In search of intra-industry bullwhips

Previous research (e.g., Cachon et al., 2007) has examined the prevalence and magnitude of the bullwhip effect across industries, measuring the bullwhip as the ratio of the variance in the stream of orders placed relative to the variance in orders received (in some cases surrogates are used for the order measures). Our contribution is to use the decomposition framework of Jin et al. (2016a) to look intra-industry, examining whether the bullwhip is created in shipping, manufacturing, or ordering - and exploring whether this varies across industries. Our analysis is based on monthly, industry-level U.S. Census Bureau data. We find (and begin to explain) significant differences across industries regarding the source of the bullwhip - be it in shipping, manufacturing, and/or ordering. We also examine how the bullwhip is impacted by the duration and end-point of the time aggregation interval; our results suggest that some managers may want to shift their reporting period.