ONE
Introduction
“In the simplest terms,” Siebert, Peterson, and Schramm wrote in Four Theories of the Press (1956), “the question behind this book is, why is the press as it is? Why does it apparently serve different purposes and appear in widely different forms in different countries? Why, for example, is the press of the Soviet Union so different from our own, and the press of Argentina, so different from that of Great Britain?”
Nearly half a century later the field of communication has made limited progress in addressing this kind of question. Though there have been attempts, particularly since the 1970s, to push the field in the direction of comparative analysis, such a research tradition remains essentially in its infancy.1 We attempt in this book to propose some tentative answers to the questions posed by Siebert, Peterson, and Schramm – though not on such a grand scale. We confine ourselves to the developed capitalist democracies of Western Europe and North America. We attempt to identify the major variations that have developed in Western democracies in the structure and political role of the news media, and to explore some ideas about how to account for these variations and think about their consequences for democratic politics. We place our primary focus on the relation between media systems and political systems, and therefore emphasize the analysis of journalism and the news media, and, to a somewhat lesser extent, media policy and law.
WHY COMPARATIVE ANALYSIS?
It is worth dwelling for a moment on one of the most basic insights of Siebert, Peterson, and Schramm: the idea that if we want to address a question such as “Why is the press as it is?” we must turn to comparative analysis. The role of comparative analysis in social theory can be understood in terms of two basic functions: its role in concept formation and clarification and its role in causal inference.2
Comparative analysis is valuable in social investigation, in the first place, because it sensitizes us to variation and to similarity, and this can contribute powerfully to concept formation and to the refinement of our conceptual apparatus. Most of the literature on the media is highly ethnocentric, in the sense that it refers only to the experience of a single country, yet is written in general terms, as though the model that prevailed in that country were universal. This, at least, is true in the countries with the most-developed media scholarship, including the United States, Britain, France, and Germany. In countries with less developed traditions of media research, another pattern often emerges: a tendency to borrow the literature of other countries – usually the Anglo-American or the French literature – and to treat that borrowed literature as though it could be applied unproblematically anywhere. We believe this style of research has often held media researchers back from even posing the question, “Why are the media as they are?” Important aspects of media systems are assumed to be “natural,” or in some cases are so familiar that they are not perceived at all. Because it “denaturalizes” a media system that is so familiar to us, comparison forces us to conceptualize more clearly what aspects of that system actually require explanation. In that sense comparative analysis, as Blumler and Gurevitch (1975: 76) say, has the “capacity to render the invisible visible,” to draw our attention to aspects of any media system, including our own, that “may be taken for granted and difficult to detect when the focus is on only one national case.” Our own comparative work began with the experience of exactly this type of insight. Comparing U.S. and Italian TV news in the early 1980s, familiar patterns of news construction, which we had to some extent assumed were the natural form of TV news, were revealed to us as products of a particular system. We were thus forced to notice and to try to account for many things we had passed over, for example, the highly interpretive character of American compared with Italian TV news, a characteristic that contradicted common assumptions about “objective” journalism in the American system (Hallin and Mancini 1984).
Comparative analysis makes it possible to notice things we did not notice and therefore had not conceptualized, and it also forces us to clarify the scope and applicability of the concepts we do employ. Comparative studies, as Bendix (1963: 535) puts it, “provide an important check on the generalizations implicit” in our concepts and forces us to clarify the limits of their application. Sociologists, for example, had assumed “urbanization” to be so closely associated with secularism and Western forms of individualism that the latter could be treated as part of the very notion of urbanism – a generalization that, Bendix argued, fell apart when we looked at India or other non–Western societies. In a similar way we will try to clarify the conceptual definitions of a number of key concepts in media studies – journalistic professionalization, for example – and to use comparative analysis to discover which aspects of those concepts really do vary together and which do not.
