Integrated Reporting and European State-Owned Enterprises: A Disclosure Analysis Pre and Post 2014 / 95 / EU

: The European Directive 2014 / 95 / EU regulating the disclosure of non-ﬁnancial information for public interest organisations is enjoying its ﬁrst years since entering into force in 2017. The emerging of social, environmental and sustainability issues in combination with the New Public Management (NPM) reforms, led public sector entities to huge demands of accountability. Long time before the European Union Directive (EUD) on non-ﬁnancial information, public sector entities were pushed to demonstrate to a broad range of stakeholders how public resources are used. Accordingly, the stakeholders’ increasing demand for social and environmental information has encouraged the adoption of di ﬀ erent types of reports by organisations, such as the Corporate Social Responsibility (CSR) Report, Sustainability Reporting (SR) and the Integrated Report (IR).In the context of State-Owned Enterprises (SOEs), the disclosure of non-ﬁnancial information gains a pivotal relevance as these type of organisations face a more comprehensive range of stakeholders than private organisations. In this vein, the present paper aims to investigate whether the mandatory disclosure directive increased the level of information provided by SOEs issuing an IR between the years 2016 and 2017 in order to demonstrate whether a mandatory regulation leads to higher disclosure.


Introduction
In recent years, global financial crises, accounting scandals, environmental accidents and employees' matters have progressively increased the concerns of investors and stakeholders [1,2]. In particular, State-Owned Enterprises (SOEs), deemed central actors in progress towards the delivery of sustainable development, have been put under the severe scrutiny of stakeholders [3,4]. In addition, the introduction of the 17 Sustainable Development Goals (SDGs) by the United Nations in 2015, aiming to create sustainable development for the future [5], further developed social expectations surrounding states and private organisations alike. As a consequence, non-financial disclosure has gained momentum as a vehicle to improve transparency and accountability [2,6], especially in the SOEs context, where the existence of multiple stakeholders with multifarious interests has created considerable pressures [7,8]. In addition, the blurred lines of ownership and accountability these organisations face [9] render the development of non-financial reporting in these organisations interesting for scholars and practitioners alike, as such organisations are providing public value by acting as private organisations [8].
The EUD addresses large undertakings considered entities of public interest that have more than 500 employees [27]. It represents an essential regulatory step towards the harmonisation of non-financial information among the EU State members, requiring the disclosure of a minimum content of information regarding environmental, social and employee matters, respect for human rights, anti-corruption and bribery matters [2,10].
According to the EUD, for each of these topics the undertakings shall provide: "(a) a brief description of the undertaking's business model; (b) a description of the policies pursued by the undertaking in relation to those matters, including due diligence processes implemented; (c) the outcome of those policies; (d) the principal risks related to those matters linked to the undertaking's operations including, where relevant and proportionate, its business relationships, products or services which are likely to cause adverse impacts in those areas, and how the undertaking manages those risks; (e) non-financial key performance indicators relevant to the particular business" [27] (p. [4][5].
The EUD does not impose a particular form of report to provide this information, leaving the possibility to choose among a separate non-financial statement or the inclusion of such information in the management report [28]. Moreover, it is also flexible about the use of the reference framework, permitting the use of national, European or recognised international frameworks, provided that entities indicate which frameworks they have relied upon [27].
Some scholars converge on the idea that IRF may represent one of the most suitable reference frameworks to disclose non-financial information under the EUD requirements [2,10]. In particular, the connections between policymakers and politicians involved both in the legislation and IR movement should foster the general adoption of IR as a tool to adhere with the EUD [2,10]. The similarity between the EUD requirements and the IRF have been explored by Manes-Rossi et al. [11], who demonstrated that organisations can adopt the IR framework in order to respond to the EUD requirements.
The high expectations surrounding the emergence of the EUD led scholars to investigate the effects of the EUD on corporate reporting disclosure. To this extent, a broad range of contributions has been observed. On a national level, Sierra-Garcia et al. [29] investigated the application of the EUD in the case of Spanish companies, and Tiron-Tudor et al. [30] analysed the change brought by the EUD on non-financial information by Romanian listed companies, for the year previous and after the entry into force of the regulation. Moreover, Venturelli et al. [6] analysed the readiness of large, Italian organisations' reports for the upcoming EUD, while Doni et al. [31] explored the quality of non-financial reports in relation to compliance toward the EUD, in the case of Italian companies foreseen by the entrance into force of the Directive. Furthermore, from a comparative perspective, Venturelli et al. [32] compared the compliance levels toward the EUD provided by Italian and British companies. Similarly, Dumitru et al. [33] investigated the quality of non-financial reports issued by Polish and Romanian listed companies before the entrance into force of the Directive. On an international level, Manes-Rossi et al. [11] explored the compliance level of the 50 largest European organisations in relation to the European Guidelines, related to the EUD.
