Political discussion,opinion leadership and trust

Personal discussion networks: Trusting one another as sources of political information?

Recent research has offered new evidence to the thesis that, despite the important and impressive presence of mass media, personal discussion still shapes the formation of political opinion. Scholars have recently focused their analyses on how the social context, particularly personal networks, are translated into a source of political influence (Huckfeldt and Sprague, 1995; Huckfeldt et al., 2004; Zuckerman, 2005). In particular, findings support Katz and Lazarsfeld’s (1955) old hypothesis that citizens depend on one another for information and advice on political matters. But what is the relationship in terms of trust among individuals belonging to the same discussion network? Might personal ties between discussion partners be an effective means of validating political information that is processed during interpersonal conversations?

First of all, it should be stressed that a great deal of political conversation usually takes place in contexts that are informal and familiar in the sense of involving people who know each other. According to research findings, family and friends have a remarkable effect on the formation of political opinions (Kotler-Berkowitz, 2005; Stoker and Jennings, 2005; Straits, 1991; Verba et al., 2005). As Zuckerman (2005) argues, this body of studies consists of a return to ‘the social logic of politics’. After several decades of analyses based on an independent and isolated citizen who had been abstracted from his/her surroundings, research on discussion networks rediscovered the role of social circles as originally stated in the 1950s by pioneers of voting studies such as Paul Lazarsfeld and his colleagues at Columbia University. As stressed by Huckfeldt and Sprague (1995: 23), ‘citizens are interdependent – that they obtain important political information through process of social communication’.

The choice of familiar discussion partners, whose political views are probably well known, is supposed to be induced not only by their proximity and direct accessibility, but also by the desire to avoid potential and unintentional conflicts and tensions within one’s own social circle. As both Noelle-Neumann (1984) and McKuen (1990) have argued, people may prefer to remain silent in a context where the majority’s views contrast with their own opinions. Despite this tendency of selecting like-minded discussants, however, if we look at the general picture, it has been found that disagreement is relatively widespread in political discussion networks (Beck, 1991; Huckfeldt et al., 2004; Mutz and Martin, 2001). This is due to the characteristics of those discussion networks that may be regarded as ‘low-density networks’ in the sense that ‘communication networks among associated individuals are overlapping but less likely to be identical . . . even though two individuals discuss politics with one another on a frequent and recurrent basis, they may still be located in communication networks that are quite distinct. . . . Hence, discussants may usefully be conceived as webs of both reciprocal and non-reciprocal informant relationships’ (Huckfeldt et al., 2004: 208–209). This does not imply the absence of political influence: on the contrary, the persuasiveness of communication that occurs between discussion partners sustains political disagreement just because networks are not cohesive, but rather overlapping with a variety of gaps and holes. As a consequence, ‘citizens evaluate each information source in the context of every other information source’ (Huckfeldt et al., 2004: 210) and ‘judge each new piece of information within the context of the information they have previously obtained’ (Huckfeldt et al., 2005: 31).

Findings about the degree of circulation of ideas in our society are considered in a very positive light since cross-cutting networks reinforce the kind of political dialogue that is needed to maintain political tolerance and democratic legitimacy. Mutz (2002) found that exposure to cross-cutting increases citizens’ awareness of legitimate rationales for oppositional views. The other side of the coin, however, is the need to give up very stringent requirements about the degree of intimacy and trust. ¹ There exists a trade-off since ‘dense networks of tight-knit social relationships and their characteristic high levels of trust may come only at the expense of exposure to cross-cutting views’ (Mutz, 2002: 123). Given the degree of heterogeneity of multicultural societies, low-density networks seem to be the only antidote to the insidious phenomenon of ‘balkanization’, in which people tend to live in separate environments with little interaction across lines of religion, race and social class.

