CHANGES OF MASS MEDIA AND THE PUBLIC SPHERE WINFRIED SCHULZ Three different models of the public sphere have gained wide attention in the academic literature. In a recent article Gerhards (1997) contrasts two of these models, the discursive public and the liberal public, respectively. The former is linked to the well-known work of Habermas on the transformation of the public sphere (1962) and to some of his more recent reflections and modifications (Habermas 1992a,b). The latter model is rooted in theories of liberal democracy which were developed in political science. Gerhards also traces influences, among others, to the writings of Luhmann (e.g., Luhmann 1971). Both models are "ideal types" in the sense Max Weber introduced this term, i.e. they give a rather abstract, idealised notion of some basic features and functions of the public sphere. However, Habermas and Gerhards both claim that their respective models have empirical value and are suitable for a description of observable phenomena (Habermas 1992b, 451; Gerhards 1997). The two models have some key features in common. They both conceive of the public sphere as an intermediary system which links the "basis" with the "top" of the political system or, as Habermas puts it, the private and collective actors of the periphery with the political institutions in the centre. Both conceptions have some obvious resemblance with political systems models of the type devised by Easton (1965). Systems models of this type contrast input processes, output processes and conversion processes. Mass media are considered by these models as just one of several channels or agents through which the interests and the will of the people are transformed or "converted" into political decisions. The models differ with respect to the position and role of political actors, particularly of interest groups. In the liberal model organised collective actors, like interest groups and political parties, dominate the public sphere and provide inputs to the political decision centre. Winfried Schulz is Professor at the Department for Mass Communication Studies, Institute for Social Sciences, University of Erlangen-Nürnberg. rN O > Ln These groups are part of the conversion processes. They aggregate the different preferences of individual citizens and represent their interests in the political arena. They mirror the variety of interests in a pluralistic society. In contrast to this, the Habermas model expects that the relevant political input to the public sphere comes directly from individual citizens and from collective actors of the civil society. Habermas makes a sharp distinction between actors of the civil society on one side and powerful actors on the other. Typical actors of the civil society are social movements and voluntary associations. Interest groups, like unions, churches, professional associations etc. are categorised as powerful actors. The role of the media is seen by Habermas as being somewhat ambivalent and predominantly sceptical. On the one hand, at least in "moments of mobilisation" under conditions of a perceived crisis, he argues, they may serve as a platform or even promoter of interests of the actors of the civil society. On the other hand, he suspects that the mass media are power-infiltrated by interest groups and serve as their means to manipulate the public. Another difference between the two models concerns the style of communications in public. While in the liberal model all communications and the actions of all actors are accepted as long as they respect other actors with different opinions, the Habermas model is quite demanding as to the style of public communications. It expects that reasons are given for issues and positions and that actors who put forward certain arguments refer to the respective arguments of other actors. This results in a public discourse of a higher level of rationality which is the precondition of reaching a consensus on political decisions, or at least at a majority opinion which is based on arguments. The legitimacy of political decisions is a function of the rationality of the public discourse. The liberal model, on the other side, expects that public opinion is produced by the aggregation of individual opinions. Majority opinions deserve no special quality whether they are based on a rational discourse or not (Gerhards 1997). There is a third model of the public sphere which has its advocates mainly in the field of mass communication research (see e.g., Blumler 1990; Schoenbach and Becker, 1995; Mancini and Swanson 1996). The label "model of a media constructed public sphere" denoted its characteristics quite well. In order to highlight the differences to the former two models a descriptive scheme which Gerhards (1997) uses in his article is adopted here and expanded by one more column which refers to the third model. Also a few labels are added to Gerhards' scheme as well as the new subcategory "role of mass media" (see table 1). The perspective of this third model is different from the other two in many respects. The most important difference is that mass media are regarded as the precondition of a public sphere. Although there are some arenas in which persons interact in order to exchange information and opinions, like coffee houses, club meetings, and party conventions, the relevance of such interactive publics for the formation of a public sphere falls far behind the relevance of the