A Study of the Economic and Non-Financial Performance Indicators in Corporate Sustainability Reports

The purpose of this paper is to identify the performance indicators disclosed in corporate sustainability reports. To perform this study we examined Italian Listed companies that produced a sustainability report in 2012. The indicators were identified using a content analysis. We analysed the core and additional indicators disclosed in sustainability reports as well as all the indicators required by sector supplements adopted by companies. Our results show that indicators are widely disclosed in Italian sustainability reports. Social indicators are on average the most commonly used indicators, particularly those concerning labour practices, followed by the economic and then the environmental indicators. The Oil and Gas and Utilities industry sectors disclosed a superior amount of indicators compared to all other sectors. These industry sectors also show a more homogeneous behaviour, also as regards disclosure of core and additional indicators. This study provides one of the first detailed analyses of the different category of GRI indicators used by Italian companies producing sustainability reports.

One of the most distinguishing features of the GRI Guidelines, even in its earliest version (G1, 2000), is the inclusion of a set of indicators designed to represent a company's economic, social, and environmental performance.Many things have changed between the first version of the GRI guidelines and the fourth version, such as the number of environmental, social and economic indicators, as well as the conceptualization of these indicators and the consideration of integrated indicators.Initially the indicators were defined according to aspects considered material (core) by the company and its stakeholders, or representing an emerging practice or one of particular interest to certain organizations (additional).
Since G2, the presentation of the indicators in sustainability reports has been supported by the existence of Indicator Protocols.The 2002 GRI family of documents began gradually including Sector Supplements (SS) proposed generally in the form of performance indicators and their associated indicator protocols.The SS provide interpretations and guidance on how to apply the GRI Guidelines in a given sector and they include sector-specific Performance Indicators.With the adoption of G3 (GRI, 2006) and again with G3.1 (GRI, 2001), the number and type of indicators to be presented in the sustainability report varied depending on the GRI Application Levels.The maximum application level "A +", provides the inclusion of the core indicators of G3 and G3.1 in the report, as well as each performance indicator of the Sector Supplement in accordance with the materiality principle and requires an explanation if these are omitted.For each of the application levels, a "+" To achieve our objective, the paper has been structured as follows: the next section contains a brief review of the literature on GRI indicators in sustainability reports and is followed by an empirical analysis that introduces the research sample and methodology; lastly, we will share our results and offer our conclusions.

Performance Indicators and GRI-based Sustainability Reports
As stated earlier, the indicators are useful tools for analysis and control of business performance and can express the complex and dynamic events that characterize enterprise management in a synthetic and integrated way (Schaltegger & Burritt, 2000;Adams & Frost, 2008).These measurements are useful in decision-making processes and in the management of a company's socio-environmental issues and they make it possible to: • identify and achieve social, environmental and economic objectives by controlling the degree of achievement and value over time; • identify areas of inefficiecy; • pursue benchmarking purposes (Jasch, 2009;Schaltegger & Burrit, 2000;Mio, 2001;2005;Perrini & Tencati, 2006).Furthermore, indicators are useful tools in external communication; through indicators stakeholders can evaluate a company's commitment and results, as well as make inter-temporal and intrasectoral comparisons.Wilburn and Wilburn (2013, p. 73) state that performance indicators offer stakeholders «[…] a means to evaluate the ethical basis of a company's CSR/sustainability programs».
Despite their informational value, however, few studies have specifically analysed the performance indicators used by companies in their CSR reports.
As previously mentioned, one of the most distinguishing features of the GRI Guidelines, even in its earliest version (G1, 2000), is the inclusion of a set of indicators designed to represent a company's economic, social, and environmental performance.After the year 2000, the GRI published three other versions of the GRI guidelines.The G2 were published in 2002 and replaced by the G3 in 2006, which were then partially revised by G3.1 in 2011.The G4, approved in 2013, is the most recent version of the GRI guidelines.
All these versions of the guidelines classify sustainability indicators from the perspective of the Triple Bottom Line (Elkington, 1997), distinguishing between economic, environmental and social indicators.GRI indicators are classified into "aspects", while the social indicators alone (with partial differences for G1) are divided into "categories" relating to labour practices, human rights, society and product responsibility.
The various revisions of the GRI guidelines have always served to enhance the indicators, thus confirming their importance.
In the transition from the first version of the guidelines to the G2, the most significant changes affected the economic and social indicators.Moreover, the distinction between "core" and "additional" indicators was introduced with G2.Specifically, the "core" indicators are considered important for both users and reporters and for most organizations, while the "additional" indicators « […] are those that have one or more of the following characteristics: represent a leading practice in economic, environmental, or social measurement, though currently used by few reporting organisations; provide information of interest to stakeholders who are particularly important to the reporting entity; and are deemed worthy of further testing for possible consideration as future core indicators» (G2, 2002, p.13).
The G3 considerably altered the division of sustainability indicators into categories and aspects by proposing a more structured and synthetic framework of indicators.These guidelines were then modified in G3.1, with significant changes to the social indicators.Specifically, more relevance was given to the impact of the company on the local community, human rights and gender.
The G4, the fourth generation of GRI guidelines, also changed the indicators, especially the environmental and social ones (in particular for the categories relating to labour practices, human rights and society), and integrated information on the supply chain.
Furthermore, G4 deleted the distinction between core and additional indicators.All indicators, in each aspect, are considered «[…] as equal inputs in the process for defining report content» (G4, 2014, p. 20 ).
In addition to the GRI indicators, the different versions of the guidelines also provide that companies can use other sector-specific performance indicators, included in the Sector Supplements, (G2; 3; 3.1; 4) (Note 1) or, as specified in the G4, indicators relating to "material topics" and not included in the list of GRI indicators (G4, 2014;GRI, 2014).
Based on these observations, the next section provides a brief review of the literature of responsibility accounting and reporting and the contribution of these studies regarding the different use of GRI indicators.

