A LASSO-BASED DEA FOR ECO-EFFICIENCY PERFORMANCE ASSESSMENT FOR THE GLOBAL FOOD AND BEVERAGES INDUSTRY

Abstract

Sustainable food systems are essential to secure food and nutrition for society and preserve the economic, social, and environmental aspects. A sustainable food system has become a significant demand for survival due to the dramatic growth of urbanization and global economic and health disruptions. Recently, food supply chains of global industries are encountering economic and environmental challenges, resulting in a significant decrease in their eco-efficiency performance. Therefore, there is a great need to identify possible reasons and their potential relationship across the eco-environmental pillars of sustainability. To this end, this thesis proposes an approach for eco-efficiency assessment integrating both the Least Absolute Shrinkage Squared Operator (LASSO) with the Data Envelope Analysis. The new approach constitutes two stages. First, the LASSO regression is applied to reduce the dimension-space of the eco- and find the relative weights estimates for each indicator in the new dimension. Second, the DEA is used to estimate the eco-efficiency ratio for all the food industries. The mathematical and operational procedures of the new approach are demonstrated using the economic and environmental footprints of 30 food and beverages industries in the USA. The new strategy is expected to provide food and beverage industries with a powerful tool for assessing their contribution toward achieving sustainable development goals.