Revisiting the relationship between corporate social and financial performance through a multivariate perspective

Resum

This PhD dissertation aims to analyse the possible relationship between the degree of Corporate Social Performance (CSP) and Corporate Financial Performance (CFP) of some business organisations. To this end, there will be an exploration of the different measurements used in the literature on CSP and CFP (sustainability disclosures, social audits and reputation rankings for CSP, and market-based and accounting-based measures for CFP). The dissertation introduces new methodological approaches for measuring the two constructs (CSP and CFP). The aim is to reduce, as far as possible, the shortcomings found in similar previous works. Although a specific objective is presented in each chapter, the overall assessment of these will determine whether the overall research objective has been achieved. The PhD dissertation is divided into five chapters. Chapters one to four consist of research essays closely related to the research question of the dissertation. These are rounded off by Chapter five, which gathers together the main conclusions, contributions and limitations identified, in addition to possible alternatives for future research in the field. The first chapter offers a critical view of the current situation and the evolution of the sustainability reporting on a global level, with special reference to Spanish organisations. This is done through an in-depth review of Social and Environmental Reports (SER) from their beginnings in the 1970s to the present. There will be an exploration of the various methodologies that have been used to set a standard for SER disclosed by different organisations, paying special attention to the provided by the Global Reporting Initiative (GRI). The aim of this chapter is to assess the possible relationship between the degree of CSP and CFP obtained by some Spanish organisations. It will evaluate stock market reaction to the disclosure of SER. The extent of corporate disclosure on matters of social concern is used to measure the degree of CSP (Herremans et al., 1993; Orlitzky et al., 2003, Wu, 2006; van Beurder and Gössling, 2008). An event study was applied in the empirical section, which ranges from 2003 to 2008. The second chapter aims to extend the debate on the hypothetical relationship between the degree of CEP and CFP of European companies. Over the last thirty years, research into the subject has not provided homogeneous results and work published on the subject presents differences concerning the relationship between the two constructs. However, the latest studies support a positive relationship between these constructs (Wu, 2006; van Beurden and Gössling, 2008). Most research in the field is based on measuring the CEP and CFP constructs using single-dimension indicators. One of the main contributions of the chapter is the application of an innovative, multivariate model to measure CEP (through social audits, CEP processes and observable outcomes, as suggested by Herremans et al., 1993; Orlitzky et al., 2003, Wu, 2006; van Beurder and Gössling, 2008) and CFP. A multivariate research method, Partial Least Squares (PLS), is applied to measure the CEP and CFP achieved by the different companies in the sample. Through the use of the PLS technique, the chapter aims to contribute to the current literature by mitigating some limitations identified previous works. It is also important to emphasise that this is the first time that a Structural Equation Model (SEM) has been applied in the field. The implications of this chapter are of great interest to both academics and company managers. They will be able to assess how environmentally-oriented strategic management could increase the firms' competitiveness in the mid-long term, in addition to raising their levels of CFP. The third chapter consists of an analysis of the possible relationship between the degree of CSP and CFP using a representative sample of European companies. In order to measure the level of CSP of these companies, CSP reputation rankings or CSP reputation indexes have been used (Herremans et al., 1993; Orlitzky et al., 2003, Wu, 2006). Annual reviews of the Dow Jones Sustainability Stoxx Index (DJSI-Stoxx) have been taken as a proxy of CSP levels. In accordance with similar research in the field (Curran and Moran, 2007), the announcements of companies being included or excluded from the DJSI-Stoxx have been taken as proxies for high and low levels of CSP, respectively. Furthermore, the chapter provides other event to be analysed which has never been taken into account by any study until now. The non-exclusion announcements have been taken as a proxy for obtaining high levels of CSP. One of the contributions of this chapter is that the empirical analysis has been carried out between 2003 and 2007, a period in which market agents, especially ethical investors, gained a representative degree of knowledge about the sustainable stock exchange indexes, an aspect not considered in studies to date. The chapter also contributes to the literature by proposing a Multivariate Regression Model (MVRM) as an event study methodology for the empirical study, with the aim of reducing, as far as possible, some of the shortcomings found in previous research, mainly focused on the market model, also known as the FFJR model (Fama et al., 1969). The fourth chapter introduces the main features of the sustainable stock exchange indexes worldwide and, specifically, in Spain. Unlike the first three chapters of the dissertation, which focus on analysing the relationship between CSP and CFP levels, the aim of this chapter is to evaluate the risk levels experienced by the only sustainable stock exchange index in Spain (FTSE4Good-IBEX) and to compare it with the risk experienced by its principal benchmark (IBEX35). This chapter complements the results of the three previous chapters, enabling not only the assessment of the level of CFP attained by portfolios investing in sustainable stock exchange indexes, but also the identification of the risk levels to which the portfolios are exposed. After reviewing the literature available on the subject, it is noted that there are few published works analysing this topic (Hoti et al., 2005; Hoti et al., 2008). Thus, this chapter contributes to the small amount of research carried out into return-risk parameters associated with sustainable stock exchange indexes. Another, very important aspect