Assessing Partisan Bias in Political News: The Case(s) of Local Senate Election Coverage

Abstract

Many studies of partisan bias in political news employ balance as a baseline. That is, the party/candidate receiving more or better coverage in any given source is automatically deemed the beneficiary of favorable treatment by the source. A study employing the balance baseline potentially exaggerates the amount of meaningful partisan bias in the source, however, for failure to control for nonpartisan, non-ideological news judgment criteria. This study models variation in the relative amount and tone of coverage received by candidates in 95 content analyses of newspapers' Senate election coverage from 1988–1992. This enables a direct test of the relative power of partisan and structural (nonpartisan, news-judgment-driven) biases in explaining the slant of election coverage. While news-organizational factors are found to dominate the amount model, a modest amount of residual slant toward the Democratic candidates remains in the tone of coverage, controlling for structural bias.