Quality and Di ﬀ usion of Social and Sustainability Reporting in Italian Public Utility Companies

: In recent decades, non-ﬁnancial reporting has been widely debated in the literature relating to both public and private sectors. Non-ﬁnancial reporting is used to increase accountability and transparency, and to adapt to external pressures and stakeholder expectations. The focus on external factors, i.e., transparency and stakeholders, has largely precluded research into the quality of non-ﬁnancial reporting. Nevertheless, the quality and reliability of sustainability reports have been widely questioned in the literature. Non-ﬁnancial reporting may provide purely symbolic actions to manage expectations. This paper analyzes the level of di ﬀ usion and quality of non-ﬁnancial reporting tools in the public utility sector. We use the principles of the GRI (Global Reporting Initiative) framework to measure quality, i.e., clarity and accuracy, timeliness and stakeholder engagement, comparability, and reliability. We use a qualitative exploratory approach with a mix of primary and secondary sources. The results show that despite the increasing use of non-ﬁnancial reporting in organizational life, it is not di ﬀ used within public utilities. We address the issue of quality, and ﬁnd that, overall, the accuracy / clarity and comparability of non-ﬁnancial reporting is satisfactory; timeliness and stakeholder engagement appear to be acceptable, while reliability does not appear to be acceptable.


Introduction
Sustainability is a popular topic within the academic literature. The concept of sustainability is linked to the concept of sustainable development, which is defined as "development that meets the needs of the present without compromising the ability of future generations to meet their own need" [1]. Sustainable development can be summarized as the inclusion of social and environmental aspects in business operations with the aim of satisfying the various stakeholders [2]. From a business reporting point of view, this idea has led to the well-known approach of the "Triple Bottom Line" (TBL), which considers non-financial aspects of a firm's performance [3].
Non-financial reporting is mostly a voluntary activity that has gained great adherence in the corporate world. This is the result of the diffusion of the concept of sustainability within organizations [4]. In this paper, non-financial reporting refers to sustainability, social, and integrated reports. The concepts of corporate social responsibility (CSR) and corporate sustainability have developed distinctly. Nowadays, the terms are used interchangeably [5] although they are conceptually different [2]. Adoption of non-financial reporting could have different internal and external benefits for an organization. It is an efficient tool to let external and internal stakeholders look "under the hood" of an organization. Non-financial reporting helps pursue engagement and transparency about how a

