Media Consumption and Crime Trend Perceptions: Comparing Results from OLS, Fixed Effects, and Dynamic Panel Models

• In Event: Rethinking Media and Perceptions of Crime

Abstract

Objectives: The American public has persisted in believing that crime is on the rise, despite the fact that U.S. crime rates have been dropping steadily. Cultivation theory holds that media is responsible for the public’s perception of crime trends. This argument, though long-standing, lacks definitive proof because of omitted variable bias and reserve causation. This study aims to fill in this research gap and examine the causal relationship between news consumption and crime trend perceptions.
Methods: We draw on three waves of the 2008-2009 American National Election Survey (ANES). We first employ ordinary least squares (OLS) and fixed effects methods to assess if greater exposure to crime news causes perceptions that crime trends are worsening. We then take advantage of the longitudinal character of the ANES data to employ generalized method of moments (GMM) dynamic panel estimators.
Results: Newspaper and TV news consumption appears to be related to crime trend perceptions when OLS is used. However, media consumption variables have no statistically significant relationship with crime trend perception when in the fixed effects and GMM dynamic panel models.
Conclusions: The results provide little evidence for the argument that media exposure influences the public’s perceptions of crime trends.

Authors

• Luzi Shi, University at Albany, SUNY

• Sean Patrick Roche, University at Albany,SUNY

• Ryan M. McKenna, Drexel University