Doron Shultziner, Hadassah Academic College Jerusalem
Yelena Stukalina, Tel Aviv-Yaffo Academic College
One important development in several established democracies over the last twenty years is the notion that “the media” is biased for and against certain politicians and topics. This presupposed bias has been the source of much controversy. It is often mentioned in public debates and exploited by politicians. This paper proposes an empirical approach to explain what partisan media bias is, how it operates, and how to measure it.
We argue that partisan media bias occurs primarily at the organization level, i.e., the impact of the media organization on news content. We discuss the ways that owners, editors and individual journalists impact partisan bias through the shaping on article content and its final physical placement. We also explain how counter factors limit and shape the way that partisan bias finds its expression. These factors include market competition; the context sensitivity of newsworthiness, including time-specific, issue-specific and event-specific considerations; and journalistic norms and ethics.
In terms of our research approach, we argue that partisan media bias is expressed in the way that different news outlets cover the same political story within the same timeframe relative to one another. This approach rests on the assumption that there are professional routines and objective media considerations that guide the work of journalists who come from different ideological backgrounds. Given that these considerations are largely similar for equivalent market competitors, the differences in coverage between such outlets are attributed to pressures and interests on the organizational level.
We present and test several possible measures of partisan media bias: volume bias, front-page bias, page-number bias, size bias, visual bias, opinion spectrum bias, and description bias. We choose a hard test case for the study: the social justice protest movement in Israel, 2011. This story enjoyed high public approval from across the political spectrum. This Occupy movement focused on a widely shared problem of high living expenses, and the protestors distanced themselves from all parties and politicians. It would be arguably more difficult to find partisan bias among general-market newspapers given this wide public support.
We analyze 1,556 news products from the four main newspapers in Israel. We find that Israel Hayom, the pro-Netanyahu newspaper, was an outlier relative to its three market competitors in every measure. Our results also suggest that individual (or single) measures of partisan bias may either not reveal partisan bias or may even yield inaccurate or false results.
We analyzed the relations between the measures and description bias. We found that newspapers appear to prioritize positive and negative articles about the story through their placement on the front-page, the size allotted to each, the attachment of visuals, and in the spectrum of opinions published by each outlet. Following this, we tested the statistical relations between the variables. We find that media outlets account for the variability in description bias (i.e., content) alongside two main mechanisms: front-page bias and size bias.
Our findings suggest that media organizations apply PMB to news production in two ways: 1. by selecting articles according to their content type; 2. by applying editorial discretion to emphasize or downplay articles according to their content. The latter is carried out primarily through decisions about front-page placement and sizing of articles.
Political Behavior 