Despite the considerable awareness that has grown around the EUD on non-financial information, no studies have been found that specifically address the impact of EUD on IR disclosure provided by SOEs even though these organisations are also affected by the regulation shift.

Integrated Reporting
Integrated reporting represents the last development in the theme of corporate reporting, based on the idea of integration in a single document of financial and non-financial information [15,34].
It constitutes a structured reaction to the growing investors and stakeholders' needs, where the former asks for more useful and relevant information about firms' financial and economic domains, mainly linked to the corporate value, and the latter require more information about social and environmental matters [13]. In this end, it is meant to overcome the drawbacks of previous stand-alone (e.g., sustainability reports) and traditional reports (financial reports), which discuss social, environmental and financial information as "unconnected silos" [13] (p. 1191), also neglecting the importance of intangible assets and governance features [34][35][36].
As such, IR is a single tool which provides a holistic representation of all the dimensions which influence the organisation's ability to create value over time in a forward-looking perspective [15,34,37,38]. It is based on principles and content elements established in the framework released in 2013 by the International Integrated Reporting Council (IIRC), "a global coalition of regulators, investors, companies, standard setters, the accounting profession and NGOs" [16] (p. 1). The guiding principles are "strategic focus and future orientation; connectivity of information; stakeholder relationships; materiality; conciseness; reliability and completeness; and consistency and comparability" [16] (p. 16). To this extent, the principle-based concept on which the IRF is based, resulting from a mimicry process with other standards and frameworks [39], sets the framework in the forefront for the inclusion and disclosure of the SDGs [12], creating further necessity to understand how current accounting techniques are, or can be, used in support of the spreading of the SDGs [40]. Notwithstanding the emergence of the SDGs and the flourishing adoption of the IRF, no contributions that empirically investigate the adoption of SDGs in the IR context have been found. However, contributions investigating the relationship between the adoption of the SDGs among GRI adopters are already available [41,42].
The content elements include organisational overview and external environment, governance, business model, risks and opportunities, strategy and resource allocation, performance, outlook, basis of preparation and presentation and general reporting guidance [16] (p. 5).
Although IRF has been primarily designed for for-profit companies, it can also be adopted and possibly adapted in the public sector and not-for-profit context [16]. Consequently, in recent years, the debate about IR application in the public sector context has been fervent [18,19,21,43,44]. In this vein, the Charted Institute of Public Finance & Accounting (CIPFA) delivered a study on the IR potential for public sector organisations [45]. According to scholars [17,20,21], IR may represent a useful tool for SOEs to evidence all the interconnections between the factors involved in the value creation process.
Moreover, as argued by Tirado-Valencia et al. [44] (p. 3), "[T]he conceptual framework of IR is well-adapted to the public sector, because state-owned enterprises usually operate under market conditions, and the primary purpose of IR is to explain to financial capital providers how an organisation creates value over time". Further, considering the full range of stakeholders orbiting around SOEs, IR can serve as an instrument to obtain their legitimacy and improve their engagement [17,43].
Although IR is in its early stages, an increasing number of studies have investigated the various issues related to IR [25,38,46,47]. However, most of the studies focused on theoretical aspects regarding the role of IR with related challenges and weaknesses and the IRF and the perceptions of stakeholders and investors about IR adoption [14,37,38,47]. Conversely, a few studies are investigating the IR disclosure practices in terms of disclosure and compliance levels with IRF. The existing studies are concentrated on private sector [14,25,38,46,47], while, the research in public sector is still limited [19,24,43,44,48].
A first strand of IR disclosure studies has been conducted before the completion of the IRF in 2013 and was based on previous guidelines introduced by the Integrated Reporting Committee of South Africa (IRCSA) in Discussion Paper issued in 2011 [49].
