Abstract

Nationwide Newspaper Coverage of Homosexuals in The Boy Scouts of America:
A Community Structure Approach

Using a "community structure approach," tested in nationwide studies by Pollock and colleagues (1977, 1978, 1994-2002), this study explores systematic links between community demographic characteristics and newspaper reporting on homosexuals in the Boy Scouts of America (BSA), especially the Supreme Court decision permitting gay exclusion.

All articles over 150 words on the topic published between January 1, 1998, and March 1, 2001, were selected from a national cross-section sample of 21 newspapers, yielding 322 articles. A single score, the Media Vector, was calculated to combine article "prominence" as well as reporting direction (favorable, balanced/neutral, or unfavorable). Media Vector coefficients (from .256 to -.188) demonstrated clear national variation.

Four major hypotheses were confirmed, the first three most strongly by subsequent regression and factor analysis: 1) A "buffer" hypothesis, predicting that high percentages of privileged groups in a city correspond with reporting favorable to human rights claims-of gays in this case [% family income over $100,000 (r = .671; p = .000); % college educated (r = .593; p = .002)]; 2) a "stakeholder" hypothesis [gay market index (r = .599; p = .002); political partisanship - Democrats (r = .596; p = .002), Republicans (r = -.602; p = .002); religious involvement - Bible/devotional readings (r = -.588; p = .003), Catholics (r = .492; p = .012)]; 3) a "media access" hypothesis [newspaper circulation (r = .477; p = .014); number of cable stations (r =.426; p =.027)]; and 4) a "vulnerability" hypothesis [ % below the poverty line (r = -.474; p = .015). Factor analysis regression found the first three hypothesis clusters accounted for 56% of the variance.