Theoretical Background
The quality and reliability of non-financial reporting have been questioned in the literature [14,16,17]. Non-financial reporting practices might not provide a higher quality of information, but they could represent symbolic actions intended to present companies as being really engaged in social, environmental and sustainability issues [15].
Few studies specifically examine the quality of reporting, which is a central issue in giving a true view on the real commitment of the company in sustainability issues [11]. Stacchezzini et al. [18] highlight that firms give limited information about their actions to achieve a sustainability performance, and tend not to disclose information when their social and environmental commitments are scarce. Michelon et al. [15] report that, in non-financial reporting, there is an increasing lack of completeness and a decreasing amount of credibility in the information reported, as well as a general concern about overall reporting practices. Companies that use these practices do not provide a higher quality of information, and the use of these practices remains symbolic. In fact, the use of a non-financial reporting template (such as GRI) could increase the symbolic performance of companies by being able to "mark various GRI boxes" and could decrease the substantive performance [15]. In this sense, the increase in information disclosed is not necessarily indicative of management purposes regarding social and environmental issues. This may cast doubt on the quality of non-financial reporting, considering that, as highlighted before, external forces may be driving the decision to adopt non-financial reporting.
The lack of credibility and quality may also depend on the lack of regulation. In fact, Joseph [22] argues that companies should be supported by some form of regulation, such as the inclusion of professional agencies in the process of elaboration and communication of non-financial reporting. Effective external assurance processes provide companies with proactive advice that helps to safeguard core values within a corporation and to ensure it meets the expectations of management, compliance, and stakeholders alike [23]. Accordingly, assurance processes relating to non-financial reporting information positively affect professional investors' evaluation of a firm's sustainability performance [24]. Assurance processes can even act as a quality reinforcement for sustainability performance by positively influencing a company's reputation in relation to sustainability [24,25]. Since the process of assurance could have a positive effect on the perceived reliability of users, the use of the process could reduce information asymmetries and agency costs and increase credibility [26]. For instance, the GRI [27] encourages organizations to use the independent assurance of non-financial reporting to improve the credibility and quality of these reports. However, the assurance process itself could be the cause of the lack of credibility in non-financial reporting. Assurance standards are vague and regulations are non-existent; consequently, content is heterogeneous and different assurance providers use different assurance processes [28][29][30]. Accordingly, assurance processes could contribute to an increase in both the expectation gap between various stakeholders and a lack of credibility [31]. Furthermore, companies using these practices may not provide a higher quality of information. This is because non-financial reporting, in some cases, tends to highlight the positive aspects of sustainability performance and to obfuscate negative outcomes [32] in order to improve reputation [33,34]. In contrast, the GRI guidelines highlight that the quality of non-financial reporting depends on a balanced reflection of both positive and negative aspects of a company's performance.
Given these considerations, non-financial reporting has been criticized for its lack of relevance and credibility [15,18,35], and some studies have recalled the need for more research about the quality of non-financial reporting [14,30]. The most common theories used to explain non-financial reporting are legitimacy theory and stakeholder theory. Legitimacy theory assumes that the actions are acceptable if they respect some socially constructed system of norms, values, beliefs, and definitions [36]. In consequence, non-financial reporting establishes conformity between the social values of the company and acceptable behavior in the social systems of the company [37]. Legitimacy theory assumes that business actions are subject to social acceptance by wider society. A company loses its legitimacy if it behaves differently or if its operation is perceived to be in contrast with the views of society [38]. Companies may adopt non-financial reporting in order to construct a legitimate image, to scale back the amount of questions asked, and to maintain a certain level of confidentiality [17]. Having legitimacy is like having a license to operate [39,40]. This idea of legitimacy is also reflected in the principal reasons for the increased publication of non-financial reporting [41]. Another important theory relating to the emergence of non-financial reporting is stakeholder theory. This assumes that companies have to be managed in the interest of all their stakeholders, not only in the interest of shareholders [42]. As a consequence, companies have to take into account different perspectives and expectations of every stakeholder who has an interest in corporate activities [42,43]. In this sense, non-financial reporting has been increasingly adopted by companies worldwide, given the demand of stakeholders for a greater transparency on both environmental and social issues [13]. Non-financial reporting can be used to align stakeholder expectations, because it goes beyond financial aspects to consider the environmental and social aspects of the firm's performance. It helps manage the relationship between a company and its stakeholders, who often have divergent and conflicting expectations [44]. The main objective of non-financial reporting is to increase dialogue with stakeholders and to guarantee their participation in decision-making processes. Non-financial reporting does not add value if it fails to involve stakeholders or to create an ongoing dialogue with them and to influence the decisions and behavior of both the organization and stakeholders [27].
In this paper, we contribute to the debate by analyzing the quality and diffusion of non-financial reporting in public utility companies. We combine legitimacy and stakeholder theory to formulate our research questions and to explain our results. As highlighted previously, public utility companies are managed and controlled directly or indirectly by public agencies. They are more sensible about non-financial information disclosure and accountability. Non-financial reporting could be used to satisfy various stakeholders' expectations. On the other hand, non-financial reporting may be used in public utility companies to legitimize decision making and to establish conformity between social value contexts and acceptable behavior. In this sense, non-financial reporting could be a response to external pressures to legitimize public utility companies' behaviors. Accordingly, non-financial reporting may be diffused significantly in these companies to legitimize behavior given the demand of stakeholders. The quality, credibility and reliability of non-financial reporting may be increasing with the involvement of stakeholders, as companies have to meet various stakeholders' expectations. Otherwise, non-financial reporting may be adopted symbolically, and public utilities may give limited and unsatisfactory information about their actions to produce a sustainability performance. To measure the quality of non-financial reporting, we developed a model based on the GRI framework; the following section describes our conceptual framework.

