then, the increasing online options for convenient news use and political discussion, would seem to have the effect of making somewhat more porous the boundaries between the private and the public spheres, at least for those with high quality Internet access, which would then facilitate increased access to the public sphere, especially for those who might not otherwise engage in news use and political discussion. If this were the case, we would expect to find that at the individual level, standard predictors of political behaviour, including education, income, political knowledge, and self-efficacy, are less strong influences or regulators of who is engaged with the public sphere. The "public" aspect of the public sphere would be then be accentuated by the Net. Classic explanations of political behaviour, rooted in rational choice theory, would seem to point us in this direction (e.g. Downs 1957). A "rational choice" involves a form of cost benefit analysis, which may in this case, be applied to strategies involved in information/news seeking and decisions to participate in political discussion. In the case of news use, for example, if the potential costs of news (e.g. time, money, mental exertion) for certain individuals outweigh its potential y benefits (e.g. uncertainty reduction), these individuals are unlikely to seek news or engage in political discussion. Of course, the exact opposite is true, if the situation is reversed and the benefits outweigh the costs. This line of reasoning suggests that if technological developments, such as the Internet, structurally reduce the cost of news acquisition, provide more convenient and less demanding forums for political discussion, people will be more likely to engage in such activities. Most importantly, those individuals with the most to gain will be the most likely benefit from these developments (for discussion see Bimber 2003). Yet human beings are not necessarily rational creatures (Katz and Rice 2002; Neuman 1991). Any technology, and especially the Internet, is shaped not only by its rational uses, but also by human psychology, which suggests that as the cost of entry to the public sphere decreases and sources of news and forums of political discussion increase, the "public sphere rich" will actually get richer, while the "public sphere poor" will remain relatively poorer (Bimber 2003). This is the fundamental proposition of the knowledge gap hypothesis (Donohue, Tichenor and Olien 1975). The psychological basis of this proposition draws on schema theory and related research, suggesting that individuals with more complex cognitive schema ^ are better able to process and incorporate new information. Tichenor, Donohue and Olien (1970), for example, contend that education creates more sophisticated communication skills and abilities that assist individuals in processing information C^ more easily and effectively. ^ Overall, it appears that those people who used the news before the Internet are quite similar to the people who use the news on the Internet. As with users of traditional political information sources, users of online information sources tend to be white males, high in socio economic status, political efficacy, and political knowledge (Bimber 2001; 2003; Scheufele and Nisbet 2002), who have an interest in politics and who are more likely to be sceptical of information (Bimber 2003; also see Shah et al. 2005). A recent Pew report furthermore finds that people who used the Internet for news and information about the 2006 U.S. midterms elections were predominantly: white (77 percent) males (53 percent) under the age of 50 (71 percent) with a high income (over 75,000 - 44 percent), and a college degree (49 percent) (Rainie and Horrigan 2007). Political discussion, highly related to news use, should also conform to the knowledge gap hypothesis. While studies attempting to predict online discussion have been somewhat limited, findings thus far seem to be reflective of the findings on face-to-face political discussion, and generally support the knowledge gap hypothesis - that is, in terms of political discussion, the online forums do not appear to be markedly more accessible than their "face-to-face" counterparts. In a field experiment using a nationally representative panel, Price, Cappella and Nir (2002) find that individuals who participated in scheduled online discussions conformed to a hierarchical model of participation - they were older, highly educated, predominantly white, more politically knowledgeable, more politically interested and active, and had higher levels of social trust. Overall, however, in terms of accessibility, human behaviour dramatically complicates the potentially rosy picture painted by some structural aspects of the public sphere. Increased ease of entry into the public sphere is an insufficient criterion for participation in it. In spite of initial hopes, we should therefore not expect the Internet to revolutionise news use and political discussion by bringing in entirely new participants. Instead, the "public sphere rich," those high in SES and political knowledge, for example, will likely get richer. This leads to a general expectation that in terms of accessibility, data on participation in the public sphere via news consumption and political discussion will provide similar portraits of both online and offline media and forums of discussion, or more specifically: H1: Online political discussion and online news use are positively related to standard predictors of political engagement (e.g., education, income, political knowledge). However, we should expect some variance in terms of individual characteristics, including socio-demographics and political attitudes, that may help us to understand what sort of "public" has access to and is engaging in the online