Among these, Marx and Mohammadali-Haji [50] investigated IR practices of the top 40 companies on the Johannesburg Securities Exchange, South Africa (JSE), finding that that the IR quality of the selected companies varies from excellent to poor with only a partial level of compliance with IIRC 2011 guidelines. Setia et al. [36] analysed the corporate reports of the top 25 companies listed on the JSE immediately before (2009-2010) and after (2011-2012) the regulation of IRs [49] by focusing on the disclosure of four capitals (human capital, natural capital, social capital and intellectual capital). Their results pinpoint that companies report more non-financial information when the disclosures are under increased scrutiny via regulated disclosure requirements and that JSE-listed companies disclose significantly more information on social and relational capital in IRs (2011/2012) compared to annual reports (2009/2010). Then, with the final development of the IR framework in 2013, other disclosure and compliance studies emerged.
Based on strategic and institutional legitimacy theoretical framework, Haji and Anifowose [50], analysed a sample of 246 IRs of large South African companies, over a three-year period (2011)(2012)(2013), to test the difference in disclosure practices pre and post IR "apply or explain" introduction. Results evidenced a significant increase in the level of disclosure following the adoption of IR practice, especially regarding human and intellectual capital. Moreover, they observed a trend towards the institutionalisation of IR disclosure across and within the different industry sectors.
Others, explored the level of IR disclosure provided by firms included in the IIRC's pilot programme. Focusing on four <IR> framework areas (the guiding principles of connectivity and materiality, and two content elements, the business model and governance), Rivera-Arrubla et al. [14] found medium levels of IR disclosure and some significant explanatory factors: environment of organisations (i.e., region and industry), assurance of the report and publication in the IIRC website. Investigating a sample of eight financial sector European companies for the year 2015. Sofian and Dumitru [46] found, on average, a medium level of compliance and that each sampled company differs from the others with respect to at least one of the guiding principles or fundamental concepts of the <IR> framework. Using the case of Johannesburg Stock Exchange (JSE) listed companies, Zhou et al. [25] examine the level of adherence with the <IR> framework and the influence on the analyst forecast error and the equity cost. They found that analyst forecast error as well as the cost of equity capital are significant and negatively associated with the firms' level of compliance with the <IRF>.
In the case of six French companies drafting IRs, Albertini [51] finds that social, relational and financial capitals were the most disclosed capitals, while natural capital was the least discussed. Moreover, she observed a prevalence of increases in capital or positive information rather than negative or decreases in capital information. Also using a small sample of five companies that are expected to be superior IR reporters, Liu et al. [15] applied a normative benchmark, to examine the level of compliance with the <IRF>. The analysis revealed both a low level of compliance with the <IRF> and a lack of connections between the various strands of information (e.g., financial and non-financial capitals). Human and social and relational capital are the most disclosed capitals, while business model, organisational overview and external environment and governance were the most disclosed content elements.
Pistoni et al. [47] conducted a comparative analysis for the years 2013 and 2014 on a sample of 58 companies included in the IIRC database. They assessed the level and quality of IR disclosure provided in accordance with the IRF requirements. Although they observed a general improvement in the quality of the content from 2013 to 2014, the overall IR quality was still inadequate with more attention paid to the form than to the content of the reports. Kilic and Kuzey [38] focused on a sample of 64 Turkish listed firms by examining the compliance level of corporate reports (i.e., traditional annual reports and stand-alone sustainability reports) with the IRF required content for the year 2015. Results showed that reports discuss generic risk information rather than company-specific, provide more positive than negative information as well as more backwards-looking information rather than forward-looking information, present financial and non-financial initiatives separately and lack a strategic focus. On the other hand, in the public sector realm, Guthrie et al. [24] investigated a sample of five Italian public sector organisations to assess whether adoption of the IR concretely contributes to the advancement of integrated thinking. Through semi-structured interviews, reports and website analysis, the authors suggest that the adoption of IR can lead to change in organisations in terms of integrated thinking adoption. Additionally, Montecalvo et al. [48] conducted a case study on a SOE based on the analysis of 15 years of annual reports and IRs, supplemented by interviews, to examine how the IR had influenced its sustainability reporting practices. They observed that IR implementation positively affected the balance and content of sustainability disclosures. Manes Rossi [19] conducted a case study on six PSOs to assess the extent to which the IR framework is followed and if the IR may represent an appropriate tool to improve stakeholders' engagement in PSOs. Focusing on four main aspects, business model, materiality, conciseness and stakeholder engagement, the author concludes that the IR framework does not provide sufficient support for public sector entities to be considered as the best reference for accountability purposes. Farneti et al. [43] conducted a longitudinal case study, from 2009 to 2017, based on content analysis and interviews with key managers of a New Zealand SOE, to examine if the adoption of the IR influences social disclosures. They observed that, during the period 2009-2012, when the GRI framework was used, there were more social disclosures than in the 2013-2017 period when disclosure was driven by the IRF. Moreover, in the managers' opinions, IR adoption had fostered the engagement of both internal and external stakeholders. Finally, Tirado-Valencia et al. [44] focused their attention on the integrated thinking dimension of the IR, investigating a sample of 17 SOEs for the period between 2013 and 2017, to examine the incorporation of integrated thinking in IR report preparation. Their analysis evidenced that SOEs are still far from the full incorporation of integrated thinking in IR disclosure and that in the external approach dimension, connections of the environment and the impact on society with value creation are relatively frequent, while connections related to commitment to stakeholder expectations are infrequent.
