Decision support systems for target setting and resource allocation in multi-unit and multi-level organisations using data envelopment analysis

This thesis is concerned with the development of decision support systems for determining performance targets and allocating resources in multi-unit organisations. These organisations are organised into networks of decision making units (DMUs) that seek to satisfy demand for services in the public sector or attract demand for services in the for profit making sector. Mathematical programming methods in general, and data envelopment analysis in particular are the methods chiefly used throughout the thesis. The decision support systems sought to address two distinct problems faced by multi-unit organisations. The first is concerned with the allocation of recurrent type of budgets to decision making units which use their resources without interference from their headquarters. This type of problem is called a-posteriori decision support and it is addressed by developing a framework of effective target setting. Data envelopment analysis models are developed for setting targets at the DMU and the global organisational levels. Two target-based resource allocation models are then developed seeking to encapsulate alternative organisational structures and objectives of resource allocation, namely equity, effectiveness and efficiency. The second case concerns problems where the allocation of resources is made directly to prespecified DMUs. This problem is called a-priori decision support which includes a phase of managerial diagnosis and planning. In the diagnostic phase performance targets for different management tiers are assessed, and systematic procedures of micro level benchmarking are developed. In the planning phase targets for improving the scale size of individual units are assessed, the long & short run viability of the network of outlets is examined and, fmally, the marginal impacts of past investments on the performance of DMUs are investigated. The two phases of the decision support system would aid management in making decisions regarding the future of individual DMUs (e.g. investment, expansion, divestment). Application of the method to a network of 154 public houses is incorporated throughout the relevant chapters of the thesis.