news use and political discussion and how that public compares to the "offline" public. This is not well understood. There is at least some evidence to suggest, for example, that online political discussion is attracting a new kind of political discussant with only some of the individual characteristics of offline discussants. Stromer-Galley (2002) finds that a need for privacy and social anxiety predicts online talk but not face-to-face talk. Two sets of findings stand out as particularly characteristic of online news use and political discussion: those on age and gender. Unlike traditional news, young people are especially inclined to use the Internet for news and information (e.g. Madden 2003; Madden and Fox 2006). In particular, findings suggest: While most citizens use the Internet to supplement other media sources, there is some evidence that suggests that there is a growing segment of the population, in particular young people (e.g. "generation X" as suggested in a early study by Shah, Kwak and Holbert 2001), for whom political engagement is most intensely connected to Internet use. Young people may then find the online public sphere more accessible than the offline public sphere. Findings suggest that women, on the other hand, are somewhat less likely to use the Internet for news and political discussion (e.g. Katz and Rice 2002), finding it less accessible than other media and forums of discussion. The usage gap between men and women has been substantially reduced as Internet technology has continued to diffuse into mainstream use. Nevertheless, a small gap remains. A recent national level survey by the Pew Research Center finds, for example, that during the 2006 midterm election, more males (34 percent of all male Internet users) got their campaign related news from online sources than did female (29 percent of all female Internet users) (Ranie and Horrigan 2007). Stromer-Galley (2002), also finds that women are somewhat less likely to access the available spaces for online political discussion. Overall, however, it is not entirely clear how relevant these demographic variables are relative to other standard predictors of political behaviour and in comparison with offline news use and discussion in various "face-to-face" forums. I therefore ask: RQ1: Do the variables predicting online news use vary from those predicting traditional news use (i.e., newspapers and television news)? RQ2: Do the variables predicting the frequency of online political discussion y vary from those predicting frequency of discussion in various "face-to-face" forums? Traversability: Characterising the Relationship between News and Political Discussion In contrast to accessibility, however, traversability (the ability to move easily or seamlessly from news to political discussion) seems an especially defining aspect of the online public sphere. Once people have crossed the individual-level and structural-level hurdles to accessing the online public sphere, important questions arise as to how the boundaries of this "new world" influence their experience of they are likely to already possess the individual characteristics (e.g. SES, political knowledge, political self-efficacy) that empower them to take advantage of whatever structural advantages the Internet holds in the way of traversability. I have argued that the more intimate the relationship between news and political discussion, the greater the traversability of the public sphere. This dynamic stems from a more general relationship between news media use and political discussion that varies in intimacy but is relatively consistent in terms of its mere existence. Indeed, an empirical relationship between news media use and political discussion has been relatively well established in research. Koch (1994), for example, finds that reading The New York Times on a daily basis causes a significant increase in political discussion. Similarly, research by McLeod and his colleagues, indicates that newspaper reading and local news viewing predict civic participation at the community level, where individuals reflect and deliberate about issues (McLeod et al. 1999). Kim, Wyatt and Katz (1999) furthermore find a relationship between newspaper reading and political conversation. Interestingly, however, they do not find a significant relationship between television news use and political conversation. This finding is in line with the notion that television is not particularly useful to the public sphere (Habermas 1989). It is not clear from these studies, however, how certain types of news media connect up to different forums of political discussion. The newspaper, for example, may lead to discussion in certain forums, whereas television news may conceivably lead to discussion at other forums not explored by these studies. Importantly, it is difficult to predict where the political discussion might take place because traditional news media seem to present no clear connection to particular forums of political discussion. It is difficult to see, for example, how a newspaper is directly connected to political discussion at the workplace (commonly invoked as an important forum for face-to-face political discussion, e.g. Scheufele et al. 2004; 2006). While it is easy to see how reading the newspaper could lead to discussion in any number of forums, a newspaper presents no obvious link to any contemporary forum of political discussion. The boundaries between newspapers and political discussion at the workplace, for example, are not blurred but easily delineated. One knows when one is reading the newspaper and when one is discussing politics at the workplace. The boundaries between these two spaces may also be less porous than those found online for at least 2 reasons: (1) Newspapers and the workplace are not aligned closely