Ranking Indian Companies on Sustainability Disclosures Using the GRI-G4 Framework and MCDM Techniques

1.1 Sustainability accounting and reporting

Sustainability Accounting is a type of accounting that goes beyond mere financial or quantifiable disclosures, encompassing non-financial aspects as well. The extensive scope of this accounting concept benefits governments, suppliers, employees, and customers. Such reporting may be accomplished through Corporate Social Responsibility (CSR), sustainability reporting, and non-financial reporting. A broader disclosure that covers various dimensions aids in fostering a healthier global economy and ensures that businesses benefit society at large. Sustainability Accounting transcends traditional financial accounting, aiming for economic and social improvement rather than overemphasizing monetary aspects, thereby ensuring transparency for stakeholders. It has expanded the accounting field to cover non-financial dimensions and the organizational impact on society and the environment, modifying traditional cost and financial accounting practices by adjusting bookkeeping and disclosures to include social and environmental dimensions.

Sustainability Reporting is a report prepared by an organization to communicate its economic, environmental, and social impact to stakeholders such as investors, government, employees, etc., to help them understand its underlying motives towards sustainability. Over the last decade, companies worldwide have increasingly adopted sustainability reporting. It enhances both the internal performance and external reputation of organizations compared to those that do not prepare sustainability reports.

1.2 GRI-G4 framework-a brief introduction

"Established in 1997 as an independent body by the Coalition for Environmentally Responsible Economies and the United Nations Environment Programme, the GRI introduced the framework for environmental reporting." Following the establishment of a multi-stakeholder Steering Committee in 1998, the GRI framework was expanded to include economic, social, and governance aspects in its reporting. The first GRI guidelines were published in June 2000, followed by the second version in 2002. The third version, G3, which contained detailed instructions for sustainability reporting, was published in 2006. G3.1 was published in 2011, and G4 was published in 2013. According to KPMG survey reports (2008, 2013), the GRI framework is a widely accepted reporting framework for economic, social, and environmental performance. The general and specific disclosure aspects of the GRI framework for sustainability reporting are agreed upon by a majority of stakeholders worldwide.

The GRI-G4 framework has four categories: Governance, Economic, Environment, and Social, each with different aspects. Companies following GRI guidelines for disclosing their corporate sustainability report on these four categories. The governance category has seven aspects, the economic category has four, the environment category has twelve, and the social category has twenty-nine. The Social category includes four sub-categories: labor practices and decent work, human rights, society, and product responsibility. Each aspect includes several indicators, as outlined in the GRI-G4 framework. The GRI-G4 Framework contains a total of 52 aspects and 149 indicators across different aspects.

"Besides the GRI framework, other guidelines for reporting sustainability issues include the Dow Jones Sustainability Index (DJSI), the Sustainability Accounting Standards Board (SASB), the United Nations Global Compact (UNGC), and the International Integrated Reporting Councils (IIRC)." While the scope of sustainability reporting in other frameworks is subjective, the GRI-G4 framework is objective. This objective nature makes the GRI-G4 framework unique among other frameworks, which are largely subjective in nature.

3.1 Content analysis technique

Content analysis technique is “used in this study for extracting information in a numeric form from the published sustainability reports of the select companies”.

A binary coding system is "used to measure the level of corporate sustainability reporting (CSR), i.e., ‘1’ if the item is disclosed or ‘0’ otherwise”.

For calculating the “quality of sustainability disclosure, a four-point scale (ranging from ‘0’ to ‘3’) is used”. ‘1’ is assigned if the item is partly disclosed, ‘2’ if the item is fully disclosed in narrative form, ‘3’ if it is disclosed in quantitative form, and ‘0’ it not disclosed.

Post obtaining “the item-wise score, the overall disclosure score for sustainability performance is computed by using the following equation”:

$\mathrm{I}_{\mathrm{k}}=\sum_{\mathrm{i}=0}^{\mathrm{n}} \mathrm{Xik} / \mathrm{N}_{\mathrm{k}}$           (1)

where, N_k is the maximum expected score for each category, X_ik “assumes value ‘0’-‘3’ in a four-point scale”.

