Benchmarking corporate social responsibility within Spanish companies.(Report)

Abstract In this paper, the authors focus on the influence of corporate social responsibility (CSR) upon business profitability. In order to be capable of working with homogeneous data, the authors' starting point is to use the criteria defined by PricewaterhouseCoopers' work on the subject, and published by the Spanish journal Actualidad Economica. In this work, an index was created which assigns between one and five points to the companies depending on the importance given by them to CSR. The CSR measurement published by the Observatorio de la RSE will also be considered. In order to measure companies' profitability, this paper will take into account their return on equity (ROE) and return on assets (ROA) of 2005 and 2006. The authors' purpose is to demonstrate that the relationship between CSR and business profitability is neutral.

Keywords Corporate Social Responsibility (CSR) * Return on Equity (ROE) * Return on Assets (ROA)

JEL M00

Introduction

In the recent past, many authors have collected information on corporate social responsibility (CSR) and tried to analyze it in a homogeneous way. There has been a lot of information, both qualitative as quantitative, obtained in many cases by people directly responsible for CSR within their companies.

Companies should try to become socially responsible in every decision made: from the product design to its selling and distribution; from personnel recruitment until departure from the company; from raw material selection as well as the kind of energy used, facilities lay out, to final product; from getting financial resources to its investment, and so forth. Also the information to be given to third parties has to be taken into account from a CSR point of view.

There are many approaches to CSR (1) and, for this reason, there is still a long way to go to achieve a common and universal [CSR] criteria. Moreover, it is necessary to establish a universally accepted model, which would also be measurable. (2)

In Spain, the 35 companies included on the IBEX index must report annually on which good corporate governance practices have been applied (3). Financial markets including these criteria and concepts, such as social responsible investments (4) appear. Furthermore, measurements on social behavior and the environment (5) also exist, and finally, citizenship pressure for CSR is growing fast.

Since the end 2006, the Asociacion Espanola de Contabilidad y Administracion de Empresas (AECA) has been developing a specific language which will be useful in order to compare annual reports on CSR published by Spanish companies. "The goal is to make measurable and comparable the CSR reports written down by companies," as AECA (6) sources explain.

It is also important to highlight the importance on getting involved with societal problems, respecting the environment, being socially responsible concerning the different groups related to the company and so on, may have for the company. These aspects might turn into a truly competitive advantage as they could be seen as a market differentiation.

Contrast Between CSR and ROE-ROA

The authors will try to establish which kind of relationship may exist between CSR and business profitability, measured taking into account the return on equity (ROE) and the return on assets (ROA). (7) With respect to CSR, the work developed by PricewaterhouseCoopers for the Spanish journal Actualidad Economica has been taken as a reference. In this study, CSR is evaluated in a range between 1 and 5. The resultant index has been built up through the analysis of stated goals and the results achieved. (8) The initial hypothesis is to establish that CSR has a direct positive influence on companies' results. Statistical tools used for analyzing the data are described below.

Statistical Background

Many authors have analyzed the influence that CSR has on capital cost, profitability, and results. (9)

Margolis and Walsh (2001, 2003) analyzed 95 studies and concluded a positive correlation between social and financial results is plausible. Figure 1 also shows that negative correlations, non-correlated and mixed results also exist. This study is available from: http://www.schlange-co.com/fileadmin/user_upload/Studien/CR_CostOfCapital_Schlange_Co.pdf.

[FIGURE 1 OMITTED]

Statistical Analysis

The authors use data published by the Spanish journal Actualidad Economica for 2005. They later attempt to verify if the same relationship maintains (or can be extrapolated to) through 2006.

The purpose of this paper is to show how the different CSR indexes (Actualidad Economica (10) and the Observatorio de la RSE) relate to business results. The main goal is to determine if better business results imply a higher CSR index or if it is possible to determine if higher CSR levels improve business results. The goal of this paper is to analyze the relationship between RCS(05) (the Actualidad Economica" classification of corporate social responsibility and ROE(05) or ROA(05) data from http://www.infomercados.com. The RCS(05) variable refers to 123 companies selected by "Actualidad Economica" (11)

The RS index is also taken into account; these classifications are made by the Observatorio RSE (committee by UGT, a Spanish trade union). According to the authors of this index (Perdiguero et al. 2007), three categories were established to create the index: transparency, sustainability importance, and managing corporate social responsibility. In order to do this, we analyzed the management results of seven big areas in the companies, evaluated twenty types of indicators and finally, used a numerical range between 0 and 100 to classify the companies. (12)

The Actualidad Economica Index

The first approach is an illustration to compare the values of ROE(05) with CSR(05) (Fig. 2).