Background and Prior Studies
We have already seen that the current literature on responsibility accounting and reporting has provided a limited or marginal analysis of one of the most significant components of the sustainability report, i.e. the indicators of economic, environmental and social performance.
Many studies analysed only one particular type of indicator: environmental, social or economic (Azzone & Dubini, 1993;Ilinitch et al., 2008;Jasch, 2000;Olsthoorn et al., 2001;Gallego-Àlvarez, 2012).Olsthoorn et al. (2001, p. 456) point out that the study of environmental indicators especially, although relatively "new", «[…] is already highly diversified with approaches based on LCA, economics, management accounting, ecology and a physical gate-to-gate analysis».Other studies have extended the analysis to the three types of GRI indicators, but analyse them within the scope of broader research covering the full content of responsibility reports, as well as the "quality" and level of compliance to GRI Guidelines and/or environmental standards (Morhardt et al., 2002;Guthrie & Farneti, 2008;Skouloudis & Evangelinos, 2009;Skouloudis et al., 2010;Asif et al., 2013;Romolini et al., 2014).
Yet few studies have specifically examined the use of GRI indicators in companies' sustainability reports (Gallego, 2006;Roca & Searcy, 2012;Samuel et al., 2013;Alazzani & Wan-Hussin, 2013).Gallego (2006) analyses the disclosure of GRI indicators (2002 version) in the sustainability reports of 19 Spanish companies.The results reveal attention to social issues, primarily in terms of indicators related to labour, practices and decent work, strategy and management, non-discrimination, freedom of association, child labour and forced and compulsory labour as well as environmental issues, primarily in terms of indicators related to energy, water, biodiversity and emissions, effluents and waste.Furthermore, the analysis shows that the sector of activity influences the type and the number of indicators disclosed.Roca and Searcy (2012) provide a review of the indicators used in corporate sustainability reports.A content analysis of the reports published by 94 Canadian companies, divided into 10 sectors, revealed the use of 585 different indicators, as well as the use of the indicators provided by the GRI G3 guidelines by 31 companies.While companies operating in the banking, engineering, construction and chemical production, mining, transportation, communications and services, and oil and gas sectors reported numerous indicators, most companies from the electricity, retail, and food industries, on the other hand, did not use the GRI indicators.Furthermore, among the 79 indicators listed in the GRI G3, all were used at least once.While the GRI's economic indicators were widely reported in many sectors, there were many differences in the environmental indicators as well as in the various categories of social indicators that were reported.Samuel et al. (2013), instead, provide an insight into sustainable production indicators used only by the petrochemical industry in Malaysia.Results indicate that the majority of the indicators disclosed are related to compliance, performance and environmental impacts.The indicators related to supply chain and product life cycle were weakly reported, and none of the indicators disclosed fell within the category that addresses sustainability issues in terms of the ecological carrying capacity.
In the same sectoral perspective, Alazzani and Wan-Hussin (2013) analyse how the use of a voluntary standard assessment system for environmental reporting could help mitigate the damage caused by oil and gas companies to developing nations.The study evaluates the environmental practices of eight oil and gas companies, and content analysis of their environmental reports 2009 indicates that they made reasonable efforts to disclose their environmental performance in accordance with the GRI G3 Guidelines.
Overall, then, there are very few studies that specifically investigate the disclosure of GRI indicators in sustainability reports of companies.This lack drove us to further explore this subject.

