Indexing Theory

The indexing hypothesis, as proposed by W. Lance Bennett (1990), states that the range of debate on public affairs appearing in the news is indexed to the range of debate present in mainstream government discourse. With this hypothesis, Bennett pointed out that a variety of theories about the relationship between media and government, particularly political economy theories of media and theories of the sociology of news production, made similar predictions about what the news would look like. Since its initial publication in 1990, more than 100 journal articles have referred to Bennett's essay.

The concept of indexing has appeared most consistently in three areas of political communication research. First, indexing has implications for theories of media autonomy. Well-functioning democracies require an independent press, but much research indicates that press autonomy is limited at best. Indexing captures an aspect of that limited autonomy, for the hypothesis claims that when political officials do not [Page 330]differ on an issue, journalists will accept that consensus, fail to seek alternative perspectives, and present one-sided news. References to the indexing hypothesis also appear in the literature on media framing. Indexing offers an explanation of the origins of news frames as well as an explanation of why news sometimes contains multiple, competing frames and sometimes contains only one. Finally, the literature on “new institutionalism” makes use of the indexing concept. Indexing in this context is an example of an institutional practice that links organizational rules and interorganizational relationships to specific dayto-day organizational activities and outputs.

Although the indexing concept synthesizes important theoretical predictions and is intuitively appealing for many scholars of media and politics, empirical support for the hypothesis itself has been mixed. While some studies seem to reveal indexed news, others do not. In its initial formulation, the hypothesis is difficult to falsify. It is not entirely clear what constitutes “mainstream government debate.” For example, if the news includes the views of foreign government officials who disagree with American officials'consensus view on foreign policy, is it indexed or not? Moreover, the original formulation of the indexing hypothesis may refer to the range of official debate or to the distribution of official perspectives. Is news that confines itself to the perspectives offered by public officials but misrepresents the distribution of official support and opposition to various views indexed or not? Bennett himself initially expected that indexing would apply to some kinds of news events more than others and has more recently suggested that indexing offers only a partial explanation of how news is covered. These limitations suggest the need for further empirical study and revision of the indexing concept.