Common weights in dynamic network DEA with goal programming approach for performance assessment of insurance companies in Iran

Abstract

Purpose Conventional data envelopment analysis (DEA) models permit each decision-making unit (DMU) to assess its efficiency score with the most favorable weights. In other words, each DMU selects the best weighting schemes to obtain maximum efficiency for itself. Therefore, using different sets of weights leads to many different efficient DMUs, which makes comparing and ranking them on a similar basis impossible. Another issue is that often more than one DMU is evaluated as efficient because the selection of weights is flexible; therefore, all DMUs cannot be completely differentiated. The purpose of this paper is to development a common weight in dynamic network DEA with a goal programming approach. Design/methodology/approach In this paper, a goal programming approach has been proposed to generate common weights in dynamic network DEA. To validate the applicability of the proposed model, the data of 30 non-life insurance companies in Iran during 2013-2015 have been used for measuring their efficiency scores and ranking all of the companies. Findings Findings show that the proposed methodology is an effective and practical approach to measure the efficiency of DMUs with dynamic network structure. Originality/value The proposed model delivers more knowledge of the common weight approaches and improves the DEA theory and methodology. This model makes it possible to measure efficiency scores and compare all DMUs from multiple different standpoints. Further, this model allows one to not only calculate the overall efficiency of DMUs throughout the time period but also consider dynamic change of the time period efficiency and dynamic change of the divisional efficiency of DMUs.

Goal programming to evaluate the profile of the most profitable insurers: an application to the Spanish insurance industry

“…(2017) have used GP to solve a dynamic multiobjective linear integer programming model to optimally distribute an insurance firm's advertising budget among five different products and Gharakhani et al. (2018) have used a GP approach to generate common weights in a data envelopment analysis (DEA) model to measure efficiency scores of 30 non‐life insurance companies in Iran. Marcenaro‐Gutierrez et al.…”

“…In the early times of GP, we could already find applications like Gleason and Lilly (1977) who developed a GP model for use in insurance agency decision-making, or Lawrence and Reeves (1982) who presented a zeroone GP model for capital budgeting in a property and liability insurance company. More recently, Aggarwal et al (2017) have used GP to solve a dynamic multiobjective linear integer programming model to optimally distribute an insurance firm's advertising budget among five different products and Gharakhani et al (2018) have used a GP approach to generate common weights in a data envelopment analysis (DEA) model to measure efficiency scores of 30 non-life insurance companies in Iran. Marcenaro-Gutierrez et al (2010) and Luque et al (2015) use a similar procedure, involving econometric analysis and reference point methods, to evaluate workers' satisfaction and students' achievement, respectively.…”

Eco-innovation analysis of OECD countries with common weight analysis in data envelopment analysis

“…So, the flexibility of weights in traditional DEA models is a weakness that results in different weight vectors, and therefore, unreliable efficiency scores (Razavi Hajiagha et al, 2018). With this weighting scheme, many DMUs will be ranked as efficient (Gharakhani et al, 2018). The CSW models are extensively used to overcome this issue by producing a distinctive weight vector for all DMUs (Hatami-Marbini and Saati, 2018).…”

Common set of weights and efficiency improvement on the basis of separation vector in two-stage network data envelopment analysis

“…Pourmahmoud and Zeynali [30] represented nonlinear model to extract CSWs in NDEA, and avoiding the selection of zero weights evaluated the performance of the units. Gharakhani et al [13] proposed a new approach for seeking a common set of weights in dynamic network DEA models based on the goal programming (GP) technique. The proposed approach makes it possible to monitor dynamic change of the period efficiency.…”