Conceptual Framework: Quality of Non-Financial Reporting
The GRI framework is currently the most widely used reporting standard in sustainability reporting [45], and is considered the most detailed and comprehensive [45]. Previous studies show that the GRI standards are the most commonly reported source for social and environmental disclosures [46]. Therefore, it is appropriate to use the GRI framework to measure the quality of non-financial reporting content [47][48][49]. A recent study of sustainability reporting focused on GRI indicators [47,50]. In this paper, we use the GRI framework to individuate what kind of information must be contained within non-financial reporting. The GRI guidelines highlighted that reports should include various criteria, for example: vision and strategy; company profile; governance structure and management systems; GRI content index; performance criteria. These criteria could be divided into two typologies: core and additional. Core criteria are generally applicable for most organizations, whereas additional criteria consider new practices applicable to some organizations but not to the majority [27].
Regarding the content of the report, the GRI individuates four principles: materiality, stakeholder inclusiveness, sustainability context, and completeness [27]. According to the GRI (2013), materiality reflects the fact that information included in the report should describe the economic, environmental, and social impacts of the organization, and be decisive in the assessments and decisions of stakeholders. Stakeholder inclusiveness signifies that the report should consider and respond to expectations and interests of stakeholders. The sustainability context is the organization's performance in the wider realm of sustainability. Completeness signifies that topics analyzed and the report's boundaries should be enough to reflect the significant involvement of the company in social and sustainable issues and to enable stakeholders to evaluate the achievement of objectives defined.
Furthermore, the GRI, G3 and G4 versions identify the following principles for defining the quality of non-financial reporting: clarity, accuracy, timeliness, comparability, and reliability. Clarity indicates that information should be clear, understandable and accessible to stakeholders. Accuracy points out that information should be sufficiently accurate and detailed for stakeholders in order to evaluate company performance. Timeliness signifies that reporting should be processed regularly and available in time for stakeholders to explain and share the decisions made. Therefore, in our analysis we refer to "timeliness and stakeholders engagement" in order to give more emphasis to the role of stakeholders. Comparability indicates that the information presented should allow stakeholders to see an overview of performance with respect to goals and results achieved in previous years over time, which supports comparisons with other organizations. Reliability relates to the ways in which information and processes used to prepare reports are gathered, recorded, compiled, analyzed, and disclosed so that they can be examined and so that the quality and materiality of the information can be established. The following table summarizes how the various measures were operationalized. Considering that "clarity" and "accuracy" refer to the type of information that should be made available for stakeholders, we combine the two dimensions in one measure. For "clarity/accuracy" and "timeliness", we developed a five-level Likert scale [51]. We take "timeliness" to relate not only to when the information was communicated, but also the decision making and sharing process for non-financial issues. This is because, as reported by the GRI, the timing of information disclosure has enable stakeholders to be an integral part in the decision-making process. For this reason, and considering the importance of stakeholder inclusiveness in the process of reporting [27], "timeliness" considers whether stakeholders were involved in the process, which also means having information in time to evaluate performance. For "comparability", we use a three-level scale that better corresponds to the nature of the variable. We take reliability to mean the adoption of assurance processes. This is because, despite the limits, the processes of assurance could reduce information asymmetries and agency costs and increase the credibility and perceived reliability [26,27].
In this paper, we use the six principles of quality identified by the GRI to measure the quality of non-financial reporting in Italian local utility companies. We exclude the first principle of quality from the analysis, as it is not possible to give an effective judgment from reading the reports. We also analyze the level of diffusion of non-financial reporting in this kind of company, which is a secondary aim of our research. In this sense, we try to analyze the importance of the semantic properties of the information; in other words, we attempt to assess not only how much, but also what kind of information is disclosed and how [52].

Methodology and Variable Operationalization
The purpose of this research is to analyze the diffusion, which refers to how companies disclose and communicate non-financial information, and the quality of non-financial reporting in local utility companies. We use a qualitative exploratory approach to analyze the content of non-financial reporting in a sample of Italian companies. Analysis was done using manual content analysis [53,54]. This method is defined as systematic and replicable technique that allows us to analyze large volumes of data easily in a systematic way by identifying, through a coding process, specific characteristics of the text and compress it into various content categories [55]. The number of companies operating Sustainability 2020, 12, 4525 6 of 17 within this sector is constantly growing over time due to the implementation of some recent sector reforms [56]. In this research, we consider, as a reference population, all the companies associated with Federutility (the federation that brings together companies operating in the public services of water, the environment, electricity and gas in the Italian context). These companies are of great interest in this context, because their economic, environmental sustainability and social aspects are closely related to each other. Furthermore, in many cases, these companies include both private and public interests and hence have a large number of stakeholders [57]. For these companies, non-financial reporting could be an efficient tool for sharing results and contributing to greater transparency and accountability [58], but first we need to determine the level of quality of non-financial reporting. The number of associated companies was 393 in July 2012 when we started to collect data. Companies operate in the sector of electric distribution, gas, water supply and cemetery services. They include publicly owned and private companies, multinational (such as Veolia or Acciona) and local cooperative companies. The analysis of the data of diffusion of non-financial reporting started with the inclusion of the whole population represented by the 393 companies registered with the Federutility association, removing from the total those companies that were owned or controlled by other affiliates. This was done because, normally, non-financial reporting assumes a business group dimension, and the calculation of the percentages of diffusion would be vitiated by a separate, double consideration of these realities. In this way, the number of the sample was reduced to 318 companies, which is the number used during the analysis. In the first phase, we gathered all the available reports on the company websites, and subsequently contacted all the companies that did not publish reports to ask for information about non-financial reporting. When the reports found or provided by companies were sufficient to clarify our research goal, we proceeded to the analysis. If not, we contacted the companies by phone in order to clarify the missing aspects. Diffusion has been measured considering the degree of use of the various reports related to social and sustainability aspects such as sustainability reports, social reports, social and environmental reports, integrated reports and responsibility reports. Moreover, to measure diffusion, we also considered other communication tools or initiatives aimed at disclosing information about social and sustainability aspects.
First, we developed a manual content analysis [53,54] to assess disclosure quality. In this way, we highlighted and coded every sentence in the reports related to the six principles of quality. However, reports collected or provided were analyzed by all authors to ensure the reliability of the results. All information collected was classified, by all the authors, according to the operationalization scheme defined in Table 1. In the case of discrepancies between the authors, the results were reviewed and discussed to overcome subjectivity problems in the evaluation. Table 1. Quality factor measurement framework.