of the chapter is that the empirical analysis is made during the financial crisis and economic recession in which economies worldwide, including the Spanish one, are involved. The empirical study has been carried out by means of univariate and multivariate GARCH models to measure the conditional volatility associated with the FTSE4Good-IBEX and the IBEX35. The implications of this chapter should prove of great interest both to academics and practitioners alike, as it proposes a new portfolio diversification framework using non-conventional financial assets, such as sustainable investment. Finally, the fifth chapter summarizes the main conclusions and contributions of the four previous chapters. This dissertation argues that, in order to obtain a more robust measurement of the CSP construct, researchers must not rely on a measure of performance relating to only one stakeholder group. Although this consideration may frequently be impossible to put into practice (because of various restrictions when collecting data about all the stakeholders of companies), the dissertation emphasizes the need to clarify what dimension of CSP the research want to measure. As an example, Chapter 3 called the measurement of environmental performance Corporate Environmental Performance instead of CSP because the environment is one of the many stakeholders that the companies interact with. One of the main contributions of this PhD dissertation is that innovative multivariate methodological approaches are developed for measuring both the degree of the CSP and CFP constructs. With the use of these multivariate models, researchers might be able to make more realistic measures of these constructs. Another aspect that has not helped to develop the scientific and academic knowledge in this field is the fact that there is no agreement about the direction of the possible relationship between the CSP and CFP constructs. The direction is of great importance, since the decision about this link direction has an impact on the theoretical framework to be considered. The analysis of what link direction is more appropriate between the two constructs previously mentioned is not an objective of the present PhD dissertation. It is clear that this decision should be made according with the objectives of the research. It is interesting to note that the Slack Resources Theory seems to be adequate as a theoretical framework for works analysing the impact of CFP on the improvement of CSP levels. This is not the case of this dissertation because the research question seeks to determine the significance of the opposite relationship (how CSP levels influence CFP). For the last thirty years, research about the impact of CSP on CFP levels has been carried out considering a broad range of theoretical frameworks: Legitimacy Theory, Agency Theory, Stakeholder Theory (mainly in its instrumental dimension). In this PhD dissertation, Stakeholder Theory has been chosen as the global theoretical framework, although some chapters emphasize others such as the Instrumental Stakeholder Theory. One of the biggest difficulties identified in this research is the fact that, nowadays, Stakeholder Theory does not provide widely-accepted tools for carrying out empirical tests. During the last twenty years, the most important evolution of Stakeholder Theory has been focused on its normative dimension, because some authors consider Stakeholder Theory as a 'philosophical approach'. Ethicists, who have provided many of the principles of the Stakeholder Theory, have not been concerned with developing a scenario in which Stakeholder Theory could be empirically tested. Research about the descriptive and instrumental dimensions of Stakeholder Theory has increased significantly in the last ten years. The development of the instrumental and descriptive dimensions of Stakeholder Theory has been made in order to configure a true theory and to be able to test in practice the normative foundations of Stakeholder Theory. However, nowadays, it is not clear that Stakeholder Theory provides hypotheses suitable for empirical testing. Instrumental Stakeholder Theory proposes a more flexible framework and it is possible to establish causal hypotheses, thereby making it possible to test the implications of different business strategies. Although Instrumental Stakeholder Theory should be an adequate framework for testing research questions like the one proposed in this dissertation, it is necessary to further develop the Superior Stakeholder Theory in order to mitigate the limitations identified in its predecessor (Stakeholder Theory). Superior Stakeholder Theory should be easier to use, more efficient, simple and even instinctively pleasing in order to give more significance to the results obtained in works in the field, such as those provided in this PhD dissertation. The global analysis of the results obtained in this PhD dissertation confirms the existence of a positive and significant causal relationship between the level of CSP and CFP of European companies, especially in Spain, and provides an answer to the research question proposed in this PhD dissertation. This dissertation emphasizes the shortcomings shown by the applicability of the Neo-classical Theory in the current business agenda. Although Neo-classical theory has been adequate to drive business for decades, no truth lasts forever and business is too dynamic to sustain an obsolete theoretical framework from many years. The results obtained in this PhD dissertation show that there is a significant and positive relationship between CSP and CFP. The main implication of this finding is that corporations need to be managed through a more flexible model that considers the claims of their stakeholders, disregarding the shareholder approach provided by the Neo-classical Theory. Corporations that consider Corporate Social Responsibility criteria or sustainable management policies as a cost might not obtain better levels of CFP and improve their competitiveness in the mid-long term. This loss of competitiveness is made even clearer when comparing the success of the companies with others operating in the same industry, which are their potential competitors. These implications are of great interest for corporate management when establishing their strategic policy because the dissertation results show that aspects relating to sustainable management are increasingly considered by stakeholders and, especially, by investors when making their investment decisions.