Media consumption and crime trend perceptions: a longitudinal analysis

Abstract

For over two decades, despite the downward crime trend, the American public has persisted in believing crime is on the rise. Cultivation theory holds that the media is responsible for the public’s crime trend perceptions. Previous cultivation studies heavily rely on cross-sectional data, which may lead to spurious conclusions due to reverse causation and omitted variable bias. This study aims to address these issues by utilizing longitudinal analyses. Drawing on three waves of the 2008–2009 American National Election Survey, we test the cultivation hypothesis using traditional OLS, OLS with lagged crime trend perceptions, fixed effects, and dynamic panel models. Newspaper and TV news consumption are related to crime trend perceptions in traditional OLS models. In other models, media consumption is not related to crime trend perceptions. The results do not support the cultivation hypothesis. It is likely that the cultivation effect of media has been overstated in the previous cross-sectional research.