Measuring the social responsibility of European companies: a goal programming approach

• Published date: 01 May 2019
• DOI: http://doi.org/10.1111/itor.12438
• Author: Francisco Guijarro,Juan A. Poyatos,Gabriel García‐Martínez
• Date: 01 May 2019

Intl. Trans. in Op. Res. 26 (2019) 1074–1095

DOI: 10.1111/itor.12438

INTERNATIONAL

TRANSACTIONS

IN OPERATIONAL

RESEARCH

Measuring the social responsibility of European companies:

a goal programming approach

Gabriel Garc´

ıa-Mart´

ıneza, Francisco Guijarroband Juan A. Poyatosc

aEconomic and Social Sciences, Universitat Politecnica de Valencia, Valencia, Spain

bI.U. de Matem`

atica Pura i Aplicada, Universitat Politecnica de Valencia,Valencia, Spain

cInstituto Valencianode Finanzas, Valencia, Spain

E-mail: gagarmar@esp.upv.es [Garc´

ıa-Mart´

ınez]; fraguima@upvnet.upv.es [Guijarro];

juanangel@voluntariadoyestrategia.com[Poyatos]

Received 15 December 2016; receivedin revised form 15 March 2017; accepted 18 May 2017

Abstract

Corporate social responsibility (CSR) can be measured by a number of different criteria, some of which

are similar to each other, while others can be manifestly contrary to the general tendency. This means

that some companies can obtain a good valuation in some criteria but a bad valuation in others, which

makes it difficult to assess the company’s overall CSR valuation. It is not easy to find a single measure

that covers all aspects of corporate social performance. This paper aims to estimate multicriteria CSR

performance through different models of goal programming and by taking into account all the dimensions

that make up CSR. An illustrative example shows the result of applyingthese models to a database composed

of 212 European companies, which enabled us to identify the most socially responsible group, regardless

of the approach considered in the construction of the multicriteria performance. The results show that

environmental and corporate governance dimensions are the most important elements in measuring this

performance.

Keywords:social behavior; EIRIS, goal programming; dimensionality reduction

1. Introduction

The financial crisis of 2008 highlighted the concern and growing interest of society about issues

related to corporate social responsibility (CSR). This has led researchers to focus on the analysis of

corporate social behavior, and as a consequence a significant number of papers on this subject have

recently been published (Aguinis and Glavas, 2012; Bilbao-Terol et al., 2012, 2013; Eccles et al.,

2014; Malik, 2015; Gasser et al., 2017).

Igalens and Gond (2005) point out that the diversity of the data sources used to measure social

behavior combined with the multitude of theoretical approaches has helped to create confusion

C

2017 The Authors.

International Transactionsin Operational Research C

2017 International Federation ofOperational Research Societies

Published by John Wiley & Sons Ltd, 9600 Garsington Road, Oxford OX4 2DQ, UK and 350 Main St, Malden, MA02148,

USA.

G. Ga rc´

ıa-Mart´

ınez et al. / Intl. Trans. in Op. Res. 26 (2019) 1074–1095 1075

on how to properly measure it. The authors confirm that although an attempt has been made to

find the best measurement system available for a given definition of the theoretical construct, most

studies have been found to include data in a nonsystematic way, which means relatively distant

approximations to the concept of CSR have been used.

The review of empirical measurements found in the literature raises serious doubts about the

accuracy with which they reflect the concept of social performance. This has led several authors

to suggest a combination of different measurement systems (Rowley and Berman, 2000). The lack

of relevant and accurate data measuring corporate social performance is another limitation of

numerous empirical studies, and many of the data and measurement systems used have not been

adequately tested.

Wood (1991) maintains that in order to review the methods for the measurement of any variable,

it is first necessary to take into account the nature of what one wants to measure. This author

lists the main topics to be measured by CSR: environmental assessment, stakeholder management,

clients and consumers, employees, suppliers, criminal conduct, among others. However, Rowley

and Berman (2000) argue that social performance measures composed of several aggregate

dimensions are often misused. Those who use them do not usually propose a theoretical model that

previously correlates these dimensions with each other. According to the authors, the measurement

model can be examined through statistical techniques such as factor analysis, which reduces

the dimension of the problem. Instead of using all the possible variables reported by different

CSR databases, we should be able to extract the most important and representative of corporate

performance.

This paper aims to provide an objective, general, and unifying method of measuring CSR using

goal programming (GP), a well-established multicriteria technique. This methodology makes it

possible to construct models thatsatisfy conflicting criteria, which in the scope of this study translates

into models that allow simultaneous consideration of the different dimensions that make up social

responsibility.

A number of studies incorporatesocial responsibility criteria into the decision-making area, and in

some cases GP models are proposed to both design the objective function and the constraints of the

decisional problem. Tsai and Hsu (2008) developed a model foroperationalizing social responsibility

programs for air transport management within the context of constrained physical resources. The

authors combine two classical multicriteria techniques: the analytic hierarchy process (AHP) and

GP. In a similar way, Tsai et al. (2010) propose an integrated approach to help the hospitality

industry to solve the problem of selection decisions and cost evaluation of CSR initiatives. Tsai

et al. (2009) propose a novel model that is applied to the case study of a small enterprise. The

model combines the decision-making trial and evaluationlaboratory (DEMATEL)method, analytic

network process (ANP) and zero-one GP (ZOGP) in evaluating socially responsible investment

(SRI) selection procedures. DEMATEL helps companies to identify the most important criterion

or the one that affects other criteria the most. ANP helps to determine the priority weights among

the alternative stocks,while the ZOGP model helps organizations to use resources without exceeding

their constraints.

Zhang (2016) considers the economic, environmental, and social implications of the tourist

industry. A combination of the (ANP and GP is used to determine the relevant variables for the

development of tourism in Tibet, simultaneously considering economic, environmental, and social

goals.

C

2017 The Authors.

International Transactionsin Operational Research C

2017 International Federation of OperationalResearch Societies