Research type is descriptive in nature.

3.2 Sampling procedure and sample size

The present study will be “based on the secondary data collected from sustainability reports published in the respective company’s website. The population of this study comprises all listed companies”. As the empirical evidence relating to the CSR in the context of India is very limited, the present study relies on disclosures uploaded in the website of GRI (2018) in selecting the sample of the study. The sample companies will be selected on the basis of the following criteria; the company must be a listed company on NSE or BSE; the company must be publishing its sustainability reports either on its own website or uploaded on GRI website; and the company must be publishing its respective sustainability report continuously at least for a period of six years as per the GRI guidelines. See Table 1, there are 116 Indian companies which are reporting on sustainability parameters as per GRI-G4 database (GRI 2018) from 2014 to 2019. From preliminary scrutiny it was found that about 53 companies which are regular in sustainability reporting on continuous basis for past 6 year’s but excludes all financial companies because financial companies following GRI-G4 report along with ‘GRI Financial Services Sector Supplement’. Therefore, the sample size is 53.

Sample Size: All the listed companies except financial sector which are disclosing sustainability report for six years (2014-2019) as per GRI G-4 guidelines have been considered for the study.

3.3 Research framework

Figure 1 presents the framework which is followed for the calculation of rank of the select companies under study. Here researcher first identify the different criteria under study. Second step is calculating the data from the sustainability reports published by the companies under study. Third step is normalizing the data and applying other different procedures which are mention in the research methodology section of Entropy to calculate the criteria weight. Fourth step is to apply TOPSIS (Technique for order preference by similarity to ideal solution) along with Entropy weight to calculate scores of sustainability disclosure. Finally, on the basis of scores obtain in the fourth step, ranking of companies are done.

1.jpg

Figure 1. Framework for ranking of companies

3.4 Entropy

The average amount of information, surprise, or uncertainty present in a random variable's potential outcomes is known as the entropy of the variable in information theory. Information entropy, sometimes known as Shannon entropy in his honor, was first developed by Claude Shannon in his 1948 publication A Mathematical Theory of Communication. Take a biased coin as an illustration, with a chance of p for heads and a probability of 1-p for tails. A coin flip has an entropy of one bit when p=1/2, when there is no reason to expect one result over another. In this instance, there is the greatest amount of surprise. The event is known and there is no entropy when p=0 or p=1, respectively. Entropy method can be explained in these four steps as follows consecutively:

Step 1: structure of the decision matrix. Supposing there are ‘n’ alternatives (Y={ Y_i |i= 1, 2,..., n}) and ‘m’ criteria(X={ X_j |j=1,2,... , m}) in any problem, decision matrix (DM) can be expressed as follows: 

X1    X2     X3   …   Xm

$\mathrm{DM}=\left[\begin{array}{ccccc}a_{11} & a_{12} & a_{13} & \cdots & a_{1 m} \\ a_{21} & a_{22} & a_{23} & \cdots & a_{2 m} \\ a_{31} & a_{32} & a_{33} & \cdots & a_{3 m} \\ \vdots & \vdots & \vdots & \ddots & \vdots \\ a_{n 1} & a_{n 2} & a_{n 3} & \cdots & a_{n m}\end{array}\right] \begin{aligned} & Y 1 \\ & Y 2 \\ & Y 3\\ &...\\ & Y n\end{aligned}=[Aij]  \mathrm{nxm}$

Note: ‘i’ represents row and ‘j’ represents column elements of a matrix and [Aij]nxm is a matrix of order n x m.

Step 2: normalization of decision matrix. The decision matrix needs to be normalized for each criterion X_j(j=1, 2, ... , m). The normalize decision matrix (NDM) will be NDM=[x_ij].

$x_{i j}=a_{i j} / \sum_{i=1}^n a_{i j}$           (2)

where, i=1, 2, 3…. n & j=1, 2, 3, …, m.

