Polarized News Media and the Polarization of the Electorate

Introduction

In democracies that usually divide into two camps like the U.S. and Canada, recent political communication research has identified a dramatic change in the media environment: from broadcasting to narrowcasting. Such change is attributed to the increase of news outlets and news programs aiming at segmented audiences. In addressing this phenomenon, some scholars continue to explore and confirm the existence of media fragmentation (e.g., Mendelsohn & Nadeau, 1996; Webster, 2005) while others are paying more attention to possible consequences of media fragmentation (e.g., Galston, 2003; Jones, 2002). Scholars have provided perspectives from different sides, but it remains a question whether a fragmented and polarized news media environment is responsible for the emergence of opinion polarization in the electorate. While there is growing concern about polarization in the news media, particularly in talk shows on TV, it is equally possible that the audience, given more options in terms of media types and program types, are less influenced by specific biased channels. In other words, before we confirm a causal link from the polarized media to the polarized electorate, we should consider if we might have overstated the concern about the influence of a polarized news media on a polarized electorate (Rosentiel, 2006).

Polarization is a process in which two separate groups are formed with opposite opinions or positions. Besides the meanings of “division” and “split,” the term “polarization” in the paper emphasizes the degree of opinion strength and extremity. A divided society may not necessarily be a polarized one, while the reverse is usually true. A 50-50 separated electorate may not be a polarized one if the voters hold mild opinions. The level of polarization, hence, cannot be properly measured or judged by whether or not a group is divided; instead, we need to take into account the level of division and the degree of opinion extremity.

Following this rationale, the term polarization in this paper is operationalized in two ways: polarization at the society level and perceived polarization at the community or network level. First, polarization at the society level will increase when there is an increase in the number of voters that hold extreme or relatively strong positions favoring or against a political camp.¹ Second, given that polarized politics is not simply a matter of the actual positions held by partisans but also a consequence of the perceptions each side holds of the other (Huckfeldt, Carmines, Mondak, & Palmer, 2005), perceived polarization at the community or network level will increase when an individual perceives or detects an increase in the number of its network members holding extreme or relatively strong positions.

What would be the effect of a polarized news environment on the polarization of the electorate? What can be the circumstances under which voters perceive less polarization within their communication networks? Given the definition of polarization and the concern of polarized media effect, one would expect that when voters access self-selected news media during a campaign season to form a voter choice, the level of polarization will increase, meaning that voters’ opinions will grow stronger and the number of people perceiving preference division within their networks will increase. But, how likely will this expectation hold true when we consider the complexity of the communication processes, such as interactions with multiple news sources, including news media, family, and close friends?

This paper employs Bibb Latane’s (1996) simulation approach and presents results of a series of computer-based experiments. Computer-based simulation allows a researcher to look beyond specific contexts and focus on causal effects of certain stimuli on phenomena of interest. Although the design of the study is not fixed to a particular context, the circumstances of concern are well specified, including polarized news media, divided electorate, and self-selected communication networks.

The next section will summarize recent empirical findings about media polarization and its possible effect on the electorate. The third section will detail the design of the computational experiment, composed of polarized news media objects and two types of citizen agents. The fourth section will illustrate experiment results, followed by a conclusion and discussion about the implications of the findings for future research.