together in time and space, allowing information garnered from newspapers to be forgotten or made less meaningful in the transition; (2) The workplace is usually not a specifically designated space for political discussion, and indeed has other, more explicit purposes, which may make the transition from news to discussion a bit more challenging. This kind of lower-level of traversability may in fact be a general feature of the industrial age. Online news, as we shall see, seems to have its most obvious expression in online forums of discussion. In this sense, the online public sphere may be more akin to the classic public sphere where salons and coffee houses were ostensibly intimately linked with the world of Arts and Letters (Habermas 1989). For an understanding of just how this might operate we turn to the structure of the online public sphere. In the eighteenth and nineteenth centuries, coffee houses, salons, and newspapers were the institutions comprising the structural realm of the classic public sphere. With the intensification of the industrial age in the late 19th and 20th century, newspapers, radio, television, and perhaps volunteer associations, church, and the workplace may perhaps, be identified as institutions of the public sphere. With the advent of the Internet, the institutions of the public sphere have taken a new structural form - they may include formal news sites, video sharing sites, blogs, wikis, social networking sites, chat rooms, website bulletin boards, email, instant messaging, and so on. Within this form, the categories of news and political discussion are tightly connected in time and space and indeed, often blur into each other. Prior to the Internet, for example, mass communication and interpersonal communication could be rather clearly delineated from one another. This delineation is substantially blurred online, where communication can occur on various levels at the same time (Dahlgren 2005). Within these online discursive domains, blurred and porous boundaries between one communicative space and the next allows for increased traversibility between online news and spaces of political discussion. This may happen in a number of different ways. For example: Most news sites now enable readers to comment on certain articles, allowing people to discuss (or at least comment on) what they read in the same space that they read it. Social networking sites allow users to post links to news articles and their opinions or thoughts on political information via status updates on Twitter and Facebook, which may then be discussed among social networks through "comments" and "tweets." Often news sites even provide links to social networking sites, which allow for easier transitions from one to the other. Indeed, most transitions between news and discussion are either seamless or just a click or two away. People may therefore read about or watch the news online and in the very next moment blog about it or post a video on a video sharing site, such as YouTube, where it may in turn be discussed, or simply email someone about what they read. Occasionally, it is altogether unclear when one is using the news and when one is discussing it (e.g. blogs, tweets). Some scholars see this form of ambiguity as fundamentally postmodern due the unprecedented amount of agency given to audiences (e.g. Landow 1997; Murray 1997; Wall 2005). As Murray (1997, 128) suggests (while clearly not referring to blogs or social-networking sites, which did not yet exist), "When things are going right on the computer, we can be both the dancer and the caller of the dance. This is the feeling of agency." All in all, in the structural realm of the contemporary public sphere, online "news" and political "discussion" appear to be intimately linked and indeed, sometimes indistinguishable from one another, facilitating high levels of traversability. £ We should therefore expect to find a similar relationship between online news and online discussion to that which exists between traditional news media and offline discussion, but perhaps an even more intimate one given the virtually unlimited number of politically-relevant online spaces and the rather porous and blurred boundaries between them. That is, there should be a stronger relationship between online news and online discussion to that which exists between traditional news and various "face-to-face" forums of discussion, such as volunteer associations, commonly mentioned as an important space of political discussion. The Internet provides direct links between news and discussion, compressing them in terms of time and space, drawing them closer together. In this regard, the online public sphere may be somewhat more like the classic public sphere and less like the public sphere of the industrial age. Empirical research supports the possibility of a particularly close connection 3 a c > between online news and online discussion, though there has yet to be a thorough model comparison between online and offline news use and political discussion. According to the survey findings of Shah and his colleagues, for example, online information seeking is positively associated with interactive civic messaging, while newspaper use and television news use are non-significant (Shah et al. 2005). A n Pew Research Center survey furthermore finds that of the people who report get- ^ ting their 2006 midterm election campaign related news online, 8 percent of them report posting their own campaign related blog, 13 percent report forwarding someone else's political commentary, 1 percent report creating their own audio or video recordings, and 8 percent report forwarding some else's political audio or video recordings. All in all then, 23 percent of people who report getting their 2006 midterm election campaign related news online also