DEA as an Operations Research based linear programming approach for evaluating the relative performance of homogenous Decision Making Units (DMUs) is applied to Swiss shopping centers. Output-to-input efficiency ratios à as for example the sales productivity à incorporate one output (sales) and one input (sales area). Peer group comparisons (efficiency rankings) are difficult if multiple inputs and / or multiple outputs of different kind of data (quantitative, qualitative, categorical etc.) or measuring units (CHF, m2, %) are to be considered.
That is where DEA shows its advantages: DEA simultaneously handles multiple input factors (e.g. sales area, parking lots, OCR) and multiple output factors (e.g. sales, sales productivity, customer satisfaction) in a single efficiency measure à without prior fixing the factor weights. Furthermore, DEA helps in evaluating the DMUs (reference DMUs or benchmarks) that inefficient DMUs could refer to in order to improve efficiency (input reduction and / or output increase).
The literature review focuses on real estate DEA applications and emphasises property types whose efficient operation is crucial for valuation à as for retail properties or for properties run by operating companies (hospitals, hotels, shopping centers etc.).
The empirical analysis implements the Charnes-Cooper-Rhodes-Model (CCR model), the Banker-Charnes-Cooper-Model (BCC model) and the Additive model, and combines different shopping center performance drivers. Sales area and sales are strong performance drivers. Ratios as factors show new insights, and a DEA model including ratios could be an alternative to the widely used sales productivity. By differentiating the factors between food and retail sales and food and retail sales area, respectively, the efficiency of the shopping center sector mix is assessed.
Practical implications using DEA as a benchmarking or rating tool à for example in a Management Information System (MIS) à are given.