Galetovic A., Anand B., Di Tella R., (2007), Information or Opinion? Media Bias as Product Differentiation, Journal of Economics and Management Strategy, vol. 16, 635-682

 

Abstract

Two aspects of media bias are important empirically. First, bias is persistent: it
does not seem to disappear even when the media is under scrutiny. Second, bias
is conflicting: different people often perceive bias in the same media outlet to be
of opposite signs. We build a model in which both empirical characteristics of
bias are observed in equilibrium. The key assumptions are that the information
contained in the facts about a news event may not always be fully verifiable,
and consumers have heterogeneous prior views (“ideologies”) about the news
event. Based on these ingredients of the model, we build a location model with
entry to characterize firms’ reports in equilibrium, and the nature of bias. When
a news item comprises only fully verifiable facts, firms report these as such, so
that there is no bias and the market looks like any market for information.
When a news item comprises information that is mostly nonverifiable, however,
then consumers may care both about opinion and editorials, and a firm’s
report will contain both these aspects—in which case the market resembles
any differentiated product market. Thus, the appearance of bias is a result of
equilibrium product differentiation when some facts are nonverifiable. We use
the model to address several questions, including the impact of competition on
bias, the incentives to report unpopular news, and the impact of owner ideology
We are grateful to the editors and two referees for their suggestions. For helpful conversations we thank Roni Fischer, Ariel Pakes, Julio Rotemberg, and Andrei Shleifer.
Anand and Di Tella gratefully acknowledge the financial support from the Division of
Research at Harvard Business School, and Galetovic the hospitality of the Stanford Center
for International Development.on bias. In general, competition does not lead to a reduction in bias unless this is accompanied by an increase in verifiability or a smaller dispersion of prior
beliefs.

Lascia un commento

Il tuo indirizzo email non verrà pubblicato. I campi obbligatori sono contrassegnati *

Post comment