A second important finding of the body of research on interpersonal communication is the strong correlation between media exposure and political discussion. The more a person collects media information the more he/she talks about politics (Campus et al., 2010; Kim et al., 1999). Direct comparisons between mass media and political discussion have shown that the latter filters media messages (Schmitt-Beck, 2003). Citizens embedded in homogeneous networks, whose members share the same political opinions, appeared more influenced than those who discuss politics with people holding different and divergent political positions (Beck et al., 2002; Campus et al., 2008). The main contribution of these studies is to re-emphasize the role of social groups, especially personal networks, in influencing the opinion formation of individuals. As Beck et al. (2002) put it, voting is a ‘social calculus’ where all intermediaries – media, social organizations and personal discussants – are ‘carriers’ of partisan and ideological messages. And, in contrast to the thesis of the massive influence of television, it may happen that discussion networks are more powerful carriers than mainstream media.

However, it is worth noting that the majority of the above-mentioned studies focus on the relationship between political discussion and exposure to traditional media sources, such as newspapers and TV news watching. One may wonder if the ongoing change in the content of media communication has been exerting an influence on political discussion habits as well. In fact, the advent of the popularization of politics has changed the scenario and made the public sphere more available to a larger group of citizens (Dahlgren, 2009; Van Zoonen, 2005). A number of studies have highlighted interesting relationships between watching talk-shows or reality shows and the nature of political discussion (Coleman, 2003; Mazzoleni and Sfardini, 2009). Further explorations can be made with a view to ascertain whether the popularization of politics fosters political discussion. In particular, being that political information is more clearly based on personal clues than on ideological issues, the popularization of politics could increase the ‘vulnerability’ of homogeneous networks. A possible implication could be an increase of reciprocal tolerance, since a discussion may be initiated before the ideological viewpoints of discussants are fully revealed.

Still time for opinion leaders?

In this literature, a notable difference with the traditional theory of the two-step flow of communication has emerged. This concerns the existence of the so-called opinion leaders. According to Katz and Lazarsfeld’s seminal book, Personal Influence, in each community there exist opinion leaders who are likely to expose themselves to mass media and to influence other persons in their immediate environment (2006 [1955]: 3). It is important to say that ‘opinion leaders’ are not leaders in a formal sense, i.e. not heads of government and political parties or business tycoons or media stars. They are ordinary people that happen to be regarded as competent and reliable within their own communities. In this sense, they are also different from ‘political entrepreneurs’ in Robert Dahl’s (1961) definition. In fact, political entrepreneurs also seek influence but with the aim of gaining resources to be used for a specific political project. This aspect does not necessarily apply to opinion leaders.

To identify opinion leaders in the political field, Katz and Lazarsfeld used a complex follow-up procedure: they asked their sample to name an expert ‘who keeps up with the news and whom you can trust to let you know what is really going on’ and then interviewed persons named as experts by the sample and then those named by the first expert group. In so doing, they found that opinion leaders in public affairs not only had a reputation for expertise but were also more inclined to acknowledge their own influence on other people. One of the most interesting findings was that the sphere of politics appeared different from the other fields investigated. In other contexts, in fact, leadership was horizontal, since people chose their reference points within their own social group and among their peers. As for political opinion leadership, by contrast, ‘the flow of influence in this arena crosses status boundaries, travelling from upper status to people beneath them in status’ (Katz and Lazarsfeld, 2006 [1955]: 331). Better educated and wealthier people tended to have more chance of participating and being informed and, consequently, were more likely to be regarded as experts. Therefore, it could be argued that, despite not being linked to particular social positions and roles, becoming opinion leaders was not independent from socioeconomic status.