report engaging in some sort of subsequent communication about it (Fallows 2007).Therefore: H2: The relationship between online news use and online discussion is stronger than the relationships between offline news use and offline forums of discussion. Online news may also provide a direct link to at least one "face-to-face" forum discussion: the workplace. This is because of the intimate connection between the Internet and many places of work. A Pew Internet Project report, for example, suggests that 57 million organisational members (62 percent of all employed workers) in the United States have Internet access (Fallows 2002). Given the central role of the Internet at many places of work, people may be reading online news and then be speaking to someone in the next cubicle or around the "water-cooler" in the very next moment. This seems less likely to be the case with a newspaper, for example. While "extra-curricular activities" such as reading the news online may be to some extent be frowned upon at some places of work, this activity is far less conspicuous than reading the newspaper at work, which would be quite awkward in almost any work setting. The relationship between online news use and political discussion should be somewhat weaker than the relationship between online news and online political discussion, however, to the extent that there is somewhat less traversability in the former context. While the boundaries between online news and political discussion at the workplace may be blurred, the boundaries may not be as porous as the online political discussion context because the workplace is not always deemed an appropriate place for political discussion. Therefore: H3: Online news use is positively related to frequency of political discussion at work but the relationship will be somewhat weaker than the relationship between online news use and online political discussion. Research Design Variables Socio-demographic Variables and Factual Political Knowledge. The age of respondents was an open-ended continuous item (M = 50.1, SD = 17.2). Sex was coded with female equal to 0 and male equal to 1 (54.8 percent female, 45.2 percent male). Education was an open-ended continuous item that asked respondents to report their total number of years of schooling (M = 14.6, SD = 3.0). Income was evaluated by asking respondents to report their total household income for the previous year (2002) by selecting from 10 categories ranging from $10,000 or less to 101,000 or more (median = $50,000 to $60,000). Political Attitudes. Several standard predictors of political behaviour were used in the analyses: Ideology (M=4.2, SD=1.4) was measured by computing the mean of two 7-point scale items. One item asked about the respondents' fiscal ideology and the other asked about social ideology, with 1 being very liberal and 7 being very conservative. The measure for ideological polarity also employed these two items. The farther along the ideology scale in either direction indicated higher polarity (M=2.5, SD=1.6). Political self-efficacy was measured by computing the mean of three items (1 = strongly agree, up to 10 = strongly disagree) that assessed respondent's beliefs about their ability to understand and influence government processes (M=5.8, SD=.93). Factual political knowledge (M = 2.6; SD = 1.2, a = .63) was an additive index of four items tapping correct identification of public figures and knowledge of current events, such as Do you happen to know what job or political office is held by Dick Cheney? Finally, social trust was assessed through one item (1 = strongly agree, up to 10 = strongly disagree), which asked about the extent to which people can generally be trusted. News Media Use (offline and online). Newspaper news use (M=6.2, SD=2.5) and television news use (M=6.0, SD=2.7) were each created by computing the mean of two items (from 0 = never to 1 = very rarely, up to 10 = all the time) that asked about attention to newspaper coverage of national and of international public affairs (newspaper news use) and attention to television coverage of national and of international public affairs (television news use). Online news use was similarly created by computing the mean of two items (from 0 = never to 1 = very rarely, up to 10 = all the time) that asked about the frequency with which respondents searched online for information on international and national issues (M=3.1, SD=3.1). Political Discussion. Offline political discussion was assessed through the use of five separate items (from 0 = never to 1 = very rarely, up to 10 = all the time) measuring the frequency of political discussion at various face-to-face forums of discussion commonly referenced to in political communication literature (e.g. Mutz gj 2006; Scheufele et al. 2004), including the workplace (M=3.6, SD=2.9), church (M=2.5, SD=2.4), non-church community/ volunteer groups (M=3.0 SD=2.5), with family (M=5.8 SD=3.0), and with neighbors (M=2.9 SD=2.4). Online political discussion (M=.82, SD=.60) 3 a o £ (fl 0 c (6 was assessed through computing the mean of two separate items (from 0 = never to 1 = very rarely, up to 10 = all the time) that asked about frequency of political discussion via chat/instant messaging and email (M=1.3 SD=1.6). Data The data used to test the hypotheses and research questions are derived from a national level survey conducted by The Cornell University Survey Research Institute in October and November of 2003, using CATI methods (N = 781). I am indebted to Dietram A. Scheufele, who was the principal investigator for the original study. The response rate was 55 percent based upon AAPOR definitions (Research definition Response Rate 3). Results I employed nine OLS regression analyses to examine the hypotheses and re-^ search questions. All hypotheses are supported by the