Thus, by considering previous studies, a gap emerges in the literature, because there is limited research assessing the level of IR disclosure in compliance with the IRF under the EUD regulatory pressure. The majority of studies have been conducted in the private sector with a lack of empirical evidence in the public sector realm.
In order to fill this gap, this research aims at extending empirical research on IRs in the public sector field, by investigating the role of Directive 95/2014 in stimulating IR disclosure in the SOEs context. More specifically, the present study proposes a two-year analysis (2016 and 2017) to analyse the level of IR disclosure provided in compliance with IRF pre and post EUD.

Theoretical Background
According to legitimacy theory, organisations' existence and the probability of survival are strictly correlated to the perceptions of stakeholders as well as to the relationship between social expectations and organisational behaviours [36,52,53]. In order to obtain legitimacy, each organisation should act demonstrating respect to norms, social values and expectations shared by the community of stakeholders in which it is rooted [7,36,54]. If the stakeholders' community perceives a lack of congruence between its system of values and norms and that of the organisation, a legitimacy gap emerges [7,50]. Legitimacy theory takes on particularly strong connotations in the SOEs context where a more comprehensive forum of stakeholders exerting political and social influences exists, demanding information about financial and non-financial matters [8,54].
In this vein, SOEs can be considered particularly interesting as such organisations "constitute an important sector in different countries, and their response to existing and future challenges can greatly influence the development (not only economic growth) of many regions in the planet" [55] (p. 207). Moreover, SOEs can be seen as instruments used to correct market failures or promote economic development; thus, such organisations represent an example for society in the way they act [56]. Accordingly, to legitimise their actions, SOEs may manage these pressures, increasing their transparency and accountability level about financial and non-financial information [8,54]. In this aim, the IR can represent a useful tool, permitting the two strands of information (financial and non-financial) to be integrated in a single document, in turn helping SOEs in improving transparency and gaining legitimacy [17,20,21]. By using the IR, SOE managers, beyond proving the efficiency and effectiveness underpinning the use of public funds, can also demonstrate the sustainability of their activities and programs, enhancing the general level of trust and credibility [44].
Scholars identified two complementary strands of legitimacy: strategic and institutional [37,50,52,57]. Strategic legitimacy is based on an internal and managerial perspective and focuses on the strategy used to acquire or repair the legitimacy [37,50]. In the first case, organisations adopt a proactive behaviour, analysing the organisational field and disclosing additional information, especially about social and environmental issues, to demonstrate that they share stakeholders' norms, social values and expectations [37,50,52]. In the second case, a legitimacy gap exists [36]. As such, organisations adopt a reactive strategy, changing internal managerial behaviours and disclosing more information to highlight the efforts made to adhere to the socially accepted norms and values and possibly change the negative stakeholders' view [2,36].
Institutional legitimacy takes an external and institutional perspective and is based on the assumption that each organisation operates within an external environment whose pressures in terms of norms, laws, rules, routines and belief systems influence its behaviours and structures [17,53]. These pressures stimulate isomorphism, which is uniformity of practices and behaviours aimed at conforming to norms and values to gain wide acceptance [17,50,53].
