RECESSION HEADLINE NEWS, CONSUMER SENTIMENT, THE STATE OF THE ECONOMY AND PRESIDENTIAL POPULARITY - A TIME-SERIES ANALYSIS 1989-1993

During the 1992 American presidential election, the media were accused of portraying the economy in a negative light, with both economic and political consequences for the country. Such criticism was based on assumptions concerning relationships among four variables: economic news coverage, public perception of the state of the economy (consumer sentiment), the actual state of the economy, and presidential popularity. This paper seeks to examine the relationships among all four variables in a way that accounts for inherent time series characteristics of the data including: potential non-stationarities (or tendencies for the series to drift over time) and co-movements among the series. Hypotheses concerning the nature and direction of influence among the four variables are proposed and time series analyses are conducted to test each hypothesis. We use recession-related headlines from the New York Times to represent economic news. Each series is analyzed to isolate its principal characteristics, and tests for co-movement (formally, cointegration) between the series are conducted. Vector autoregression is used to model the joint determination of the series, and tests for Granger causality are conducted. The results show some causal evidence for a media effect: recession headlines were a significant prior influence on the determination of consumer sentiment in this study. There is some limited evidence of an adversarial press effect, wherein the president's growing popularity rather than real world economic conditions appears to have led an increase in the number of recession headlines.