This paper presents a rationale for studying the effect of a community's news media resources on the diversity of topics that the public considers important (the agenda). Research on media effects has followed a model that implicitly assumes that the important effects consist of the persuasion of individuals to adopt specific positions. This paper demonstrates the feasibility of effects research that is free from this assumption and focuses on diversity in agenda holding within the community as the main criterion variable. To illustrate this type of research, secondary analyses of two bodies of archived survey data are presented. Communities are grouped according to their comparative media richness, as indicated by the number of competing newspapers published in each locale. Both the mean number of problems mentioned per person and the entropy in the distribution of problems across a set of categories are higher in communities with competing daily newspapers. (Author/JM)