data. Research Question 1 LD Research question 1 asked whether the variables predicting online news use and traditional news use would vary. A series of three regression analyses were run examining the predictors of online news use, newspaper news use, and television news use. Online news use was inversely predicted by age (p =-.21, p < .001) and positively predicted by education (p =.15, p < .001), sex (male) (p =.07, p < .05), political knowledge (p =.20, p < .001), political self-efficacy (p =.11, p < .01), and while falling just short of significance, income (p =.06, p < .10), newspaper news use was positively predicted by education (p =.13, p < .001), political knowledge (p =.12, p < .01), age (p =.24, p < .001) and social trust (p =.09, p < .05). Finally, television news use was positively predicted by education (p =.09, p < .05), political knowledge (p =.13, p < .01), age (p =.15, p < .001), and though falling short of significance, ideological polarity (p =.09, p < .10). Out of the three models, the one predicting online news use is the strongest, with an adjusted R2 of .18, compared with .13 for newspapers and .07 for television (See Table 1). Table 1: OLS Regressions Explaining News Use: TV, Newspaper, Online Variable TV NP Online Education .09* 13*** 15*** Sex (male) .02 -.01 .07* Income .01 .06 .06+ Age 15*** .24*** _21*** Political Knowledge .13** .12** .20*** Ideology (conservatism) .03 -.04 -.03 Ideological Polarity .09+ .05 -.003 Political Self-efficacy .01 .03 .11** Social Trust .06 .09* .05 N 692 696 696 Adj. R2 .07 .13 .18 F 7.1*** 12.8*** 18.3*** + p<.10; * p<.05; ** p<.01; *** p<.001 Note: Regression entries are standardized Beta coefficients. Research Question 2 / Hypotheses Research question 2 asked how the variables predicting offline forums of political discussion and those predicting online political discussion would differ. A set of regression analyses therefore examined the influence of several standard predictors of political engagement on frequency of political discussion online, at work, with family, at church, with neighbours, and at volunteer associations. Again the strongest model overall is that predicting the online activity. Frequency of online discussion is inversely predicted by age (p =-.24, p < .001), and positively predicted by income (p =.08, p < .05), political knowledge (p =.08, p < .05), and importantly, online news (p =.33, p < .001). Frequency of discussion at work is inversely predicted by age (p =-.28, p < .001), and positively predicted by sex (male) (p =.09, p < .05), online news (p =.08, p < .05), and while falling just short of significance, newspaper news use (p =.09, p = .06) and television news use (p =.07, p < .09). Frequency of discussion with family is inversely predicted by sex (male) (p =-.09, p < .05) and age (p =.-12, p <. 05), positively predicted by education (p =.10, p < .05), newspaper news use (p =.10, p < .05), and television news use (p =.22, p < .001). Frequency of discussion at church is positively predicted by ideology (indicating greater conservatism) (p =.16, p < .001), ideological polarity (p =.14, p < .001), newspaper news use (p =.14, p < .01), and while falling just short of significance, television news use (p =.06, p < .06). Frequency of discussion at volunteer associations is positively predicted by newspaper news use (p =.11, p < .05), television news use (p =.11, p < .05), and falling just short of significance, sex (male) (p =.07, p < .08), and inversely, income (p =-.07, p = .09). Finally, frequency of discussion with neighbours is positively predicted by newspaper news use (p =.16, p < .01) and television news use (p =.11, p < .01), and falling just short of significance, ideological polarity (p =.17, p < .06), and inversely by education (p =-08, p = .08). Among these relationships, it was hypothesised that the relationship between online news use and online political discussion would be stronger than the relationships between offline news use and offline forums of political discussion, which would suggest greater traversability. This hypothesis was confirmed, with online news use being the strongest media use predictor in general, and by far, the strongest predictor of online political discussion (p=.33, p<.001). Hypothesis 3 was also confirmed, with online news predicting discussion at work. And as further hypothesised, the relationship was rather small relative to the relationship between online news use and online political discussion (p=.08, p<.05). Notably, discussion at work and discussion online were the only discussion variables that were predicted by online news use (See Table 2). Table 2.: OLS Regressions Explaining Frequency of Political Discussion £ 3 a 0 £ * 1 * (I) 0 c B KD Variable Online Work Family Church Volunteer Neighbor Education .06 .03 .10* .04 .05 -.07+ Sex (male) -.01 .09* -.09* -.04 .07+ .04 Income .08* .03 .003 -.03 -.07+ -.01 Age -.24*** -.28*** -.12** -.01 -.05 .01 Political Knowledge .08* .03 .16*** .00 -.04 .07 Ideology (right) -.05 -.01 .04 .16*** -.02 .02 Ideological Polarity -.03 .04 .07+ .14*** .06+ .07+ Self-efficacy -.003 .05 .05 -.02 .03 -.05 Social Trust .05 .02 .001 -.01 .06 -.004 Online news .33*** .08* .004 .03 -.001 -.02 Newspaper news .06 .09+ .10* .14** .11* .16** TV news -.02 .07+ .22*** .08+ .11* .12** N 692 681 685 683 680 687 Adj. R2 .26 .13 .16 .08 .04 .06 F 21.6*** 9.1*** 11.7*** 6.2*** 3.3*** 4.8*** + p<.10 * p<.05; ** p<.01; *** p<.001 Note: Regression entries are standardized Beta coefficients. Discussion The purpose of this article was to examine the accessibility and traversability of the contemporary public sphere. In terms of accessibility, it appears the typically "public sphere rich" are in fact getting richer. As this applies to news use in particular, the overall model of online news, which includes