Quality Principle Measurement
Clarity and accuracy Five possible levels were defined: -Level 1: the information is unclear and not accurate: there are few quantitative data, the presentation has not been carried out clearly and the application of the triple bottom line approach is not explicit; -Level 2: the information is partially clear and accurate: some quantitative data appear and the presentation is not too accurate; however, the triple bottom line approach is satisfactorily explained; -Level 3: the information is quite clear and accurate: there is an adequate number of data and the presentation is quite accurate; the triple bottom line approach is clear and explicit; -Level 4: the information is clear and accurate: there is a significant number of data and the presentation is accurate; the triple bottom line approach is clear and explicit; -Level 5: the information is very clear and accurate: it includes all data considered important for external communication, using a presentation framework that is well structured and easy to read for each category of stakeholder; all this information is developed in a context that respects and is consistent with the triple bottom line approach.

Quality Principle Measurement
Timeliness and stakeholder engagement Five possible level were identified as follows: -Level 1: stakeholder engagement is not addressed at all; -Level 2: stakeholder engagement is addressed, but without any clear and explicit reference to the communication processes, elaboration of the report and measurement of results; -Level 3: stakeholder engagement is dealt in a fair manner, with a particular focus on communication initiatives, but also with poor or absent references to the profiles of involvement during the process of drafting the document and measuring the results; -Level 4: stakeholder engagement is addressed clearly, with a particular focus on communication initiatives and on those related to the process of drafting the document; however, there are no references to the involvement of the stakeholders in the results measurement processes; -Level 5: stakeholder engagement is clear and complete, with a particular focus on communication initiatives and on those aspects related to the processes of preparation of the report and measurement of the results.

Comparability
This principle is equivalent to the existence of a comparison system between the results obtained and the previously stated objectives. We consider also the presence of a possible intermediate level, i.e., the presentation in a comparative manner of the results of the past year compared with the previous ones, although without explicitly mentioning the relationship between declared goals and results. We subsequently attribute a score ranging from 1 to 3 as follows: Level 1: absence of any comparative reference between the results obtained in the past and previous years; Level 2: presence of a comparison of data and results, without explicit reference to the relationship of these results with the previously stated objectives; Level 3: presence of a comparison of data and results, also related to the previously stated objectives.

Reliability
Adoption of an assurance process We used seven stages of empirical research during almost five years of activity, which led to the collection of the results contained in this section:

1.
Analysis of websites and direct contact with the companies (when necessary), finalized for the collection of non-financial reporting documents (from July 2012 to November 2012); 2.
First data analysis, related to the diffusion of non-financial reporting statistics (from January to March 2013); 3.
Participation in research dissemination activities related to the first results; debate with the academic community and acquisition of some ideas from colleagues about the development of the quality factor analysis (from May to September 2013); 4.
Elaboration of the quality factor measurement framework (as shown in Table 1), to be used for the analysis of the collected documents (from November 2013 to June 2014); 5.
Content analysis made by the single researchers of this project on the collected documents and first score assignment by each one, using the quality factor measurement framework (from September 2014 to July 2015); 6.
Analysis of the single score assignment made by the researchers, comparison and analysis of the differences and sharing of a unique final score for every company (from September 2015 to July 2016); 7.
Elaboration and group validation of the final statistics, charts and table creation (from September 2016 to January 2017).

