Alternative statistical regression studies of the effects of Joint and Service Specific advertising on military recruitment

Three different regression approaches use a large database developed by the Wharton Center for Applied Research (WCAR) to study the effects of Joint versus Service Specific advertising on military recruitment. ( Here 'Joint' refers to advertising designed to serve recruitment for all four services simultaneously. Service Specific refers to advertising administered separately by each of the four services.) These regression approaches and the data and models are examined with special reference to US Army recruitment. The WCAR study led to a recommendation to replace Service Specific with Joint advertising. This recommendation was called into question by the RAND Corporation in its study that used a different regression approach. A third study that combines regressions with data envelopment analysis (DEA) is presented in this paper. This study utilizes recently developed methods based on DEA which, when incorporated in the regression, make it possible to distinguish between efficient and inefficient performances. The resulting regression yields results that show Joint advertising to be not only less efficient but also to attract potential recruits from the Army to other services. Implications for further research are set forth, which can also cast light on commercial practice by regarding Joint as a type of 'category advertising' and Service Specific as a type of 'brand advertising'.