Comparing groups of units through composite indicators in a non-convex approach: corporate social responsibility for the food and beverage manufacturing industry

Abstract

This paper compares the performance of groups of units by composing indicators of corporate social responsibility (CSR) from an efficiency and productivity perspective, applicable across various industries. From a methodological perspective, our work extends the traditional input-oriented Benefit-of-the-Doubt (BoD) model in the multiplier form, by first adapting it to accommodate the non-convexities of the production set, and second, by innovatively applying it to compare indicators across groups of firms. This adaptation, pioneered in our study, leverages the framework previously established in the literature to address more complex scenarios. From an empirical perspective, we contribute by comparing the efficiency and productivity in CSR activities of food and beverage companies across regions of Europe, the United States and Canada, and Asia–Pacific over the period 2009–2018. The paper reveals that USA-Canadian firms tend to perform best with respect to CSR, followed by European firms, and that Asian-Pacific firms achieve the worst efficiency and productivity results. The study also shows that regional catching up in CSR productivity occurred over the analyzed period.