Empirical Results
As reported in the previous sections of the paper, the aim of the empirical research is twofold: • To measure the level of diffusion of non-financial reporting for sustainability in the Italian local utilities sector; • To analyze the level of quality of this kind of information, based on the principles of (1) clarity and accuracy, (2) timeliness and stakeholder engagement, (3) comparability, and (4) reliability.

Level of Diffusion of Non-Financial Reporting
The 318 companies were initially subdivided into three groups, depending on the kind of ownership structure (entirely public, public-private or entirely private), as presented in Table 2.  Table 2 shows that a clear majority of the companies are publicly owned. This first consideration helps in the analysis of the results about the diffusion of non-financial reporting in the sample. The data regarding the presence of non-financial reporting, considering the whole sample, are presented in Table 3.  Table 3 shows that only 11 per cent of the companies are involved in processes of producing reports about their own non-financial reporting, with an average of 4.57 years since adoption. This means that non-financial reporting has assumed a certain regularity and a certain duration. Therefore, non-financial reporting cannot be considered as a simple "trend" or a temporary event for those companies. The figure could be as high as 13 per cent if we consider those companies that drafted a social responsibility or sustainability report for at least one year, starting from the financial year 2005, but later decided to abandon it.
Before proceeding with a deeper analysis of these results, we will display the specific labels adopted by the companies for their non-financial reporting documents. The results of this analysis are presented in the Table 4. This table shows, that for these kinds of companies, "sustainability report", which is inspired by the GRI guidelines, is the most common label. The use of different denominations, like "social", "social and environmental", "integrated", and "responsibility report", which are less common, does not imply particular differences in the standards, reporting scheme or contents of the document, but rather Sustainability 2020, 12, 4525 9 of 17 seems to express the willingness of the company to emphasize, on a communicative level, a certain aspect of reporting (social, socio-environmental, responsibility). Integrated reporting partially deviates from this picture, because these documents (albeit only two cases) make explicit the desire to integrate financial and non-financial information and this reflects the structure and general purposes of the document. However, we should consider that this picture shows the situation before the publication of the IIRC (International Integrated Reporting Council) framework ((2013) for the development of integrated reporting practices. Therefore, the structure of these integrated reports is not inspired by that specific standard, but by some reporting practices already existing before 2013.
Regarding ownership type, the following table distinguishes the results considering public, public-private and private companies.
Results (Table 5) show that only 6.25 per cent and 5 per cent of, respectively, entirely public and entirely private companies publish non-financial reporting, whereas for public-private companies, this percentage grows to the unexpected level of 47.37 per cent, almost one in two. Each of the seven companies that decided to abandon non-financial reporting is publicly owned. This highlights a possible peculiarity of the strategic choices made by wholly publicly owned companies. Those strategic decisions seem less stable and durable than those that also involve private investors. From this perspective, management decisions might be influenced by politics, especially in a sector like this, i.e., the public/local utility sector. For companies that are entirely privately owned, it may seem surprising to find similar percentages to entirely publicly owned companies, and such a big difference from public-private companies. This may be explained by the fact that these companies are almost exclusively very small, and most of them are cooperatives that mostly supply electricity to small municipalities, in some cases in the mountains or on small islands. Therefore, such companies do not have much in common with the majority of the public-private companies and, at the same time, they probably do not need tools of social and environmental reporting either, thanks to the close contact with their users and due to their small size.
Another observation supporting what we have already stated in this paper is in relation to the presence of communication tools or initiatives when a social responsibility or sustainability report is absent. Although they cannot be considered reports, these tools and initiatives are intended to account for at least some aspects of a company's social and environmental performance. Moreover, they are not the result of sporadic episodes, because they have a certain regularity over time. Table 6 shows the diffusion of this kind of initiative.  Table 7 summarizes the level of attention to social and environmental performance in our sample of public utility companies. The data considers companies that use non-financial reporting (included in Table 5) and companies using other initiatives that relate to accountability for at least some aspects of social and environmental performance ( Table 6). The overall result shows that 21.70 per cent explore social and environmental responsibility in a more or less extensive way (in the first instance, with the presence of a social responsibility or sustainability report). However, about 74 per cent of public-private companies pay extensive attention to social and environmental responsibility, which is much higher than the overall average.