If comparison can sensitize us to variation, it can also sensitize us to similarity, and that too can force us to think more clearly about how we might explain media systems. In the United States, for example, media coverage of politicians has become increasingly negative over the past few decades. We typically explain that change by reference to historical events such as Vietnam and Watergate, as well as changes in the conduct of election campaigns. This trend is not, however, unique to the United States. Indeed, it is virtually universal across Western democracies. The generality of this change, of course, suggests that particular historical events internal to the United States are not an adequate explanation. Comparative analysis can protect us from false generalizations, as Bendix says, but can also encourage us to move from overly particular explanations to more general ones where this is appropriate.
Of course, comparative analysis does not automatically bring these benefits. It can be ethnocentric itself, imposing on diverse systems a framework that reflects the point of view of one of these – though this is probably most true of work that, similar to Four Theories of the Press, purports to be comparative but is not in fact based on comparative analysis. We will argue later in this chapter that ethnocentrism has been intensified in the field of communication by the strongly normative character of much theory. Comparison can indeed be ethnocentric. We believe, however, the comparative method properly applied provides a basis for systematic critique of work that falls into these patterns of overgeneralization and conceptual narrowness.
The second reason comparison is important in social investigation is that it allows us in many cases to test hypotheses about the interrelationships among social phenomena. “We have only one means of demonstrating that one phenomenon is the cause of another: it is to compare the cases where they are simultaneously present or absent,” wrote Émile Durkheim (1965) in The Rules of Sociological Method. This has become the standard methodology in much of the social sciences, particularly among those interested in analyzing social phenomena at the system level, where variation will often not exist in a single-country study. There are, of course, many epistemological debates surrounding the effort to find “sociological rules” in Durkheim’s sense. Some believe social theory should follow the natural sciences in the search for laws that are “always and everywhere the case”; others believe that the generalizations of social theory will necessarily be relative to particular systems and historical contexts. Some believe explanation requires a clear identification of cause and effect, “dependent” and “independent” variable; others think in terms of identifying patterns of coevolution of social phenomena that might not always be separated into cause and effect. In the field of communication, those who do analysis at the system level often tend to be skeptical of “positivism”; the “positivists” in the field tend to be concentrated among people working at the individual level. For many years empirical research in communication was almost synonymous with the media effects paradigm, which was concerned not with larger media structures but with the effects of particular messages on individual attitudes and beliefs. This may be one reason systematic use of comparative analysis has developed slowly. We believe, however, that it is not necessary to adopt strong claims of the identity between natural and social science to find comparative analysis useful in sorting out relationships between media systems and their social and political settings.
Let us take one example here. Jeffrey Alexander, in an unusual and very interesting attempt to offer a comparative framework for the analysis of the news media, poses the question of how to explain the particular strength of autonomous journalistic professionalism in the United States. One hypothesis he offers is that “it is extremely significant that no labor papers tied to working class parties emerged on a mass scale in the United States” (1981: 31). He goes on to contrast U.S. press history with that of France and Britain, and advances the claim that the absence of a labor press in the United States explains the development of autonomous professionalism. We will discuss Alexander’s important theoretical framework in greater detail in Chapters 4 and 8. As for the particular hypothesis about the labor press, comparative analysis allows us fairly easily to dismiss it, once we go beyond the comparison between the United States and France. There are a number of cases in Europe where a strong labor press and strong professional autonomy of journalists both developed; indeed we argue that this pattern is typical of most of Northern Europe. What other factors might account for journalistic autonomy we take up later (as well as a number of questions about how to define it).
The use of comparative analysis for causal inference belongs to a relatively advanced stage in the process of analysis. Our own study is primarily exploratory in character, using comparative analysis to serve the first cluster of purposes previously outlined, for conceptual clarification and theory development, much more than for the second, for hypothesis testing and causal inference. Our purpose here is to develop a framework for comparing media systems and a set of hypotheses about how they are linked structurally and historically to the development of the political system, but we do not claim to have tested those hypotheses here, in part because of severe limitations of data underscored in the following text.
