An inverse DEA model for intermediate and output target setting in serially linked general two-stage processes.

This article focuses on developing and applying an inverse data envelopment analysis (DEA) model for a serially linked two-stage production process under constant returns to scale technology. The proposed inverse DEA model determines feasible intermediate and output targets for a hypothetical input-augmented decision-making unit (DMU) to maintain a pre-specified relative efficiency level without altering the efficient frontier established by observed DMUs. The model is formulated as a multi-objective optimization problem solved via the ε-constraint method, enabling the identification of Pareto optimal solutions that reveal trade-offs among outputs and intermediates. Empirically, the model is applied to Australian superannuation funds, demonstrating how fund managers can use the output and intermediate targets to plan for fund growth while preserving efficiency at both overall and stage levels. The study highlights that including the hypothetical DMU with these targets in the observed set does not change the efficiency scores of existing DMUs, ensuring the practical feasibility and managerial relevance of the targets.