Contents lists available at ScienceDirect Journal of Retailing and Consumer Services journal homepage: www.elsevier.com/locate/jretconser Using a productivity function based method to design a new shopping centre Rafael Suárez-Vega a,b,∗ , José Luis Gutiérrez-Acuña c , Manuel Rodríguez-Díaz b,c a Departamento de Métodos Cuantitativos en Economía y Gestión, Universidad de Las Palmas de Gran Canaria, Spain b Instituto Universitario de Turismo y Desarrollo Económico Sostenible (Tides), Universidad de Las Palmas de Gran Canaria, Spain c Departamento de Economía y Dirección de Empresas, Universidad de Las Palmas de Gran Canaria, Spain ARTICLE INFO Keywords: Productivity function Local MCI model Goods/services differentiation Affordability index ABSTRACT Previous studies of competitive location models for shopping centres have mainly focused on the commercial attraction, rather than considering a real cost function, which would require a comprehensive financial analysis of the shopping centre and of the stores to be installed in it. The aim of this paper is to develop a method to determine the optimal design of a new shopping centre based on the productivity function that is, taking into account both its attractiveness and the forecast costs. Thus, we propose an optimisation model that can be applied to determine an attractive commercial offer that ensures the economic viability both of the investment and of the maintenance and management of the shopping centre. The proposed method is eminently practical; the model constructed includes the variables that determine the probabilities of attracting demand and also those which influence the costs and profitability of the shopping centre. All the variables considered are in- terrelated in some way with the gross leasable area. The method described furnishes a tool that is flexible, adjustable to the urban characteristics considered, and which facilitates the decision-making process for a po- tential investor. Finally, an application of this methodology is presented. 1. Introduction The construction of a new shopping centre calls for a major long- term capital investment that must be recovered by means of the rents paid by the retail firms occupying the centre. The complex nature of such a construction project makes decision-making a significant but difficult process. Moreover, all these decisions are interconnected, be- cause it is impossible to analyse the entire project without taking an overall view, including the demand attraction, the commercial offer, the location, urban planning restrictions and economic viability. The demand attraction determines the extent to which a shopping centre is capable of producing enough sales to ensure its economic viability (Nevin and Houston, 1980; Mcgoldrick and Thompson, 1992). This factor depends on the competitiveness of the commercial offer presented and on the effectiveness of the centre's management (Howard, 1997). However, the offer made by a shopping centre is composed not only of its gross leasable area (GLA), but also of services such as parking, gardens and leisure areas (Adkins Lehew et al., 2002; Oppewal and Timmermans, 1999). Since Weber (1909) established the basis for locating services, various alternatives have been proposed in order to solve increasingly complex situations. Competitive localization models are one of the variants of these problems. In this case, one of the most important as- pects to take into account is the demand capture estimation for the facilities. Gravity models are widely used in the analysis of retail pro- blems (some reviews can be found in Berman et al. (2009) and Eiselt et al. (2015)). These models are based on the idea that individual movements between points are inversely proportional to the distance between them. Huff (1964) proposed a retail model in which a facility's attraction for a customer is inversely associated with its distance, and positively with its size. This model was later generalised as a multi- plicative competitive interaction model (MCI) to take into account additional characteristics of the facility and to better define its attrac- tion. MCI models are widely used because the weights of the char- acteristics defining the facility's attraction can be easily estimated using ordinary least squares following the transformation proposed by Nakanishi and Cooper (1974). As Ghosh (1984) pointed out, parameters involved in the MCI model can be affected when socio-demographic differences exist in the study area. He faced the problem of locating a new store using a MCI model by adding for each demand point a new variable to absorb the spatial heterogeneity of the demand. Later, Suárez-Vega et al. (2015), in order to estimate the capture for a new hypermarket, used a local Huff model in which the parameters were locally estimated by geographically https://doi.org/10.1016/j.jretconser.2019.06.008 Received 3 October 2018; Received in revised form 11 April 2019; Accepted 12 June 2019 ∗ Corresponding author. E-mail addresses: rafael.suarez@ulpgc.es (R. Suárez-Vega), jose.gutierrez110@alu.ulpgc.es (J.L. Gutiérrez-Acuña), manuel.rodriguez@ulpgc.es (M. Rodríguez-Díaz). Journal of Retailing and Consumer Services 51 (2019) 176–185 0969-6989/ © 2019 Elsevier Ltd. All rights reserved. T