ABSTRACT
A study of the universal health insurance mandate compared hypotheses linking newspaper coverage and different city characteristics using the “community structure” approach to research, as elaborated by Tichenor, Donohue, and Olien (1973, 1980), Demers and Viswanath (1999), Pollock (2007) and others. [See November/December (14:6) special issue of Mass Communication and Society on “community structure” scholarship.] Sampling all articles of 250 words or more from 22 major metropolitan newspapers in a national cross-sample from February 10, 2007 to November 1, 2009, the resulting 262 articles were analyzed using content and statistical analysis. A “prominence” score, including the article’s headline size, placement, length, and presence of photographs, was combined with the article’s “direction” (favorable, unfavorable, or balanced/neutral coverage of the issue) to produce a “Media Vector.” The cities’ Media Vectors ranged from 0.5761 to −0.1630; twenty of the 22 cities portrayed universal healthcare favorably. Pearson correlations revealed that characteristics falling under the “stakeholder hypothesis” (percent Hispanic: r=0.545, p=0.004) and the “buffer” hypothesis (percent professional: r=0.401, p=0.032) had significant relationships with favorable coverage of universal healthcare. Initial regression analysis found that percent Hispanic accounted for almost 30% of the variance, percent professional almost another 6%. A factor analysis revealed five significant factor clusters, and regression of factors with Media Vectors yielded two variables accounting for approximately 35% of the variance: Hispanic/Low Healthcare Access (27.5%) and Public Healthcare/Vulnerability (7.2%). Thus, these clusters reveal that “marginality” (Hispanic) and low healthcare access demographics had the greatest influence on newspaper coverage of universal healthcare.