The Quality of Non-Financial Reporting
We now turn our attention to an analysis of the quality of the collected non-financial reports, in order to highlight clarity and accuracy, timeliness and stakeholder engagement, comparability and reliability. As stated previously, the GRI guidelines are the main references we use.
Firstly, we consider the number of non-financial reporting practices that are officially inspired by the GRI, as reported in the documents. This shows that 27 out of 35 (87.10 per cent) reports are explicitly consistent with the GRI guidelines. This result only provides an initial assessment of the levels of consistency and comparability among reports, and therefore it is not sufficient to express a solid evaluation of the intrinsic informative quality of such documents.
Again, the analysis of the quality of non-financial reporting documents was based on four parameters: (1) clarity and accuracy, (2) timeliness and stakeholder engagement, (3) comparability, and (4) reliability.
With reference to the criteria of clarity and accuracy, the five levels shown in Table 1 led us to the following result in Figure 1:

Total 34 283 318%
The overall result shows that 21.70 per cent explore social and environmental responsibility in a more or less extensive way (in the first instance, with the presence of a social responsibility or sustainability report). However, about 74 per cent of public-private companies pay extensive attention to social and environmental responsibility, which is much higher than the overall average.

The Quality of Non-Financial Reporting
We now turn our attention to an analysis of the quality of the collected non-financial reports, in order to highlight clarity and accuracy, timeliness and stakeholder engagement, comparability and reliability. As stated previously, the GRI guidelines are the main references we use.
Firstly, we consider the number of non-financial reporting practices that are officially inspired by the GRI, as reported in the documents. This shows that 27 out of 35 (87.10 per cent) reports are explicitly consistent with the GRI guidelines. This result only provides an initial assessment of the levels of consistency and comparability among reports, and therefore it is not sufficient to express a solid evaluation of the intrinsic informative quality of such documents.
Again, the analysis of the quality of non-financial reporting documents was based on four parameters: (1) clarity and accuracy, (2) timeliness and stakeholder engagement, (3) comparability, and (4) reliability.
With reference to the criteria of clarity and accuracy, the five levels shown in Table 1 led us to the following result in figure 1: These data show that accuracy and clarity overall is satisfactory, with excellence in six cases out of 35 (17.14 per cent) and good quality in 10 cases (28.57 per cent). Therefore, these criteria showed a These data show that accuracy and clarity overall is satisfactory, with excellence in six cases out of 35 (17.14 per cent) and good quality in 10 cases (28.57 per cent). Therefore, these criteria showed a positive result, with an average level of 3.40.
In terms of timeliness and stakeholder engagement in the processes of non-financial reporting, such engagement can emerge in very different ways. This could consist of a simple external communication of the data in the report or in the actual engagement of stakeholders in the measurement of the variables of social impact, or within the processes of drafting documents. Note that this criterion does not provide an evaluation of the company's social involvement capability in general, but rather its capability to directly engage stakeholders in the processes of drafting and communicating non-financial reports.
The data show (Figure 2) that the most common level of stakeholder engagement is the medium level, equal to three, but also that the distribution is not balanced, with more cases of level 1 than 5 (four versus one) and of level 2 rather than 4 (11 versus seven). Only one company seems to have attempted to account for the full engagement of stakeholders, involving measuring results and presenting activities ranging from document drafting to the final communication of data. This leads to an average level for this criterion of 2.71. The data show (figure 2) that the most common level of stakeholder engagement is the medium level, equal to three, but also that the distribution is not balanced, with more cases of level 1 than 5 (four versus one) and of level 2 rather than 4 (11 versus seven). Only one company seems to have attempted to account for the full engagement of stakeholders, involving measuring results and presenting activities ranging from document drafting to the final communication of data. This leads to an average level for