Sample
The study sample is made up of all 292 companies, both Italian and foreign, listed on the Italian Stock Exchange on March 31, 2014.The companies are analysed according to their relevant "industry" membership as identified on the Italian Stock Exchange website.
In order to verify the number and type of economic, environmental, and social indicators included in the GRI-based sustainability reports compiled in 2012 by the companies listed on the Italian Stock Exchange, it was first necessary to verify how many of the listed companies produced such reports.Table 1 provides the final sample analysed.

Initial sample 292
• Companies suspended from listing on the date of the analysis 7 • Companies that produce voluntary reports but not GRI-based 17 • Companies that produce no voluntary reports 221

Final sample analysed 47
In Table 2, the 47 firms in the sample analysis are divided according to industry sector, the GRI guidelines adopted and GRI Application Level (AL).
Table 2 shows that the Financial Services sector contains the largest number of companies, followed closely by the Industrial sector.Meanwhile, no enterprises operating within the Basic Materials and the Health Care sectors compiled GRI-based sustainability reports.Furthermore, the Table shows which GRI Guidelines the companies adopted, revealing a widespread use of the G3.1 (30 companies out of 47), primarily in the Utilities sector where they were used by all eight companies in question, and in the Consumer Goods and Industrial sectors where they were used by most of the companies.With regard to the AL, there is a prevalence of the A+ level, both for companies that adopt the G3.1 (18 companies out of 30), as well as for those that adopt the G3 (9 companies out of 17).The B + level is the second most popular, again for both of the guidelines, although the number of companies is much lower compared to the former level.

Methodology
Our analysis focused on the sustainability reports of the sample companies, all dating from 2012 and present on their corporate websites.To answer our research questions, we studied the GRI content index present in the corporate sustainability reports of the sample (43 cases) and, when absent, referred to the GRI index available on each corporate website (4 cases).The verification of economic, environmental, and social indicators in the sustainability reports was performed by using content analysis (Krippendorff, 2004).The presence in the sustainability reports of each indicator included in the GRI index was verified in relation to the page(s) indicated within the GRI index.Thus, an excel spreadsheet was created which catalogued the indicators included within the GRI G3 (79) and G3.1 (84), (Note 2) codifying an indicator's presence in the report with the value "1" and its absence with the value "0".Each indicator was given the same "weight" (Marston & Shrives, 1991).To improve the reliability of data analysed, the verification of the GRI index and of the content of sustainability reports in the section relating to indicators, was carried out by two separate researchers.The results were then compared and the final data codification table was compiled.

The Categories of Performance Indicators in Sustainability Reports
This section presents an overview of the different categories of economic, environmental and social indicators, distinguishing between core and additional indicators.See Table 3.As Table 3 illustrates, all of the indicator categories are, on average, well represented in the reports of the selected companies (with percentages exceeding 60%).This trend is mainly generated by the core indicators of all categories, with percentages of reporting close to or above 70%, while the additional indicators generally had much lower percentages, starting from values of 35%.
The Social Indicators related to LA, with an average of 12 (or 80%), represent the most frequently mentioned category in the sustainability reports (with values ranging from a minimum of 3 to a maximum of 15).The next most common category is that of EC indicators (76%) which feature a higher concentration around the average value (standard deviation of 1.99) than the others.The EN indicators, on the other hand, have an average frequency of approximately 67% and the highest standard deviation (6.4).
These results are replicated, and even improved, in the disclosure of the core indicators.
The core indicators LA and EC show the highest percentages of disclosure (with values of 87% and 84%, respectively) compared to all other categories of core indicators.The core indicators EN and PR follow, both with values of 75%.As regards the additional indicators, however, only the LA maintain the top position (with 72%), followed by additional indicators SO (with 65%).
On the other hand, the other categories of Social Indicators (core and additional), related to HR and PR, are the ones reported on average less frequently (with percentages of 61% and 62%, respectively), highlighting cases of companies that report none of these performance measures (minimum values of 0).
Overall, the average and the median values in all of these indicator categories are remarkably close to one another, thus revealing a considerable symmetry in the distribution of these indicators in each category.Incidentally, half of the categories have median values higher than their average values, indicating that more than half of the companies studied report a higher number of indicators from each category than the average.The opposite is true only for the HR, SO, and PR indicators, in which the average values are higher than the median.This trend is also confirmed, except for indicators SO, in the analysis of additional indicators.The median values are higher than the average for less than 5 of the 7 categories of indicators.In the core indicators, however, all categories of indicators have a numerosity above average, as is evidenced by the median values being always higher than the average values.
Of particular interest are the values for the Total Indicators.The total number of indicators reported by the companies ranges from a minimum of 17 to a maximum of 83, with a markedly larger difference between the average values (58) and the median values (61) as compared to the other categories of indicators.Furthermore, the standard deviation of 17.26 indicates a significant contrast between companies in their disclosure of Total Indicators.Similar considerations may also be made for core and additional indicators within the category of Total Indicators, specifying that such differences between mean and median values, compared to other categories of indicators, rise significantly for the core indicators and are reduced for additional indicators.This confirms the greater importance that all companies give to the disclosure of core indicators in all the categories analysed. www.ccsen