Comparative analysis, particularly of the broad synthetic sort we are attempting here, is extremely valuable but difficult to do well, especially when the state of the field is relatively primitive. It is risky to generalize across many nations, whose media systems, histories, and political cultures we cannot know with equal depth. This is why we have undertaken this project as a collaboration between an American and a European. Some might wonder why we did not try to organize a broader collaboration. There are, of course, many practical difficulties in such an enterprise, but the fundamental reason is that our purpose in this book is to produce a cogent theoretical framework – or at least to move toward one. Multinational collaborations in our field have often tended to fall back on the least common denominator in terms of theory, or to leave theoretical differences unresolved. We hope that scholars will find our general arguments interesting enough to excuse occasional errors or lack of subtlety in dealing with particular cases. In comparative research, much of the real collaboration is of course indirect. Our study builds on a growing body of scholarship across Europe and North America, and we hope that many of these scholars will eventually carry the ideas proposed in this volume much further than we can do here.
SCOPE OF THE STUDY
This study covers the media systems of the United States, Canada, and most of Western Europe, excluding only very small countries (e.g., Luxembourg, much of whose media system is actually directed toward audiences in neighboring countries). Our study is thus based on a “most similar systems” design. As Lijphart (1971) stresses, one of the greatest problems in comparative analysis is the problem of “many variables, few cases.” One of the principal means of solving that problem, he notes, is to focus on a set of relatively comparable cases, in which the number of relevant variables will be reduced. This approach will reduce the number of cases; but in a field such as communication, where the existing literature and available data are limited, this is often a benefit as well in the sense that it is impossible for analysts to handle competently more than a limited number of cases. One of the problems of Four Theories of the Press, as we noted, is that its scope is so grand that it is almost inevitably superficial: like a photo with too much contrast, it obscures too much of the detail we need to see.3 By limiting ourselves to North America and Western Europe we are dealing with systems that have relatively comparable levels of economic development and much common culture and political history. This is a limitation, obviously: the models developed here will not apply without considerable adaptation to most other areas of the world, though we hope they will be useful to scholars working on other regions as points of reference against which other models can be constructed. One advantage of this focus is the fact that the media models that prevail in Western Europe and North America tend to be the dominant models globally; understanding their logic and evolution is therefore likely to be of some use to scholars of other regions not only as an example of how to conduct comparative research but also because these models have actually influenced the development of other systems.
Our study, as mentioned previously, is an exploratory one, and the main purpose of the “most similar systems” design is not to hold certain variables constant for purposes of demonstrating causality, but to permit careful development of concepts that can be used for further comparative analysis, as well as hypotheses about their interrelations. The fact that it is an exploratory study also means that the geographical definition of its scope is in some ways arbitrary: we did not already have a theoretical framework that could provide the basis for selection of cases. Instead we followed the familiar strategy of limiting the study to a region on the assumption that this would result in a reasonably comparable set of cases. “Comparability,” as Lijphart (1971: 689) says, “is not inherent in any given area, but it is more likely in an area than in a randomly selected set of countries.” The area approach also made the study more manageable in a practical sense – we were able to visit the countries more easily, for instance, and to take advantage of the relatively large amount of comparable data compiled on European media systems. We could probably have added Australia and New Zealand – whose historical connections make them very similar to Western European countries – to our study without making the conceptual framework significantly more complex. We suspect, however, that most other cases we might have added would have introduced important new variables, straining our ability to master the relevant literatures and present the resultant framework in a coherent way. In Chapter 4 we introduce a triangular drawing on which each of our cases is represented in relation to three media-system models. Any significant multiplication of cases would probably have made such a two-dimensional representation impossible!
The desire to “reduce the property space of the analysis,” in Lijphart’s terms, is also reflected in our decision to focus primarily on news media and media regulation. A comparative analysis of media systems certainly could include much more about cultural industries – film, music, television and other entertainment; telecommunication; public relations; and a number of other areas. But this would involve other literatures and require very different sets of concepts and we will not try to take it on here.