this criterion of 2.71. In relation to the comparability of the reported data with respect to previous reports, this requirement essentially comprised a reference system, made explicit in the sustainability documents, comparing obtained results with previously stated objectives. During the quantitative analysis, we also considered the existence of a potential intermediate level for the presence of this requirement, as reported in Table 1. In order to assign a quantitative scale of assessment that would be comparable with the other criteria of quality, we assigned these quantitative values: • a quantitative value of one for level 1 (the absence of any comparative reference between the results obtained in the last and previous years); • a quantitative value of three for level 2 (the presence of a comparison of data and results, without explicit reference to the relationship of these results with the previously stated objectives); • a quantitative value of five for level 3 (the presence of a comparison of data and results related to the previously stated objectives).
The following chart summarizes the results obtained. Figure 3 shows that there is only one case where a company reported self-referential information, limited to the last financial year only. The most common profile is the comparison of the data from different financial years, but without a specific reference to the objectives stated in previous financial years (23 cases out of 35, equal to 65.71 per cent). Eleven companies provided a good level of comparability. These data provide an average value of 3.57-slightly higher than the combined criteria of clarity and accuracy. Overall, this shows that companies are aware of the importance of providing a clear picture of the temporal trend of their values (in the best cases, even in comparison with their previous statements), even if there is still some room for improvement. In relation to the comparability of the reported data with respect to previous reports, this requirement essentially comprised a reference system, made explicit in the sustainability documents, comparing obtained results with previously stated objectives. During the quantitative analysis, we also considered the existence of a potential intermediate level for the presence of this requirement, as reported in Table 1. In order to assign a quantitative scale of assessment that would be comparable with the other criteria of quality, we assigned these quantitative values: • a quantitative value of one for level 1 (the absence of any comparative reference between the results obtained in the last and previous years); • a quantitative value of three for level 2 (the presence of a comparison of data and results, without explicit reference to the relationship of these results with the previously stated objectives); • a quantitative value of five for level 3 (the presence of a comparison of data and results related to the previously stated objectives).
The following chart summarizes the results obtained. Figure 3 shows that there is only one case where a company reported self-referential information, limited to the last financial year only. The most common profile is the comparison of the data from different financial years, but without a specific reference to the objectives stated in previous financial years (23 cases out of 35, equal to 65.71 per cent). Eleven companies provided a good level of comparability. These data provide an average value of 3.57-slightly higher than the combined criteria of clarity and accuracy. Overall, this shows that companies are aware of the importance of providing a clear picture of the temporal trend of their values (in the best cases, even in comparison with their previous statements), even if there is still some room for improvement. specific reference to the objectives stated in previous financial years (23 cases out of 35, equal to 65.71 per cent). Eleven companies provided a good level of comparability. These data provide an average value of 3.57-slightly higher than the combined criteria of clarity and accuracy. Overall, this shows that companies are aware of the importance of providing a clear picture of the temporal trend of their values (in the best cases, even in comparison with their previous statements), even if there is still some room for improvement.  Considering reliability, we used a "yes/no" criterion for assurance processes, with only two possibilities: presence or absence. The results show that 10 of the 35 analyzed reports had been submitted for external assurance (28.57 per cent of the overall total). This is an unsatisfactory result, with less than one-third of reports meeting this criterion for quality. For this factor, it seems that a culture of complete transparency still has some room for improvement.

