Bidirectional Causality for Word of Mouth and the Movie Box Office: An Empirical Investigation of Panel Data

Abstract

ABSTRACT Word-of-mouth (WOM) is informal communication between consumers about products and services. By using text mining techniques, WOM measured for volume and valence at the movie box office in Taiwan. A simultaneous regression of a panel Granger causality test for WOM and corresponding film performance was performed. The empirical results show that dynamic causality only runs from box office to WOM volume in the short run analysis, and the causality between WOM volume and box office is bidirectional in the long run analysis. Potential customers attend movies because of WOM in the short run, but in the long run, box office forms a signal and creates WOM. In WOM valence analysis, causality runs from positive critics to the box office in the short run, and it runs from negative critics to box office in the long run. WOM providers may need to develop different strategies for encouraging WOM behavior among their users. The implication for film managers and marketers is that a reliable way to affect box office is to stimulate positive critics in the short run and negative critics in the long run.