Several decades later, scholars who have analysed political discussion in modern electoral campaigns have been unable to identify those opinion leaders who, according to Katz and Lazarsfeld, should be able to influence the less politically active people by virtue of the trust they inspire (Campus and Pasquino, 2007; Huckfeldt and Sprague, 1995). Therefore, although it may be true that political discussion, especially within homogeneous networks, reinforces consistent individual opinions and challenges contrasting views (Beck et al., 2002; Schmitt-Beck, 2003), it is unclear whether the process of validation of information depends on evaluating in a special way the expertise and wisdom of specific discussion partners. Perhaps, one should consider that in the individuals’ agenda of social contacts there still exist some relationships in which some people provide others with the interpretation of political information obtained from mass media. In other terms, there may still be political experts that invest time and effort in acquiring political knowledge and occupy more influential roles in the communication of political information within their micro-environments. However, the ‘low-density’ of current discussion networks, that is to say the fact that two discussion partners are often embedded in different and sometimes politically divergent networks of political communication (Huckfeldt et al., 2004), makes it more difficult to identify stable opinion leaders.

This holds in general for interpersonal communication. With respect to the traditional discussion on the matter, one should also consider that the framework today is more complex because of the emergence of online political discussion. As Chadwick (2006: 25) says, ‘much of what goes on in cyberspace is talk. Hundreds of thousands of forums have sprung up, in which people in their diverse identities can argue, compete, collaborate, or simply share thoughts.’ In some ways it could be argued that the advent of new media has redirected the focus of attention on how much people depend on others to form their own political opinions. According to Dahlgren (2009: 73), ‘interaction between citizens’ is ‘crucial’. ‘While this interaction perspective was prominent in the older sociology of the media (I am thinking here of the “two-step-flow” tradition: see e.g. Katz and Lazarsfeld, 2006 [1955]), it is sometimes ignored in discussion about the contemporary public sphere (and very often in regard to public opinion), even if such a focus is now returning in the context of the increasingly mediated civic interaction taking place via the Internet and other forms of interactive digital communication.’

The literature comparing political discussion in real-world interpersonal networks and Internet community networks has highlighted both similarities and differences. The main element of convergence is that, in both contexts, a two-way flow of information prevails. In contrast to the typical one-way communication of mainstream media (Thompson, 1995), Internet users are partners in a reciprocal process of communicative exchanges through social networking and online discussion forums. On the other hand, it should be stressed that the strength of ties makes an important difference with face-to-face interactions. ² Even if friends and relatives may use the Internet to connect to each other, online ties with discussion partners, especially on blogs, forums, chat rooms, etc., are supposed to be weaker than those within interpersonal networks, which are mostly composed of family members, friends and workmates. Such a difference begs the question of whether the web offers special and/or alternative ways of sustaining networks. One may wonder whether weak ties among discussion partners are more an incentive rather than an obstacle to political talk. According to Granovetter’s (1973) classic thesis, weak ties are actually necessary to individuals’ opportunities and to the diffusion of information. But does the online environment have a different impact on the relationship between the strength of ties and the nature of political discussion? For instance, one of the peculiarities of the web is its potentiality for initiating ties among people who do not know each other well. This implies key benefits for the political discussion: ‘Initiating such weak ties then has a positive impact by broadening an individual’s knowledge base, exposing them to ideas and approaches different from their own, and increasing their ability to recognize and take advantage of new opportunities’ (Haythornthwaite, 2002: 388).

The issue of the strength of ties between discussants offline and online also deserves further attention in relation to the nature and scope of opinion leaders. While in face-to-face communication influential people have to be, if not intimate, at least proximate, on the net new kinds of people could potentially take up the role of opinion leaders. On the other hand, even if the majority of Internet discussion, in particular that taking place on blogs and forum, is based on personal experience, this does not necessarily imply an authentic communicative exchange among people. According to Castells (2007: 247), a large number of bloggers confess they blog for themselves rather than for their audience. ‘Thus, to some extent, a good share of this form of mass self-communication is closer to “electronic autism” than to actual communication. Yet, any post in the Internet, regardless of the intention of its author, becomes a bottle drifting in the ocean of global communication, a message susceptible of being received and reprocessed in unexpected ways.’ In such a volatile context, it is less likely that those intense and regular interactions, which are required to exert an effective influence, will be developed.