In this step, each data entry in a column is divided by sum of all the column data. Which make the data standardized i.e., in probability form.

Step 3: How to calculate entropies: The entropy B_j of the set of attributes j defined as given in Eq. (3):

$B_j=-\mathrm{K} \sum_{i=1}^n x_{i j} * \log \left(x_{i j}\right)$           (3)

where, i=1, 2.., n & j= 1, 2,…….., m and K is constant whose value is K = 1/log(m), m is number of criteria that gives value of B_j between 0 to1 (0≤B_j≤1).

Step 4: How to calculate Entropy weight: Degree of divergence (dd_j) is calculated for each criteria X_j(j=1,2,…m) by this formula (4),

$d d_j=1-B_j$           (4)

The higher the value of dd_j shows criteria X_j is more important for the problem, the criteria weight can be calculated by this formula (5),

$c w_j=\frac{d d_j}{\sum_{j=1}^m d d_j}$           (5)

where, j=1, 2, ……m.

Entropy (B_j), Degree of Diversion (dd_j=1-B_j) and Criteria Weight (CW_j) values are calculated by given equation above. Degree of divergence is calculated one minus Entropy (1-B_j). Since Entropy values show randomness in the data, so high value of Entropy means randomness is high in the data, so degree of divergence is low. Degree of divergence is actual information in the given data set by removing randomness in the data. Criteria Weight is calculated dividing each column of degree of divergence by sum of all degree of divergence i.e., weightage average.

3.5 TOPSIS

Technique for order preference by similarity to ideal solution (TOPSIS) method was developed by Hwang and Yoon [24], is most popular and commonly used as multi-criteria decision making (MCDM) methods and is work on the concept the selected alternative should have the minimum distance from the positive-ideal solution and the maximum from the negative-ideal solution. The positive ideal solution maximizes the benefit criteria and minimizes the cost criteria and negative-ideal solution maximizes the cost criteria and minimizes the benefit criteria. TOPSIS method assumed that each criteria hasn’t equal weight, so, it needs criteria weights from the decision maker [24]. At the end of the procedure, it produces performance score between 0 to1 on that basis alternatives are ranked. These are the steps of TOPSIS method as follows:

Step 1: construct the decision matrix: Supposing there are ‘n’ alternatives (Y={ Y_i |i = 1, 2, ..., n}) and ‘m’ criteria (X={ X_j |j = 1, 2,... , m}) in a problem, decision matrix (DM) is expressed given in Entropy section.

DM= [Aij] nxm

Note: ‘i’ represents row and ‘j’ represents column of a matrix and [Aij]nxm is a matrix of order n x m.

Step 2: normalization of decision matrix (NDM): The decision matrix needs to be normalized for each criterion X_j(j=1,2,... ,m). The normalize decision matrix will be NDM=[x_ij].

$x_{i j}=a_{i j} / \sqrt{ }\left(\sum_{i=1}^n a_{i j}^2\right)$           (6)

where, i=1, 2, 3…. n & j=1, 2, 3……, m.

In this step, each data entry in a column is divided by sum of all the column data. Which make the data standardized i.e. in probability form.

Step 3: calculate the weighted normalized decision matrix (WNDM): Elements in each column of matrix NDM=[x_ij] are multiplied with the relevant criteria weight (cw_j) (Eq. (5)) value and matrix WNDM=[b_ij] is created.

$\mathrm{WNDM}=\left[b_{i j}\right]=\left[c w_j\right] *\left[x_{i j}\right]$           (7)

where, i=1,2,…n & j=1, 2, …, m.

