A test of rivaling approaches to explain news effects: A multi-wave panel study of agenda setting, social and economic conditions, the tone of the news, and horse race news

Different "paradigmatic" approaches to explain news effects on voting may supplement each other, because their starting points are based on different news types in political campaign news: news on issue positions of parties, news on real-world developments, news on support or criticism for parties, and news on success and failure of parties. Daily content analysis data and a weekly multiwave panel survey from the 2003 election campaign in the Netherlands are used for a simultaneous test. A logistic regression analysis demonstrates that the paradigmatic approaches supplement each other. The data reveal a huge impact of the news from a campaigner's point of view in spite of a huge variety in responses to the news at the level of individual respondents. © 2007 International Communication Association.