The Use of the Different Categories of Performance Indicators
In the following Tables 5, 6 and 7 we present a more detailed descriptive analysis by illustrating the employment of all the types of GRI indicators by industry sector, and total of companies in the sample.Core and additional indicators of each category of GRI indicators are marked in the tables in gray and black respectively.Table 5 shows the frequency of use of each EC indicator by sector and total of companies.prevalence of core indicators, are widely used by Italian companies that produce sustainability reports.This result is certainly the consequence of the high application level of GRI guidelines (A+), reported by more than half of companies in the sample analysed (about 57%).The links previously mentioned between the application level and the number of indicators to be reported led us to expect the presence, at least, of core indicators (G3 and G3.1) and of SS indicators (where adopted).
From the point of view of the total numerosity, the reporting of the total of standard indicators reaches values that are indisputably significant (about 70% of those provided), with disclosure percentages of core indicators equal to 77% of those accountable.The number of core indicators disclosed increases further if you also consider the indicators reported by companies employing the SS.Among the companies in the sample, those that made the most extensive use of SS are those of the Utilities sector, and over half of the Financial Services sector.The possible use of an SS allows any company that adopts it, to report corporate performance more comprehensively, using specific performance indicators or adding comments to standard indicators.
On average, the category of indicators most frequently reported in sustainability reports of companies of different sectors is that related to Labour Practices and Decent Works (LA).These indicators of social dimensions are followed in number by those of the economic dimension and finally by those of the environmental dimension.
As regards the different sectors, the Oil and Gas and the Utilities sectors showed a higher quality and quantity of indicators than the others.These sectors show a more homogeneous distribution of companies reporting a higher number of indicators from various categories, as opposed to other sectors' reports which are characterized by greater discrepancies in the number and types of indicators reported.The Oil and Gas and the Utilities sectors typically have a significant social and environmental impact (Clarkson et al., 2008;Pattern, 2002;Cho & Patten, 2007) and are likely to experience more pressure from stakeholders, consumers, and government bodies (both local and not) (Cowen, Ferreri, & Parker, 1987).These considerations may therefore explain the attention given to a fuller disclosure of the indicators, both core and additional.Considering both the overall analysis of indicators reported, and the frequency of use of different types of standard indicators, we observe that the most reported indicators in the various sectors are: EC1 (core), which provides information on the economic value generated and distributed, EN3 (Direct energy consumption by primary source Energy) and EN8 (Total water withdrawal by source), again both core.As regards the social dimension, LA1 (pertaining to the main characteristics of total workforce), LA7 (pertaining to the rates of injury, occupational diseases, lost days etc.) and LA10 (pertaining to the average hours of training).These results are consistent with those found in similar studies applied in other areas (Gallego, 2006;Roca & Searcy, 2012).The absolute importance of the EC1, EN3 and LA1 indicators also emerged from the sample of Spanish companies (Gallego, 2006) and Canadian companies (Roca & Searcy, 2012).
The same seems to apply to the data used in the construction of many indicators of economic performance, largely recovered from the management and financial accounting system and/or by other company structures (finance, accounting, personnel management, etc.).The construction of environmental indicators can be rather more problematic especially if the company does not have an adequate system for measuring and monitoring the environmental impacts of business activities.In fact, while some data can be obtained from invoices or delivery notes (e.g.data for the construction of EN3) or from water metres or bills (e.g. for the construction of EN8), for other indicators data may have to be obtained from ecological accounting systems or life cycle assessment, etc. (e.g.EN7, EN9, EN12 etc.) which are not always available in companies.The analysis of indicators for each sector shows differences in the disclosure of the various indicators of companies.This result is also confirmed by other studies that highlight that the sector to which a company belongs influences the level of their social, environmental and sustainability disclosure (Roberts, 1992;Tilt, 1994;da Silva Monteiro & Aibar-Guzmán, 2010;Legendre & Coderre, 2013;Kansal, Joshi, & Batra, 2014).The significant disclosure of GRI indicators highlighted by our analysis of sustainability reports can thus be considered positively.As noted, the ruling by GRI of a set of indicators related to a three-year period and defined, with reference to the core indicators, through GRI multistakeholder processes, can be a useful support tool to decision-making processes and to accountability reports.However, if you observe the construction model of indicators proposed by the GRI, you can understand that the cognitive-informative potential of indicators is limited by the absence of integrated, cross-cutting and systemic indicators, both in the third and in the fourth generation of the GRI guidelines.Sustainability reports are unable to propose a structured system of indicators, but as Moneva et al. (2006, p. 134) observe, a «[…] collection of (non-integrated) indicators» unable to support «an integrated and systemic view of business and the environment».Buhr, Gray and Milne (2014, p. 63) state that «[…] the GRI has not managed to gain agreement on a full set of indicators which together might constitute something approaching a social and/or environmental accountability».Improving the system of indicators is certainly possible (for example by building an advanced business information system and by extending the stakeholder engagement process) and is auspicious.It is believed that the establishment of a system of indicators that are not only representative of the three dimensions of sustainability but also properly integrated and comparable over time and space and then based on uniform criteria and objective, i.e. independent from the observer, as well as demonstrable and built on topics considered material and understandable by stakeholders, can provide real added value to the assessment of socio-environmental activities of companies.
We have already seen that the disclosure of economic, environmental, and social indicators in a company's sustainability report has been the object of interest of a small number of scientific papers.Our paper, therefore, aims to contribute to the awareness of the use of indicators in sustainability reports produced by Italian listed companies.The results only refer to the sample of listed companies that produce a GRI-based sustainability report and cannot be extended to the wide spectrum of non-listed Italian companies that frequently produce CSR reports.There are several possible lines of development of this research.We might think of enriching the research by analysing not only the indicators in sustainability reports but also the indicators that Italian companies must disclose, in compliance with the provisions of art.2428 of the Civil Code, in their management reports.Moreover, the recent development of Integrated Reporting highlights the need to overcome the separation between information reported in various statements and in their supplements, in order to achieve an integrated reporting system.In this perspective, the presentation of a system of performance indicators, which summarize the total performance of a company, could offer a concise overall look at the company's system that would be useful in collecting data and information considered material by various stakeholders.The extension of the analysis, within a comparative framework, to other countries would be no less interesting.