Step 4: calculate the distance from the best ideal solution (BIS) and the worst ideal solution (WIS): The distance of each alternative from positive ideal solution and negative ideal solution is calculated as given equations:

$\begin{gathered}\mathrm{BIS}=\left\{\left(\max _i\left[b_{i j}\right] \mid j \in J_1\right),\left(\min _i\left[b_{i j}\right] \mid j \in J_2\right) \mid \mathrm{i}=1,2,3 \ldots \mathrm{n} \& \mathrm{j}=1,2,3, \ldots, \mathrm{m}\right\}\end{gathered}$           (8)

$\begin{aligned} & =\left[b_1^{+}, b_2^{+}, b_3^{+} \ldots \ldots, b_j^{+}, \ldots ., b_n^{+}\right]=V_{+} \\ \text {WIS }= & \left\{\left(\min _i\left[b_{i j}\right] \mid j \epsilon J_1\right),\left(\max _i\left[b_{i j}\right] \mid j \epsilon J_2\right) \mid \mathrm{i}=1,\right. \\ & 2,3, \ldots \mathrm{n} \& \mathrm{j}=1,2,3, \ldots \mathrm{m}\} \\ = & {\left[b_1^{-}, b_2^{-}, b_3^{-}, \ldots \ldots \ldots b_j^{-}, \ldots \ldots . b_n^{-}\right]=V_{-} }\end{aligned}$           (9)

where, J₁ and are benefits and cost criteria, respectively.

V+ is best ideal solution which is calculated by finding the maximum value in the criteria column.

V- is worst ideal solution which is calculated by finding the minimum value in the criteria column.

Step 5: calculate the separation from best ideal solution (BIS) and worst ideal solution (WIS) between the alternatives: Calculation of the separation from best ideal solution (V+) and worst ideal solution” (V-) between the alternatives: with the help of distance formula given by Euclidean which is define as:

$S_i^{+}=\sqrt{ }\left(\sum_{j=1}^m\left(\left[b_{i j}\right]-V_{+}\right)^2\right)$           (10)

where, i=1, 2, 3, …, n.

$S_i^{-}=\sqrt{ }\left(\sum_{j=1}^m\left(\left[b_{i j}\right]-V_{-}\right)^2\right)$           (11)

where, i=1,2, 3, …, n.

Step 6: calculate the coefficient of separation:

$P_i=\frac{s_i^{-}}{s_i^{+}+S_i^{-}}, P_i \in[0,1] \forall i=1,2,3, \ldots \ldots, n$           (12)

On the basis of value of P_i preferred order can be obtain and chose the rank in descending order of values P_i.

The study revealed the evolution of better transparency and quality of sustainability disclosures by the Indian companies in accordance with the GRI standards. The research unveiled the factual observation that companies in India are becoming more serious about sustainability reporting to communicate to their stakeholders about their responsibility toward the economy, society and environment. The study also concludes the statistically significant impact of the transparency and quality of the sustainability disclosures of the companies over the progress of the company. The results are well supported by the fact that the improvement in the quality and transparency of the sustainability reporting not only makes a company compliant with the GRI standards but also creates its goodwill in the eyes of the stakeholders that company is serious about the benefit of the stakeholders and environment at large which in turn increases the investment as well as turnover of the company. Thus, though the sustainability disclosures are the part of non-financial aspects of a corporate but still play a whip hand in the image and progress of the company in the foreseeable future which is one of the most favourable factor for the corporates in the modern world in light of growing investor awareness round the world. Though the selected companies were found to be compliant with the disclosure standards of GRI sustainability reporting guideline, yet the analysis revealed a need to upgrade the disclosures of such corporate for achieving better transparency and quality of reporting. It has also been observed that quality of corporate disclosures in terms of governance is better in most of the companies but majority of them lack the transparency and quality of the aspects of economic disclosures in their sustainability reporting. The conclusion of the study is justified by the fact that in the recent past, various laws and Government of India as concentrated on ‘good corporate governance’ which eventually increased the quality of governance disclosures. Further, strict implementation of the environmental laws governing the liabilities and disclosures on ‘no fault basis’ lead to improvement in the aspects of environmental aspects of the sustainability disclosures of the companies in India. But comprehensive guidelines and regulations on economic disclosures is still awaited which is the major reason for Indian corporate lacking in the transparency and quality of economic disclosures in their sustainability reporting.