media publics which are constituted by mass communication. Empirical data show that during election campaigns almost every voter is reached by mass communicated messages, and most of them many times, while only a relatively small minority attends meetings which provide possibilities for personal interactions. For instance, during the 1990 campaign for the German parliamentary elections only 12 per cent of the voters attended a party event (Noelle-Neumann and Köcher 1993, 699). If people are asked to rank different sources according to their importance for political information and opinion formation, television and the press are mentioned most often. Personal interactions play only a minor a (A u j5 3 IL (U o M (U o z (U a ■Q (0 I- CT"Ö c c Ü £ TJ C m O (D £i {3 tj (.o £i™CT rote 2 § I | I o - E tc o <^ro ^ ro (3 " E tj C o u re TD OOro aal: a o o <2 <5 12 tj ž Ü o .!= U C ^ E ^ I ° (J O O o .(D CO > ct 'ctì re L. O r r (J L. r o <1 n u CI O (j <5 CT D . E CT u E oo "CT ^ CD £Ì -D 3 a s: (A C o . o . LD role (Schulz and Blumler 1994, 212; Semetko and Schönbach 1994, 73, 78). Not only during election campaigns but also for keeping abreast with current events every day, people rank mass media much higher than personal interactions. Table 2 gives an example from a representative survey of the German population conducted in autumn 1995: Table 2: Sources of Political Information Question: Where do you get most of your information on what is currently happening in the world? I will read to you some possibilities: - on radio 53 % - on television 87 % - in the newspapers 66 % - in magazines 12 % - from conversations with other people 25 % Source: Massenkommunikation 1995 (n=6,000; for methodological details see Kiefer, 1996) A second reason which serves to support the argument of a preponderance of media publics over interactive publics is derived from the hypothesis of media dependency and related empirical research. Briefly stated, the argument holds that even interactive publics are very much dependent on information and viewpoints which people have picked up from the mass media. For instance, mass media are frequently mentioned in personal conversations as sources of information or for substantiating ones' own opinions (Kepplinger and Martin 1986). In addition, public events which form interactive publics, like political meetings, party conventions, or demonstrations, can be "upgraded" by media coverage to a much higher level of publicity and usually become part of the general public sphere only through mass media. By selecting among different events the media have control over the transformation of interactive publics into media publics. As gate-keepers they decide on which events, actors, issues and arguments become public. More than that, they actively construct the public according to media formats and news media criteria such as personalisation, negativism, drama, or emotion. The following analysis concentrates on two aspects, (1) on changes of publics and the public sphere and (2) on consequences of these changes. These are derived from the model of a media constructed public sphere in order to confront some implications of this model with empirical data. Our intention is neither to prove this model nor to disprove the other two models of the public sphere, but rather to add a perspective, which is neglected by the other two models. I may be seen that this perspective which has its tradition in empirical communication research, is compatible with the theoretical reflections on the structure of the public sphere. Changes of the Public Sphere Changes of publics are the focal point of Habermas' famous book on the transformation of the public sphere (Habermas 1962). He analysed the transformation processes which took place between the seventeenth and the nineteenth century (with a few critical looks on our century until the late fifties). In his more recent publications he expands his views somewhat and adds a few remarks on more modern developments like the introduction of television (Habermas 1992a,b). Changes of the media and its publics are a key topic of mass communication research. Changes which are discussed quite frequently are those which originate from the introduction of the electronic media in the fifties and sixties and from the deregulation and commercialisation of the European media systems in the eighties. In Germany, for instance, an empirical finding gained some attention which seemed to demonstrate a considerable increase in political interest parallel to the diffusion of television among the population. Figure 1 shows these processes. The diffusion of TV was interpreted as a major stimulation to people's interests in politics, although the covariations which the data display do not necessarily prove causality (Noelle-Neumann 1979). A closer look at the data with statistical methods of time series analysis failed to corroborate the influence of television on the development of political interest among the German population and offers some alternative explanations (Schulz 1995). Figure 1: Assumptions on the Structure of the Media Public Size of the Channel-Repertoire low high little Overlapping of the Channel-Repertoire big segmented fragmented integrated expanded In the United States several researchers observed a quite different development resulting from the establishment of a media constructed public sphere. Robinson (1975, 1976, 1977), for instance, referred