Abstract
A vital role of public media according to the theoretical literature is to enable consumers to cross-check the information provided by slanted private media. To check the neutrality of public media, we conducted a randomized field experiment in Japan in collaboration with the nation's public service broadcaster in which the capacity for viewing its programs was randomly increased. Contrary to our expectations, we find that the treatment increased the evaluation of certain policies by raising the viewing time of its programs. To study the mechanism, we use unsupervised machine learning to measure media slant using the semantic similarity between TV programs and the official statements. Using this metric, we find that the positive effect is driven by subjects’ exposure to information slanted in favor of the domestic government during the experiment. These findings highlight the importance of creating institutions that secure the independence of public media.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call