Page 279 - Comparing Political Communication Theories, Cases, and Challenge

P1: JZL/KDD/KAB P2: JZZ
0521828317c11.xml CY425/Esser 0521828317 June 8, 2004 22:9

Political News Journalists

Rank:
a photo showing dark smoke emerging from a plant’s smoke
stacks
a photo of the chemical industry spokesperson at the press
conference called to protest the new regulations
a graph that shows the decline in air pollution over the last
ten years
a graph showing the projected improvement in air quality as
a result of the new regulations

Asinthiscase,eachsituationdealtwithanissuethatisasourceofpartisan
conflict. In addition, seventeen of the twenty-four news decisions were
purposely framed in a way that favored a partisan view (the other seven
were purposely neutral in tone). For example, the proposed headline in
thepollutionsituation(“ChemicalIndustryPredictsHighCostandLittle
Effect from New Regulations”) was presumed to have a right-of-center
bias because it conveyed the chemical industry’s view of the situation
rather than the regulatory agency’s perspective. On the other hand, the
last of the visual options (“a graph showing the projected improvement
in air quality as a result of the new regulations”) highlighted the expected
benefits of the new regulations and hence was assumed to reflectaleft-
of-center and proenvironmental protection bias.
In developing the survey’s four news situations, we aimed to con-
struct decision options where the partisan bias was subtle. We sought
to create plausible options that the respondents might actually face in
the newsroom rather than blatantly partisan options that a professional
journalist would reject out of hand. In this way, if the respondents ex-
pressed a preference for options that were slanted toward their point
of view, we could reasonably infer that partisanship had influenced the
decision.
We correlated journalists’ decisions with their partisanship as mea-
sured by our Left-Right scale (see previous). Because of the small size
of the samples (the average n is about 250 respondents), we examined
the significance of the aggregate distribution of decisions. Each of the
seventeen news decisions can be compared to the toss of the coin. If
the relationship between partisanship and news decisions is random, a
single test is as likely to yield a negative correlation as a positive one.
On the other hand, if partisanship affects news decisions, a single test
is more likely to yield a positive correlation and most of the seventeen
decisions will be positive in direction. The probability of a particular

259