Table 2 .
Number of companies by industry sector, GRI guidelines adopted and AL

Table 3 .
Descriptive statistics: number and category of GRI indicators disclosed in sustainability reports Consumer Services only for the PR indicators.The worst performance is by the Consumer Services sector as regards Social indicators related to Human Rights (HR), with 36% of these reported.A comparison between the various aspects of social indicators among the different sectors, confirms the previous statement regarding the increased use, in all sectors, of the LA indicators compared to the others in the social category.The LA indicators are exceeded only by the SO indicators (91%), in the Utilities sector, and by the HR indicators (55%) in the Telecommunications sector.This trend in the sectorial disclosure of the four aspects of social indicators also exists for this category's core indicator.In fact, in all sectors, the reporting of the LA core indicators dominates compared to the other social indicators, with maximum levels (95%) in the Utilities and Oil and Gas sectors.The additional indicators for the four aspects of social indicators have relatively lower percentages of disclosure than the core indicators of this category, as well as also those shown for EC and EN additional indicators.The lack of reporting of HR additional indicators in the Consumer Services, Telecommunications and Technology sectors is also noticeable.These last two sectors do not even show the SO additional indicators.In all sectors, the percentage reporting of LA additional indicators prevail over the remaining aspects of social indicators.The Oil and Gas sector has the highest in all four aspects of additional indicators of the social category.Moreover, the already high disclosure of LA additional indicators (equal to 90%) is exceeded by the 100% use of SO additional indicators only in this sector.

Table 5 .
EC indicators by industry sector and total of companies Comparatively, the behaviour of the companies in the Utilities sector is excellent, showing full coverage (100%) of the EC core indicators, although they do not report the EC5 and EC9 additional indicators.The disclosure in the Oil and Gas sector is also significant, with the four companies producing a similarly full coverage in four EC core indicators (EC1, EC2, EC6, EC8).On the other hand, the EC5 additional indicator (on the range of ratios of standard entry level wage by gender