to survey results which demonstrated that reliance on television for political campaign information is associated with political inefficacy, distrust and cynicism. He coined the term "videomalaise" in order to express that the introduction of television resulted in a feeling of political malaise among the public. Gerbner and his collaborators (Gerbner and Gross 1976, Gerbner 1990) in several studies presented data which showed that television viewing "mainstreams" the outlook and social reality beliefs of the public and "cultivates" fear, alienation, and interpersonal mistrust. Attempts to replicate such findings in a European setting arrived at negative or mixed results (see particularly Holtz-Bacha 1990, also for a review of research). This may be explained by the differences between the American and the European media systems at that time. While American television was — and still is — highly entertainment-oriented and saturated with crime and violence, in practically all European countries television was information oriented and committed to the idea of broadcasting as a public service. This has changed in the mid-eighties when private television was introduced in almost all European countries. More or less, private channels adopted American TV formats with a heavy load of entertainment, crime and violence. As a consequence, the viewing habits and the television intake of the population have changed dramatically. The following changes are very well documented empirically for Germany as well as for other European countries: • Television viewing time increased, particularly among the younger generations • Many people make use of a wide range of different TV channels and programmes • Many people watch TV more superficially and as a secondary activity • The over-all supply of TV programmes became very much entertainment oriented • Entertainment viewing increased much more than information viewing • Information programmes on television became more tabloidised with more emphasis on sensationalism, negativism, personalisation, and soft news. Consequences of a Media Constructed Public Sphere It is quite plausible to assume that these changes have consequences for the structure of the public sphere. Different consequences are hypothesised depending on what role is assigned to mass media for "communication in public," either as part of the intermediary system, or as platform or manipulator, or as a precondition of the public sphere (see Table 1). The following remarks and data refer to two types of consequences which may be derived from the media construction model of the public sphere. One consequence of the recent changes of European media systems concerns the television audience's viewing habits and styles. How do people cope with the proliferation of channels and to what degree is the structure of the TV audience and, more generally, the structure of the media constructed public sphere affected by the media changes? A distinctive feature of the transformation processes in the recent past and, perhaps more distinctively, in the future, may be an increasing fragmentation of the public (McQuail 1994, 71; Mancini and Swanson 1996). Fragmentation results from the enormous expansion of programme choices and, at the same time, the diversification and specialisation of TV programmes. Quite similar tendencies may be expected to result from the increase of radio programmes, magazines, local papers and on-line media, at least in media affluent societies like Germany. An implicit assumption of the fragmentation hypothesis is that people concentrate on certain media or channels and neglect others so that there will be only little or no overlapping of audiences. As a consequence, different segments of the society are attuned to different streams of information, world views, and value systems. The common ground of experience for all members of society dissolves and the public sphere breaks to several fragmented publics, even esoteric circles. Figure 2 tries to represent, in a rather abstract and idealised way, the assumptions of the fragmentation hypothesis with reference to potential structures of a media constructed public sphere. The "integrated" structure is the ideal pattern of a media public in a public service broadcasting system. This was the dominant normative paradigm for media policy in post-war Germany until the eighties. Since then it has been substituted by the "expanded" structure as the ideal pattern. The other two constellations, the "segmented" and the "fragmented" patterns, can be interpreted as empirical and critical views, respectively, of audience structures with little overlapping. How much concentration or diversification is manifested in people's media choices and to what degree different audiences do overlap is an empirical question which can be answered on the basis of audience research data. For instance, the range of people's choices among television channels is indicated by the size of their channel repertoire. Channel repertoire is a term which characterises the number of available TV channels that viewers choose to watch (Ferguson and Perse 1993). A representative survey among the German population in 1995 shows that 25 per cent of the German population still have a rather low repertoire consisting of just two channels which are watched regularly or frequently. In spite of the expansion of offers the channel repertoire of these Figure 2: Diffusion