Proper modelling of industrial production systems with unintended outputs: a different perspective

The question of how to properly model production systems with unintended outputs has proven both controversial and of particular interest to the productivity and efficiency community. The paper explains why some of the arguments put forward in these controversies are hardly convincing for industrial and other processes. Among other things, there is a lack of clear conceptual labelling of the different types of joint production, especially coupled production, which is the main source of undesirable and other unintended outputs, unless neglected. It is largely ignored that the desirability of such by-products may depend on the quantity produced. This is also true for reduction processes such as waste incineration or end-of-life vehicle dismantling, which in turn generate new unintended outputs. As a rule, industrial material and energy balances are modelled implicitly. Koopmans' activity analysis is the standard approach in modelling production systems with undesirable outputs in the literature of business economics on sustainable production and supply chain management. With data envelopment analysis (DEA), instead of entire production possibilities, it is sufficient to know only certain local properties in the relevant range of input and output quantities of the observed activities. This lowers the challenge to verify their empirical validity.