In the current political landscape, the media play an increasingly dominant role. But while political communication scholars have accumulated much evidence over recent decades that questions the limited media effects model, electoral research often tends to neglect media effects in their models, perpetuating the limited effects hypothesis by omission. This paper makes a case for a systematic inclusion of media variables that measure, usage, type and tone of messages received to further our understanding of how media affect political competition.
The study of media impact is not without difficulties. Media effects can be small and difficult to detect, although this doesn’t make them trivial. In addition, media effects often are short term effects that require targeted data collection to allow for meaningful analysis. And lastly, because media systems tend to be idiosyncratic the measurement of media effects in cross-country comparative research is challenging to say the least.
However, we argue that the study of media effects has never been more promising. Panel voter studies – crucial to pick up individual level variation and hence media impact during the campaign – and electronic access of media outlets – easing large scale content analysis – are becoming the standard. Our paper sets out how these developments can be of use in future research on electoral competition. We discuss the requirements of an ideal media impact study, with regards to data requirements and analytical implications, and explore how cross-national research complicates this. We then confront the ideal design with the data and options available to present day researchers, and set out suggestions for future improvements, with regard to data collection as well as analytical design.