of TV and Interest in Politics in Western Germany m»Intersil. n Politici* - Houirfiolds with » TV «t (%) ■)Question: GtntMlpipoktm Afs ■y o j InCErE^lod 'n (wlitfct? Ti^rid lin? ihcwì |he per ùf rciQrdcnCl arnnsfing "irei-; tvcig-cs baisd on d numb*! iurveyi in every year, cnty Western Germany: wurm- Allensbach ifchirtt viewers remained quite low and their viewing habits still correspond with the "integrated" pattern which was typical for the era of the monopoly of public service broadcasting when most people had only two or three TV channels for choice. By 1995, a typical German TV household had around 30 different channels available via cable and more than 50 channels via direct broadcasting satellite.1 On the other hand, 32 per cent of the German population have a relatively high repertoire with six or more channels watched regularly or frequently (the rest of 43 per cent can be put into a "medium" group with a repertoire of three to five channels). Although viewers with a repertoire have not expanded their range of choices at the same rate as the number of channels has increased they make use of a sizeable portion of the new TV offers. However, contrary to the assumption of the model of a media-constructed public, the viewers with a high channel repertoire exhibit a remarkable overlapping of choices, too. Most of them watch both the major private channels (RTL, SAT 1, Pro 7) and the public channels ARD and ZDF (for details see Schulz 1997 b). With reference to the typology of figure 2, it seems to be appropriate to characterise this viewing pattern as "expanded" rather than as "fragmented." Thus, the development in Germany so far has not led to a considerable fragmentation of the television public. Other consequences of the transformation processes in the European media systems relate to the changes in programme content and style. These changes have often been labelled as "Americanisation." Thus it seems reasonable to ask whether those concomitants of American television which were observed by Robinson, Gerbner and others have been imported together with the commercial programmes. For the first time after the deregulation of the German broadcasting system we have now data available to test this question. The results which are presented here focus on just one aspect, namely the "videomalaise" hypothesis.2 A representative survey of the German population in autumn 1995 included questions which are considered to indicate political cynicism (or alienation) and which have been used quite often in other studies, particularly in the United States (see McKean et al. 1995; Pinkleton et al. 1996; Vetter 1997). We tested relationships of political cynicism, among others, with: • access to cable or satellite transmission • preference for commercial channels (as opposed to public service channels) • total time devoted to watching TV (heavy viewing) • size of channel repertoire • extensive usage of information programmes on TV • extensive usage of TV entertainment • extensive usage of information on radio • extensive usage of information in newspapers • personal conversation as a source of "much" or "very much" political information.3 Each of the television viewing indicators is significantly related to political cynicism if we look at bivariate relations. Table 3 gives an overview of the results. The three cynicism indicators in the table were selected on grounds of a factor analysis from a more comprehensive set of statements (for more methodological details, see Schulz 1997 b). Although the percentage differences between the total population and the different population segments are not dramatic, the pattern of results seems to be quite consistent. In contrast to this, people who use information in newspapers extensively as well as those who mention personal conversations as a source of "much" or "very much" political information are less inclined to subscribe to the negative statements about politics. The results for people with extensive usage of information on radio are mixed. One may suspect that the results shown by table 3 are spurious because media usage is systematically linked to factors like age, education and interest in politics, and that these linkages account for the associations of TV usage with political cynicism. In addition, the different TV usage indicators are correlated. To control for possible spu-riousness and to find out about the specific relationship of each media usage indicator to political cynicism we tested all variables in a multiple regression analysis. Also included as independent variables in the regression model were the indicators for information behaviour beyond television viewing, namely newspaper reading, radio listening and personal conversations. After rigorous simultaneous controls for eight demographic factors, interest in politics and all communication variables, most of the relationships of political cynicism with TV usage remained statistically significant. 4This was the case for • access to cable or satellite transmission, • preference for commercial channels, • extensive usage of TV entertainment, and • extensive usage of information programmes on TV Only TV viewing time and channel repertoire lost their association with political cyni-^ cism. The same happened to newspaper reading and personal conversations, whereas ra- 1_ 3 O ■j> (O .C :3 (D O -C -d ^ -ct' Cl 00 00 CD oo CD o6 oó 00 vd c< X co co £ ^ re ^ .O št <2 o ^ u CU Jk ■;=