This project will apply economic tools to study the social impact of the media. Debates raging in the popular press, academic journals, and the halls of Congress point to effects of media on everything from test scores to murder rates to election outcomes. Despite the existence of large literatures across several disciplines, however, the empirical evidence on these effects remains largely inconclusive. The present research will make two broad contributions to filling this gap: (i) developing new data that will be a valuable new resource for the study of media effects; (ii) applying this data to study three specific topics of particular interest.
The first part of the project will measure the effect of television viewing on the cognitive and educational development of children. A new dataset on the timing of television's introduction to different local markets, combined with 1964 test scores from the Coleman study, will identify the causal effect of preschool exposure to television on standardized test scores later in life.
The second part of the project will study the determinants of "slant" in newspapers' political coverage. New data on newspapers' use of politically charged language, selective coverage of events, and citation of authorities will provide the first comprehensive picture of slant across newspapers and across time. Second-stage analysis will then identify the impact of slant on demand, the dependence of this impact on consumers' own political views, and the components of newspaper owners' objective functions.
The third part of the project will ask how the profusion of news sources available online has impacted consumers' news consumption. Two leading theories--one viewing consumers as motivated by information, the other viewing them as seeking confirmation of their prior beliefs--make sharply different predictions about the response to an expanding choice set. Panel data on online news consumption are used to test these competing theories and ask how the expanding choice set has changed the distribution of political information across the electorate.
Broader Impacts: Beyond its contribution to knowledge in the social sciences, this project will have several broader impacts. It will contribute to learning by providing information to educators and parents about the cognitive effects of television, as well as engaging undergraduate and graduate students as active participants in the research process. It will create important data resources that will be made freely available to other researchers, as well as a new methodological template for measuring political slant in news coverage. The results will be disseminated both through leading economics journals and through contacts with policy makers, practitioners, and the press. Finally, the findings will directly inform a number of ongoing debates with direct consequences for parents, educators, media industry participants, and policy makers.
