Testing a Political Economic Theory of the Media: How Were Steel Tariffs Covered?*

Abstract

Objective. The objective of this article is to test Strömberg's (2001) prediction that newspapers will devote more space to costs of tariffs than to their benefits, using the recent steel tariff issue as a test case.Method. A content analysis was performed on a sample of 123 New York Times (NYT) and 177 Wall Street Journal (WSJ) articles published between January 1 and September 10, 2002. The articles were coded sentence by sentence for pro‐, anti‐, and neutral‐tariff content. Comments or information deemed supportive of the tariffs, that is, touting their benefits, were coded +1, while those emphasizing the costs of tariffs were coded –1. A statement that was neutral with regard to the tariff issue was coded 0. The proportion of sentences coded –1 was then compared with the proportion coded +1.Results. The coders found 395 sentences devoted to the negative impacts of steel tariffs versus 124 sentences devoted to their benefits out of a combined total of 3,207 sentences. The results also reveal a mild “slant” toward free trade in the WSJ relative to the NYT. In the WSJ, the proportion of lines criticizing tariffs (0.11) exceeds the proportion supporting tariffs (0.02) by a wider margin than in the NYT (0.13 vs. 0.06).Conclusions. This article provides empirical support for Strömberg's (2001) political economy model of the media. The model implies that mass media will weaken the power of special‐interest lobbies relative to unorganized interests. Specifically, the data support Strömberg's prediction that more newspaper space would be devoted to the costs of steel tariffs—which are widely dispersed—than to their benefits—which are narrowly targeted.