In an environment of this sort, Manuel Castells’ reflection points to the impossibility of the emergence of opinion leaders in a traditional sense. According to him, the peculiar power relationship of the ‘network society’ allows the existence of ‘switchers’ who control ‘the connecting points between various strategic networks. . . . For instance, power evolves out of the switcher’s ability to connect political leadership networks, media networks, scientific and technology networks and military and security networks to achieve a geopolitical strategy. Switchers may also advance a religious agenda in a secular society by solidifying relationships between religious and political networks; or they may link academic and business networks through facilitating the exchange of knowledge and legitimation of financial sponsorship in order to further an intellectual and/or economic agenda’ (Arsenault and Castells, 2008: 490).

Although switchers may correspond to a single individual with a special role, like Rupert Murdoch (at least until his recent partial fall from grace) – who controls connection points between business, media and economic networks – usually they are not to be viewed as a specific person or a specific group. ‘This is not an old boys’ network’ or ‘the idea of a power elite’. Switchers are rather ‘networks of actors engaging in dynamic interfaces that operate specifically in each particular connection process’ (Arsenault and Castells, 2008: 491). ‘The power of the switch is ultimately at the service of the goals that are programmed into the networks. But in a world of multiple power networks, it is the switcher that facilitates the performance of the programmes’ (2008: 509).

Are contemporary opinion leaders equivalent to Castells’ notion of switchers? Or could it be argued that there are other types of online (and offline) opinion leaders that deserve a closer look? As we will see in the next section, not all are in agreement with Castells. For instance, the blogosphere is regarded as a new potential reservoir of opinion leaders. More generally, both the studies on face-to-face political discussion and the research on mechanisms of political influence in the new media, point to the revision of the traditional notion of opinion leadership in search of a definition that may better accommodate the growing volatility and interchangeability of low-density discussion networks. To do so, further empirical research is needed. Some phenomena have already been the subject of a number of studies, such as bloggers. Might bloggers be the ‘new opinion leaders’? Why do some of them become influential and trusted? In the following section, we look at how scholars have tried to answer these questions.

Bloggers: A new breed of opinion leaders?

In the light of the previous argument about the need for identifying and analysing possible new kinds of influential people, we deal with an instance of new opinion leaders which is among the most investigated and discussed: bloggers. The issue of whether bloggers can be considered as opinion leaders in Katz and Lazarsfeld’s sense is controversial. Watts and Dodds (2007: 447), for instance, give a negative answer: opinion leaders are supposed to ‘exert interpersonal influence’, while ‘the influence of the blogger seems closer to that of a traditional newspaper columnist or professional critic than to that of a trusted confidant or even a casual acquaintance’. Of a different opinion are Song et al. (2007: 971) who rather observe: ‘Social influence, which describes the phenomenon by which the behaviour of an individual can directly or indirectly affect the thoughts, feelings, and actions of others in a population, is present in the conversations in the blogosphere. . . . Opinion leaders absorb information through various means and then transmit this information to certain opinion receivers. The important role of opinion leaders has attracted growing attention recently since massive quantities of network data are available through the Internet.’ Along the same line, Kavanaugh et al. (2006) argue in favour of a link between offline and online opinion leaders by showing how blogs are regarded by social and politically active people as an additional way of exerting informal influence on members of their social networks. The debate is, therefore, open and clearly the distinction between influence exerted through face-to-face communication and some forms of online influence, such as that exerted via blogs, social networking, forums, etc. deserves further analysis.