Discussion and Final Remarks
We contribute to the literature on non-financial reporting by providing a preliminary understanding of the pattern of reporting practices in Italian public utilities. Our research contributes to analyzing both the diffusion and quality disclosure patterns of non-financial reporting. Regarding diffusion, unlike our expectations and despite the growth of non-financial reporting in organizational life [59][60][61], our results show that it is not particularly diffused within public utilities. Reasons why non-financial reporting is not widespread in public utility companies remain to be discovered. What surprises us is that, despite the importance of non-financial reporting for these companies, these tools are not widespread. These companies interact with various stakeholders and need to legitimize their behavior, yet they do not use non-financial reporting to be more accountable and to legitimate their choices. This is in contrast with what is stated by the stakeholder and legitimization theories. However, this confirms that non-financial reporting is advancing very slowly in the public sector [62]. The results also show that just 11 per cent of the total number of companies publish non-financial reporting. Of these, non-financial reporting was more prevalent within public-private companies than either public companies or private companies. In fact, about 6.25 per cent and five per cent, respectively, use non-financial reporting, whereas about 47.37 per cent of public-private companies do. Considering that a public-private company is normally an evolutionary form of a (previously) totally public company, the choice of undertaking non-financial reporting could be influenced by this passage from traditional public management towards a form of management undertaken with a private partner. This appears to be a new possible explanation supporting legitimacy and stakeholder theories [12], paying particular attention to the role of non-financial reporting to gain legitimacy through transparency. The key to adopting non-financial reporting appears to be the private partner. The latter could pressure public utility companies to adopt non-financial reporting in order to account for the various stakeholders and legitimize their behavior.
We also considered the use of other communication tools or initiatives to disclose non-financial information. Results show that about 11 per cent used these less formal methods. This is means that about 21.70 per cent regarded social and environmental reporting as important, with about half using non-financial reporting and the other half using other communication tools. Our results confirm the public sector's increasing attention to social, environmental and sustainability issues [41,[59][60][61]63]. However, the results reveal an avenue for further research to discover why these companies prefer not to use formal non-financial information-reporting typologies. On the other hand, our analysis confirms that the GRI guidelines are widely used by organizations that disclose non-financial information. The GRI guidelines are currently the most widely acknowledged guidelines for non-financial reporting [45]. This legitimizes the use of GRI as a reference to measure the quality of non-financial reporting.
Our results also show that about two per cent of companies completely abandoned non-financial reporting-all were public companies. The reasons for abandoning could be a lack of interest from stakeholders, or the cost of the process [41]. However, non-financial reporting can also be considered a management trend, and not a conscious process of organizational change [64,65]. Organizations tend to abandon practices if there are no short-term benefits, as these practices do not depend on rational behavior [65]. Our document analysis confirms that organizations use terms such as social, sustainable or other types of non-financial reporting interchangeably [5], although they are conceptually different [2]; however, most used sustainable reporting.
Regarding the quality of non-financial reporting, the results depict a gray picture, as summarized in Table 8. While the levels of accuracy, clarity and comparability are satisfactory, timeliness, stakeholder engagement and reliability are less so. In particular, the reliability of non-financial reporting seems unacceptable, with only 28 per cent of companies using some form of external assurance. This reduces the overall level of credibility of the process. Regarding clarity and accuracy, our results show that the analyzed reports include detailed and quantified indicators, which provide a gauge of the company's commitment to sustainability policies. Comparability indicates how the company is getting better or worse compared with previous periods. This is also important for stakeholders to gain a sense of whether the company is achieving its sustainability objectives. Clarity, accuracy and comparability could not be sufficient to measure quality. Because, as highlighted by Stacchezzini [18], a company could provide information about its sustainability performance when its social and environmental results are poor. Accordingly, the quality of non-financial information has to be assessed not only by considering clarity, accuracy and comparability, but also the other principles, such as stakeholder engagement and reliability, which provide credibility for non-financial reporting [15]. Moreover, our results show that reliability, timeliness and stakeholder engagement are critical, and require more interventions to improve them. Stakeholder engagement represents a fundamental aspect to increase the quality of non-financial reporting. The reasons why stakeholders are involved so infrequently require further study.
A lack of reliability, measured by the use of external assurance, could affect credibility and the perceived quality of non-financial information provided [15,18,66]. A lack of reliability could also put non-financial reporting in doubt as a reliable public source of information and knowledge [67]. In addition, the lack of timeliness and stakeholder engagement in decision making regarding sustainability policies could create the perception that non-financial reporting is only a window-dressing strategy [68]. Accordingly, non-financial reporting could be mainly used as a means to improve an organization's reputation and legitimacy [33,34], and not as a means to assess sustainable performance and to communicate with stakeholders [67].
Overall, our results represent a base for further reflections and analysis. In fact, as highlighted by [69], Directive 2014/95/EU, which introduced the mandatory requirement for non-financial information and the guidelines that detail how to prepare mandatory information, could improve the quality and credibility of that information. Directive 2014/95/EU ruled that "public interest entities" (e.g., companies whose transferable securities are traded on a regulated market, or that are of significant public relevance, such as credit institutions and insurance companies with more than 500 employees) have an obligation to communicate non-financial information concerning environmental and social concerns. This is particularly the case because the quality requirements of the Directive exceed those of the GRI. Further research is therefore needed to assess the impact of the introduction of Directive 2014/95/EU. However, further research is also required to verify the use of non-financial reporting as a managerial tool. This is because the use of a non-financial reporting template (such as GRI) could increase the symbolic performance of companies by being able to "tick more GRI boxes", and could decrease the substantive performance [15]. In our research, we analyzed how the type of ownership (public, private and mixed) could influence the adoption of non-financial reporting. Other variables like industrial sector, size, and financial structure, could be also relevant in investigating the factors influencing the adoption. Further research is needed to explain how those variables, or others, could affect the adoption of non-financial reporting. The practical and academic implications of our findings are substantial. This paper provides researchers with a practical model to measure the quality of non-financial reporting. The key element in the diffusion of non-financial reporting appears to be the presence of a private partner. Public regulators must also encourage purely public companies to use non-financial instruments in order to improve their management and make them accountable to stakeholders. The quality of non-financial reporting is still questionable, particularly due to the lack of stakeholder engagement and reliability. Policies to increase the involvement of stakeholders and to use more external assurance could help to enhance the quality of non-financial reporting. Finally, this research is limited in its focus to a sample of Italian public companies, and it adopted a qualitative scoring system. The results, therefore, call for more studies providing additional knowledge and understanding of the diffusion and quality of non-financial reporting.
Author Contributions: The paper was the result of a joint effort made by the authors, however, the paper's sections can be attributed as follows: Sections 1, 3, 5.