most standard predictors of political behaviour, is the strongest when compared with offline news counterparts. This means that the socio-economic and cognitive hurdles are actually greater for accessing the online public sphere than accessing the offline public sphere. Why is online news less accessible than offline news? One possible reason is that "searching" for internationally and nationally relevant news online may be a more cognitively engaged activity than simply attending to such information as it appears in the newspaper or especially on television. Engaging in online searches for news requires a cognitive schema relating to the kinds of information one wishes to acquire. This may furthermore require a higher level of education, which according to Nie, Junn, and Stehlik-Barry (1996), develops in citizens the cognitive skills required for political engagement and political knowledge creation. Political self-efficacy may furthermore endow citizens with the expectation that their online searches will indeed be fruitful. Conversely, people with low political self-efficacy may not have the same expectation and may therefore be less likely to engage. Still, online news use may actually be more accessible for at least one typically less politically engaged group: Younger people seem to find online news more accessible than traditional news, perhaps, as has been suggested, because it fits more readily into their already high levels of Internet use. Overall, however, given the socio-demographic and cognitive hurdles citizens must overcome when searching for online news, online news seems somewhat less accessible than many observers of the Net had previously imagined. Online political discussion, on the other hand, does appear to be somewhat more accessible than online news, but not substantially more accessible than offline forums of political discussion. Unlike online news use, education is not a significant predictor. Furthermore, in spite of their relatively equal accessibility overall, one interesting difference between online discussion and many forums of "face-to-face" discussion, is that sex is not a significant predictor of online discussion. This stands in contrast to the earlier findings of Stromer-Galley (2002) and is consistent with the possibility that online discussion is more gender neutral than "face-to-face" political discussion (see Flanagin et al. 2002). In terms of traversability, I proposed that blurred and porous boundaries between online news and online discussion create a more intimate relationship between the two than seen in traditional media domains, by connecting news and discussion in both time and space as citizens move seamlessly and with relative ease between various categories of discourse (Dahlgren 2001; 2005). The results generally support this proposition. Of all the news use and political discussion variables, online news use and online discussion do indeed have the strongest relationship, suggesting that they are more intimately linked, more bound in time and space. The only other forum of discussion to be significantly predicted by online news use is discussion at work, though somewhat less so. Yet, the "Industrial Age" public sphere may not be as low in traversability as originally theorised in this article. The strong relationship between television news use and political discussion with the family was unanticipated, but clearly makes sense - unlike the reading of newspapers, television watching is done almost exclusively in the home and is very often a collective activity. Unlike neighbours or volunteer association members, for example, families may commonly watch the news on television together and then discuss what they see as they see it, which has the effect of linking news and political discussion in time and space. Television news and discussion about politics within the family may actually be a fairly traversable discursive environment, though somewhat less so than the online environment. Certainly, with the addition of Internet to the Industrial Age options, we live in an increasingly traversable media environment. It should be noted that this study is not without important limitations. It is first unclear that online news use directly leads to or causes online political discussion. Though the two are highly correlated, while controlling for a number of theoretically relevant variables, there are no questions in the survey that ask respondents if their discussion online was a direct result of their online news use. It may then be that the two variables are simply highly related for other reasons than those related to the "traversability" between them. The argument for traversability does, however, become more compelling when comparing the strength of the relationship between online news and online discussion and the relative weakness of the relationships between traditional news use and "face-to-face" forums of political discussion. Second, the data may be criticised on the grounds that they are too old and that the Internet has evolved greatly beyond what it was in 2003. Indeed, the number of people who actually report discussing politics online in this survey is quite low. Yet I would argue that the theoretical arguments put forth in this study have actually been strengthened with recent online developments, suggesting that the relationships found in this study would actually be stronger today than in 2003. In order to more fully examine blurred and porous boundaries in the context of traversability, future research should provide more direct tests for the specific mechanisms that lie between news use and political discussion. Such tests might y include experimental research aimed at examining the precise micro-level-processes involved in traversing the online public sphere in combination