Focusing on political influence in particular, the Institute of Politics, Democracy and Internet has published a study on ‘poli-influentials’, that is to say people interested in politics who exert ‘their influence as outspoken leaders among their families, friends, neighbours and colleagues’ (Darr, 2007: 1). In terms of political influence, the study underlines the occurrence of a bigger change: with the advent of digital networking tools, those influential people ‘may transcend geographical barriers, helping local political opinion leaders achieve the reach of national opinion leaders’ (Rosenblatt, 2007: 13). Bloggers’ influence appears to be of particular interest: many people now turn to them for their political news, which often includes interesting stories ignored by mainstream media. But, in particular, it has to be stressed that a large number of users take political actions upon reading them, for instance make a political contribution after the suggestion of a blog. ‘Clearly blogs can have great influence on this behaviour’ (Schlacter, 2007: 20). Therefore, bloggers may be regarded as a new incarnation of those opinion leaders who are traditionally courted by politicians as ‘a force multiplier by swaying the opinion of their peers’, people who are ‘not obvious community leaders, but who do, however, influence their peers’ decisions and behaviours’ (Spielthenner et al., 2007: 27).

The power structure in the blogosphere has recently received empirical attention. Scholars like David Karpf ³ have tried to measure the authoritativeness of political blogs. Recent studies advance some interesting hypotheses about what kinds of blogs create political information and which act as followers (Barzilai-Nahon et al., 2010). A relevant aspect is the relationship between mass media and bloggers: as Drezner and Farrell (2008) found, media read blogs and transmit their information to the political elites. In general terms, it is well known that ‘mainstream media are using blogs and interactive networks to distribute their content and interact with their audience, mixing vertical and horizontal communication modes’ (Castells, 2007: 247). Media ‘relentlessly scan the blogosphere to select themes and issues of potential interest for their audience . . . and political elites across the entire political spectrum increasingly use the ways and means of mass self-communication’ (Castells, 2007: 252). Therefore, it becomes crucial to consider the interconnection between political figures and media personalities, whose ‘influence is exerted indirectly through media or authority structures’ and opinion leaders as defined by Katz and Lazarsfeld, whose influence is direct and derives from their informal status as individuals who are highly informed, respected, or simply ‘connected’ (Watts and Dodds, 2007: 442). In the case of bloggers and social networking, the line of division between the two groups may be blurred. In Italy, for instance, research has identified the 35 most influential opinion leaders on Facebook: in the top 10 are Nichi Vendola, an emerging leftist leader, and Beppe Grillo, a former comedian who has latterly become a political leader by launching a very popular blog, and many famous journalists and columnists. ⁴ In this case, and presumably in many others, one may detect a certain degree of overlap between different roles and notions of opinion leadership.

In conclusion, recent developments in political communication, including the popularization of politics and the introduction of new technologies, lead inevitably to questions about the meaning and the purposes of opinion leadership in contemporary democracies. As the discussion on bloggers as possible opinion leaders has shown, there is a notable divergence in the notion of opinion leadership between different scholars, a clear indicator of the need for clarification and further conceptualization of the concept of opinion leadership. What clearly emerges from the above discussion is that Katz and Lazarsfeld’s notion of opinion leader has to be somehow revised and integrated. As argued above, the new characteristics of the political context, and in particular of contemporary political discussion networks, suggest that the latest ones are no more the cohesive groups of the past, but should rather be conceived as webs of relationships that may also be loose and volatile. Often, even people talking frequently to each other, sometimes trusting each other, are not influenced by reciprocal and recurrent interactions, because they are embedded in highly divergent informational contexts which provide resources to reduce vulnerability to other people’s views. In such a framework, to think of opinion leaders as people who are more interested and informed, who acquire and maintain credibility and trust over time in a stable circle of relatives and acquaintances, is clearly too restrictive and unrealistic.

In order to understand and classify new opinion leaders, the first step is to broaden the notion of opinion leadership to accommodate all such new developments. This means that, although empirical analyses in the field should be promoted and intensified, the first step of a research agenda should be to elaborate a conceptual mapping of the relevant characteristics of an opinion leader. My tentative suggestion is that the nature of opinion leadership may well vary depending on different political fields and especially on different communication environments. As we have seen, offline and online communications retain different features that may produce different underlying processes of influence. However, only more focused empirical research could lend supporting evidence to this argument.