THE LEGACY OF FOUR THEORIES OF THE PRESS
Since we began with Four Theories of the Press, a work that remains remarkably influential around the world as an attempt to lay out a broad framework for comparative analysis of the news media, it makes sense to follow Siebert, Peterson, and Schramm’s argument a bit further.4 “The thesis of this volume,” they continue, “is that the press always takes on the form and coloration of the social and political structures within which it operates. Especially, it reflects the system of social control whereby the relations of individuals and institutions are adjusted. We believe that an understanding of these aspects of society is basic to any systematic understanding of the press” (1–2). Here again, we think the problem is well posed. We shall follow the agenda set out by Siebert, Peterson, and Schramm in attempting to show how different media models are rooted in broader differences of political and economic structure. We will argue that one cannot understand the news media without understanding the nature of the state, the system of political parties, the pattern of relations between economic and political interests, and the development of civil society, among other elements of social structure.
On one point, we will leave matters a bit more open than the authors of Four Theories of the Press. Note that Siebert, Peterson, and Schramm seem to assume that the media will always be the “dependent variable” in relation to the “system of social control,” which it “reflects.” In this sense, their formulation is ironically similar to a traditional Marxist base and superstructure theory (though as we shall see in a moment they quickly stand Marx on his head). In many cases it may be reasonable to assume that the media system essentially “reflects” other aspects of social structure – the party system, for example. But there is good evidence that media institutions have an impact of their own on other social structures.
There is also historical variation in the degree to which media are reflective or independently influential, and many scholars have argued that there is an important trend in the direction of greater media influence, particularly in relation to the political system. The belief that the media have become an important “exogenous” variable affecting other political institutions is one reason scholars in comparative politics have begun to pay attention to media institutions they previously ignored. It is worth noting that, just as communication scholars have paid little attention to comparative analysis, scholars of comparative politics have paid little attention to the media. One can search the index of the classic works on political parties and find virtually nothing on the press or media, even though politicians have certainly been preoccupied by – and occupied in – the latter as long as political parties have existed, and even though those classic works often define parties as communicative institutions (Deutsch 1966; Sartori 1976), a theoretical perspective that would seem to suggest they would have an important pattern of relationships with other institutions of communication.
Today this is beginning to change, due in part to a growing feeling that the media are less “reflective” than they once were. Sometimes this change may actually be exaggerated. Media scholars – following the tradition of McLuhan – often tend to have a professional bias toward overstressing the independent influence of media. And scholars from other fields sometimes do so as well, perhaps out of a sense that the media are “overstepping their bounds” as they become more powerful relative to other sorts of institutions. Bourdieu’s recent work, On Television (1998), might be an example here, as well as much speculation in comparative politics about “videocracy.” In Chapter 8, we will address the question of the reciprocal influences of the media and the political system, and try to sort out some of the arguments about the relative influence of media system change in shaping contemporary European political systems.
Siebert, Peterson, and Schramm go on:
To see the differences between press systems in full perspective, then, one must look at the social systems in which the press functions. To see the social systems in their true relationship to the press, one has to look at certain basic beliefs and assumptions which the society holds: the nature of man, the nature of society and the state, the relation of man to the state, and the nature of knowledge and truth. Thus, in the last analysis the difference between press systems is one of philosophy, and this book is about the philosophical and political rationales or theories which lie behind the different kinds of press we have in the world today (2).
At this point, we part company with Siebert, Peterson, and Schramm. To be sure, we too believe that political culture is important, and we will try to show how differences in media systems are connected with socially shared conceptions about state and society, objectivity, the public interest, and the like. But the focus on “philosophies” of the press – or as one might also call them, “ideologies” of the press – points to what we see as a key failing of Four Theories of the Press. Siebert, Peterson, and Schramm did not, in fact, empirically analyze the relation between media systems and social systems. They looked neither at the actual functioning of media systems nor at that of the social systems in which they operated, but only at the “rationales or theories” by which those systems legitimated themselves. “In arguing that ‘in the last instance the difference between press systems is one of philosophy’ the book disregards the material existence of the media” (Nerone 1995: 23).