with survey research employing more direct questions about the ways in which individuals navigate their online experience. Moreover, interpretive research might examine the complexity likely involved with people's experience of accessibility and traversability, helping to answer the question of what this really means for people and democratic life. The theoretical constructs and research findings presented in this article provide a starting point for this potential line of research. Overall, the findings suggest that the online public sphere is somewhat less or at best, equally accessible, but also substantially more traversable for those with the ability, skill, and motivation to access it. In the case of accessibility, the presumably more porous boundaries between the private and the public sphere created c > by the structure of the online public sphere do not seem to be enough to counteract strong forces embodied by the knowledge gap hypothesis. The blurred and porous boundaries between online news and online discussion do, on the other hand, appear to be enhancing the ease with which people transition from news to political discussion - potentially connecting news and political discussion in new and powerful ways, conducive to public opinion development. At least in terms of the criteria discussed in this article, the online public sphere seems not remotely akin to the questionably accessible but highly traversable "clas- C^ sic" bourgeois public sphere, but somewhat further removed than the somewhat ^ more accessible but less traversable "industrial age" public sphere. As noted by Papacharissi (2002, 21) both are relatively low in accessibility: "This virtual sphere is dominated by bourgeois computer holders, much like the one traced by Habermas consisting of bourgeois property holders." Both are also comparatively high in traversability - created by distinctly configured porous and blurred boundaries between information and discussion. Clearly, traversability and accessibility are not the only normative requirements for a healthy public sphere. Even when these requirements for the institutions of the public sphere (news and forums of political discussion) are met, the normative quality of the discourse that takes place within the structures may still be relatively low. One might even argue that the kind of traversability experienced online, absent of any commitment to high quality deliberation, may in fact lead to less civil, off the cuff, less processed, political talk, or less of what we might traditionally call discussion. Posting a comment on somebody's facebook page would seem to be a far cry from the high minded debate that ostensibly took place in the coffee houses and salons of classic public sphere. Nevertheless, I would suggest that accessibility and traversability are necessary while not sufficient criteria for a healthy public sphere. A public sphere with increased traversability and limited accessibility, suggests a newly dynamic communication environment for a certain segment of the citizenry. References: Anderson, Rob, Robert Dardenne and George M. Killenberg. 1994. The Conversation of Journalism: Communication, Community and News. Westport: Praeger. Bimber, Bruce. 2001. Information and Political Engagement in America: The Search for Effects of Information Technology at the Individual Level. Political Research Quarterly 54, 1 53-67. Bimber, Bruce. 2003. Information and American Democracy: Technology in the Evolution of Political Power. Cambridge: Cambridge University Press. Bimber, Bruce and Richard Davis. 2003. Campaigning Online: The Internet in U.S. Elections. New York: Oxford University. Bimber, Bruce, Andrew J. Flanagin and Cynthia Stohl. 2005. Reconceptualizing Collective Action in the Contemporary Media Environment. Communication Theory 15, 4, 365-388. Blumler, Jay G. 1970. The Effects of Political Television. In J. D. Halloran (ed.), The Effects of Television. London: Panther. Blumler, Jay G. and Dennis McQuail. 1968. Television in Politics: Its Uses and Influence. London: Faber and Faber. Blumler, Jay and Elihu Katz. 1974. The Uses of Mass Communications. Beverly Hills: Sage Publications. Bryce, James. 1888/1973. The American Commonwealth (Vol. 3). London: Macmillan. Cappella, Joseph N., Vincent Price, and Lilach Nir. 2002. Argument Repertoire as a Reliable and Valid Measure of Opinion Quality: Electronic Dialogue During Campaign 2000. Political Communication 19, 1, 73-93. Cammaerts, Bart and Leo Van Audenhove. 2005. Online Debate, Unbounded Citizenship, and the Problematic Nature of a Transnational Public Sphere. Political Communication 22, 2, 179-196. Dahlgren, Peter. 2001. The Public Sphere and the Net: Structure, Space, and Communication. In W. L. Bennett and R. M. Entman (eds.), Mediated Politics: Communication in the Future of Democracy, 33-55. Cambridge: Cambridge University Press. Dahlgren, Peter. 2005. The Internet, Public Spheres, and Political Communication: Dispersion and Deliberation. Political Communication 22, 2, 147-162. Dewey, John. 1954. The Public and Its Problems. Chicago: Swallow. Donohue, George A., Phillip J. Tichenor, and Clarice N. Olien. 1975. Mass Media and the Knowledge Gap. Communication Research 2, 1, 3-23. Downs, Anthony. 1957. An Economic Theory of Democracy. New York: Harper and Row. Dürkheim, Emile. 1974. On Anomie. In W. E. Connolly and G. Gordon (eds.), Social Structure and Political Theory 75-96. Lexington: D.C. Heath. Eveland, William P. Jr., Krisztina Marton and Mihye Seo. 2004. Moving Beyond "Just the Facts:" The Influence of Online News on the Content and Structure of Public Affairs Knowledge. Communication Research 31, 1, 359-386. Fallows, Deborah. 2002. Email at Work. Pew Internet and American Life Project. . Fallows, Deborah. 