[FIGURE 2 OMITTED]

Figure 2 shows the distribution of the majority of the 123 companies, with the six possible values of CSR (0,1,2,3,4,5), which is rather concentrated between ROE values of-0.35 and 0.25.

In Fig. 3, a box plot with the median, quartiles, maximum, minimum and average of the variable is shown simultaneously within each group. Therefore, a greater level of information on the two variables analyzed will be obtained.

[FIGURE 3 OMITTED]

Thus, by illustrating a box plot of ROE(05) for each single value of CSR(05), the Fig. 4 is obtained.

[FIGURE 4 OMITTED]

It is possible to detect several anomalies, such as medians (horizontal lines within each rectangle) and averages (the values beside each box plot) lacking any clear tendency (linear or growing).

ROA(05) is substituted for ROE(05) and shows an almost horizontal line for medians and a non-defined behavior for averages (Fig. 5).

[FIGURE 5 OMITTED]

The calculations and subsequent results (Figs. 6 and 7) show there cannot be a relationship identified between the variables ROE(05)/ROA(05) and CSR05. The relationship between business profitability and CSR is specified.

[FIGURE 6 OMITTED]

[FIGURE 7 OMITTED]

An ANOVA test between ROE(05) and CSR(05) is conducted and the results are presented in Fig. 6.

The average values of ROE(05) are the same in each level of CSR(05), with a p-value of 0.917.

If ROA(05) is used instead of ROE(05), the results in Fig. 7 appear.

Therefore, a similar conclusion appears; the average values of ROA(05) are the same for each level of CSR(05), even with a superior p-value (0.997).

In order to undertake a more detailed analysis, the quartiles of ROE(05) are used (Fig. 8). Companies are classified into four levels depending on their ROE(05); the first (QROE05=1) includes companies with a ROE(05) lesser than the first quartile (Q1 = 0.0213); the second (QROE05 = 2) are companies with a ROE(05) between Q1 and Q2 (0.0213 and 0.1240, respectively); the third (QROE05 = 3) contains companies with a ROE(05) between Q2 and Q3 (0.1240 and 0.1748); and, finally, in the fourth quartile, there are companies with a ROE(05) higher than Q3=0.1748 (QROE05 = 4). By using these four levels, the variable QROE(05) is created.

Some categories of CSR(05) have few values; in order to evaluate the impact of these categories, the variable CSR(05) is grouped into 6 categories (0, 1, 2, 3, 4, 5). This classification was done by rounding up to the whole superior to get the variable QCSR(05). The relation between QCSR(05) and QROE(05) is analyzed.

Figure 9 shows the distribution for each QROE(05) category (the column), related to the different QCSR(05) categories (expressed as a percentage).

[FIGURE 9 OMITTED]

Twenty-five of the companies with the highest ROE(05) have relatively few cases of a QCSR(05) value equal to 0 (six cases, specifically) (Fig. 9). However, it cannot be stated that these companies are among the best because only seven have a maximum value of QCRS(05).

In order to finish this comparison, a regression analysis was undertaken to determine if CSR(05) depends on ROE(05) or ROA(05). Figure 10 shows that, despite very small [R.sup.2], the p-value (0.052) indicates that the model explains the CSR (05) variable; however, there are large variations within each category.

[FIGURE 10 OMITTED]

Moreover, it is necessary to highlight the fact that the signs of the ROE(05) and the ROA(05) variables within the regression model are different. Thus, an increase of ROE(05) would imply an increase of CSR(05), while the opposite would occur for ROA(05) [an increase of ROA(05) would imply a decrease of CSR(05)].

The next step is to analyze 2006, taking into account the known values of both ROE(06) and ROA(06). Will a good level of CSR(05) improve the financial results for companies in 2006? In order to answer this question, the comparison between CSR(05) and ROE(06) and ROA(06) is undertaken. Similar graphs to those shown for 2005 are given (Figs. 11 and 12). By examining them, one can determine that a higher level of CSR(05) does not imply better results for ROE(06) or ROA(06).

Print Get the Mag Weekly Updates