Nor was their analysis actually comparative. In part, this was because of the background of the Cold War: because it is so preoccupied with the dichotomy between the contending U.S. and Soviet models, Four Theories of the Press has little room for the actual diversity of world media systems. In tracing the origins of the four theories, for example, Siebert, Peterson, and Schramm make reference almost exclusively to three countries – the United States, to which they trace the libertarian and social responsibility theories; Britain, to which they trace both the authoritarian and, along with the United States, the libertarian theories; and the Soviet Union. All the models, moreover, are really “defined . . . from within one of the four theories – classical liberalism” (Nerone 1995: 21). The four theories are of limited use in understanding the European experience. One could say that Western Europe has combined the libertarian model (manifested in the relatively unregulated commercial and party press and the tradition of advocacy journalism); the social responsibility model (public broadcasting, right-of-reply laws, press subsidies, press councils); and the authoritarian tradition (Gaullist state broadcasting or the British Official Secrets Act, as well as the controls exercised in periods of real dictatorship). One could probably say that any system combines these elements in some way. But this is far too thin a framework to begin a real comparative analysis.
Four Theories of the Press has stalked the landscape of media studies like a horror-movie zombie for decades beyond its natural lifetime. We think it is time to give it a decent burial and move on to the development of more sophisticated models based on real comparative analysis.5
MEDIA SYSTEM MODELS
One reason Four Theories of the Press has proved so influential over so many years is that there is a great deal of appeal in the idea that the world’s media systems can be classified using a small number of simple, discreet models. Is it possible to replace the four theories with a new set of models, better-grounded empirically but sharing something of the parsimony of the originals? Only with great caution. We will in fact introduce three media system models. These will be elaborated more fully in the following chapter, but briefly they are the Liberal Model, which prevails across the Britain, Ireland, and North America; the Democratic Corporatist Model, which prevails across northern continental Europe; and the Polarized Pluralist Model, which prevails in the Mediterranean countries of southern Europe. The Liberal Model is characterized by a relative dominance of market mechanisms and of commercial media; the Democratic Corporatist Model by a historical coexistence of commercial media and media tied to organized social and political groups, and by a relatively active but legally limited role of the state; and the Polarized Pluralist Model by integration of the media into party politics, weaker historical development of commercial media, and a strong role of the state. We will try to show that the characteristics that define these models are interrelated, that they result from a meaningful pattern of historical development, and do not merely co-occur accidentally. We will also use these models to organize the discussion of the media systems of individual countries, trying to show how each country’s media system does and does not fit these patterns.
Many qualifications must be introduced as soon as we begin to use these models. They are ideal types, and the media systems of individual countries fit them only roughly. There is considerable variation among countries that we will be grouping together in our discussion of these models. The British and American media systems (which we will discuss as examples of the Liberal Model) are in fact quite different in many ways, even though it is common to talk about the Anglo-American model of journalism as though it were singular. Italy, with a “consensus” political system and a full half-century of democratic government is quite different from Spain, with a majoritarian system and a much later transition to democracy, though both are close to the Polarized Pluralist Model in many characteristics. We will discuss Germany in relation to the Democratic Corporatist Model, though it is quite different from the small democracies that represent the classic cases of that model. We will discuss France in relation to the Polarized Pluralist Model of the Mediterranean countries, but we shall see that it is something of a mixed case between the Polarized Pluralist and Democratic Corporatist Models, as Britain is a mixed case between the Liberal and Democratic Corporatist Models. In part we hope that the models will be useful precisely in bringing these variations to light. It should be stressed that their primary purpose is not classification of individual systems, but the identification of characteristic patterns of relationship between system characteristics.