2007. Election News Hounds Speak Up: Newspaper, TV, and Internet Fans Tell How and Why They Differ. Pew Internet and American Life Project. Flanagin, Andrew J., Vanessa Tiyaamornwong, Joan O'Connor and David R. Seibold. 2002. ComputerMediated Group Work: The Interaction of Member Sex and Anonymity. Communication Research 29, 1 66-93. Habermas, Jürgen. 1989. The Structural Transformation of the Public Sphere. Cambridge: MIT. Habermas, Jürgen. 1991. The Public Sphere. In C. Mukerji and M. Schudson (eds.), Rethinking Popular Culture: Contemporary Perspectives in Cultural Studies, 398-404. Berkeley: University of California. Horrigan, John, Kelly Garrett, and Paul Resnick. 2004. The Internet and Democratic Debate. Pew Internet and American Life Project. Katz, Elihu, Jay G. Blumler, and Michael Gurevitch. 1974. Utilization of Mass Communication by the Individual. In J. G. Blumler and E. Katz (eds.), The Uses of Mass Communications: Current Perspectives on Gratifications Research, 19-32. Beverly Hills: Sage. Katz, Elihu, and Ronald E. Rice. 2002. Social Consequences of Internet Use: Access, Involvement and Interaction. Cambridge: The MIT Press. Kim, Joohan, Robert O. Wyatt, and Elihu Katz. 1999. News, Talk, Opinion, and Participation: The Part Played by Conversation in Deliberative Democracy. Political Communication, 16, 4, 361-385. Koch, Tom. 1991. Journalism for the 21st Century: Online Information, Electronic Databases, and the News. New York: Greenwood. Landow, George. 1997. Hypertext 2.0: The Convergence of Contemporary Critical Theory and Technology. Baltimore: Johns Hopkins. Madden, Mary. 2006. Internet Penetration and Impact. Pew Internet and American Life Project. Madden, Mary, Susannah Fox. 2006. Riding the Waves of "Web 2.0." Pew Internet and American Life Project. Marx, Karl. 1844/1974. On Alienation. In W. E. Connolly and G. Gordon (eds.), Social Structure and Political Theory, 97-109. Boston: D.C. Heath. £ McLeod, Jack M., Dietram A. Scheufele, Patricia Moy, Edward M. Horowitz, R. Lance Holbert, Weiwu Zhang, Stephen Zubric and Jessica Zubric. 1999. Understanding Deliberation: The Effects of Discussion Networks on Participation in a Public Forum. Communication Research 26, 6, 623-654. Murray, Janet H. 1997. Hamlet On the Holodeck: The Future of Narrative in Cyberspace. New York: Free Press. Mutz, Diana C. 2006. Hearing the Other Side: Deliberative Versus Participatory Democracy. New York: Cambridge University. Nie, Norman H., Jane Junn and Kenneth Stehlik-Barry. 1996. Education and Democratic Citizenship in America. Chicago: University of Chicago. Neuman, Russell W. 1991. The Future of the Mass Audience. Cambridge: Cambridge University. Page, Benjamin I. 1996. Who Deliberates? Mass Media in Modern Democracy. Chicago: University of Chicago. £ 3 a (D o c (B o 00 Papacharissi, Zizi. 2002. The Virtual Sphere: The Internet as a Public Sphere. New Media and Society 4, 1, 9-27. Price, Vincent, and Joseph N. Cappella. 2002. Online Deliberation and Its Influence: The Electronic 00 Dialogue Project in Campaign 2000. IT and Society 1, 1, 303-329. Price, Vincent, Joseph N. Cappella and Lilach Nir. 2002. Does More Disagreement Contribute to More Deliberative Opinion? Political Communication 19, 1, 95-112. Putnam, Robert D. 2000. Bowling Alone: The Collapse and Revival of American Community. New York: Simon and Schuster. Rainie, Lee, John Horrigan, Michael Cornfield. 2005. The Internet and Campaign 2004. Pew Internet and American Life Project. Scheufele, Dietram A., Bruce W. Hardy, Dominique Brossard, Israel S. Waismel-Manor and Erik C. Nisbet. 2006. Democracy Based on Difference: Examining the Links Between Structural Heterogeneity, Heterogeneity of Discussion Networks, and Democratic Citizenship. Journal of Communication 56, 4, 728-53. Scheufele, Dietram A. and Matthew C. Nisbet. 2002. Being a Citizen Online: New Opportunities and Dead Ends. Press/Politics, 7, 3, 55-75. Scheufele, Dietram A., Matthew C. Nisbet, Dominique Brossard, and Erik C. Nisbet. 2004. Social Structure and Citizenship: Examining the Impact of Social Setting, Network Heterogeneity, and Informational Variables on Political Participation. Political Communication 21, 3, 315-338. Shah, Dhavan V., Jaeho Cho, William P. Eveland and Nojin Kwak. 2005. Information and Expression in a Digital Age. Communication Research 32, 5, 531-565. Shah, Dhavan V., Nojin Kwak and Lance Holbert. 2001. Connecting and Disconnecting with Civic Life: Patterns of Internet Use and the Production of Social Capital. Political Communication 18, 2, 141-162. Stromer-Galley, Jennifer. 2002. New Voices in the Public Sphere: A Comparative Analysis of Interpersonal and Online Political Talk. Javnost- The Public 9,2, 23-42. Tarde, Gabriel. 1899/1989. Opinion and Conversation (J. Ruth, Trans.). Unpublished Translation of "L'opinion et la conversation," from G. Tarde, L'opinion et la foule. Paris: Presses Universitaires de France. Tewksbury, David and Scott L. Althaus. 2000. Differences in Knowledge Acquisition among Readers of the Paper and Online Versions of a National Newspaper. Journalism and Mass Communication Quarterly 77, 3, 457-479. Tichenor, Philip J., George A. Donohue, and Clarence N. Olien. 1970. Mass Media Flow and Differential Growth in Media Knowledge. The Public Opinion Quarterly 34, 2, 159-170. de Tocqueville, Alexis. 1840/1945. Democracy in America, vol. 2. New York: Vintage Books. Trenaman, Joseph and Dennis McQuail. 1961. Television and the Political Image. London: Methuen. Wall, Melissa. 2005. Blogs of War: Weblogs as News. Journalism: Theory, Practice and Criticism 6, 2, 153-172. Zhou, Yuqiong and Patricia Moy. 2007. Parsing Framing Processes: The Interplay Between Online Public Opinion and Media Coverage. Journal of Communication 57, 1 79-98. Zillman, Dolf, and Jennings Bryant, eds. 1985. Selective Exposure to Communication. Hillsdale: Erlbaum.