Institutional isomorphism can be coercive, mimetic and normative [48,50,53,58]. Coercive isomorphism can be identified with the external pressures exerted by regulatory bodies such as the European Union, the state or local governments, which require compliance with rules, laws, decrees or regulations [48,50,59]. The Directive 95/2014 may stimulate a form of coercive isomorphism among EU Member State entities of public interests, provided that the search for standardisation and comparability of the information does not outweigh the quality of disclosure [6,59]. It can also represent a stimulus to provide a high level of disclosure through the IR [10]. Mimetic isomorphism occurs when organisations belonging to the same environment or industry sector share similar behaviours and mimic the best practice of the best organisations [50,53]. In this vein, organisations pertaining to the same industry sectors, although coming from different geographical areas, follow the routines and reporting patterns of the industry leader, to reduce uncertainty and gain more legitimacy from the environment [17,50]. Such practices, in turn, can foster a process of institutionalisation of reporting practices [50], especially in those sectors, which being particularly environmentally sensitive, are more socially exposed [8,36].
In the end, normative isomorphism arises from professional networks or industry bodies in terms of values, beliefs and social norms that provide structures of legitimate behaviour [17,50]. In this kind of isomorphism, in order to gain legitimacy, organisations tend to both adopt structures, systems and processes and adhere with values and social norms laid down by relevant professional groups [17,59].
Under a similar perspective, it is possible to assume that given the complex network of resource providers SOEs are exposed to, such organisations may try to satisfy different interests of different stakeholders that may affect their activity [60]. In this vein, stakeholder theory is grounded on the existence of social contracts between the organisation and its stakeholders [61]. Such theory permits an analysis of the relationship existing between an organisation and its stakeholders [62].

Sample Collection
The present contribution outlines changes in corporate disclosure behaviours for the years pre (2016) and post (2017) EUD adoption, in the case of European SOEs implementing the IR.
With the aim to draw a comprehensive and broader representation of IR adopters, the research sample was gathered from the official Integrated Reporting Example Database, collecting the mentioned companies from the section <IR Reporters> for the "Europe" region. Subsequently, the selection, a number of 187 companies, results in the first-step sample.
Commencing from the given list of organisations, a further two-step process of filtering was conducted. Firstly, organisations belonging to the European Union were selected. Secondly, with the extent to identify SOEs, which are "enterprises where the state, regional governments or cities have significant control, through full, majority or significant minority ownership" [63] (p. 275), an ownership analysis was pursued through the employment of Thomson Reuters EIKON. Table 1 presents the composition of the final sample per sector and country.  Hence, an overall number of 22 firms resulted from the application of the specified filters, comprising a final sample of 44 reports, providing in this manner a balanced sample for the two years analysed. Furthermore, the IR or annual reports were downloaded from the entities' official websites for the analysed years.

Analysis
The present contribution investigates the level of disclosure by a content analysis. Widely adopted in disclosures studies [35,36,50,53], content analysis grants "replicable and valid inferences from texts (or other meaningful matter) to the contexts of their use" [64] (p. 18).
Furthermore, the employed methodology was performed on the collected reports according to an already established checklist available in the literature, with the aim to guarantee the validity and reliability of the research [65]. In this vein, the present research adopted the checklist proposed by Zhou et al. [25] as its eight categories and 31 items cover all the elements foreseen by the IIRF.
The reports were evaluated based on a dichotomous approach. The mentioned scoring method was used to analyse the presence or the absence of each element included in the proposed checklist [11,25,38]. In this vein, a score of (1) was awarded in the case of presence of a determinate element and (0) whenever the absence of the element was recognised. The adoption of a checklist already existent in the literature ensures the reliability of the study [66].
Furthermore, the Integrated Reporting Disclosure Score (IRDS) was calculated as follows: where m j=1 d j m represents the average obtained by dividing the total number of disclosed elements (dj) per the total number of elements (m) included in the checklist implemented by Zhou et al. [25]. Table 2 includes the elements and categories foreseen by the adopted checklist. As regards the analysis, two of the authors separately analysed the reports, and once all the reports were analysed, results were discussed by all the research team and the final results were assigned.

Results and Discussion
A broad adoption of the Global Reporting Initiative (GRI) framework was observed in the analysed reports. In this regard, Table 3 outlines the number of SOEs adopting the GRI framework in their IR.
According to the results presented in Table 3, it is possible to note that the GRI, usually used to develop sustainability reports, is adopted by 18 organisations in both years, demonstrating that the GRI framework plays a pivotal role in providing guidance to organisations developing their IR. In addition, primary evidence of mimetic isomorphism [58] can be observed, as organisations are following what similar organisations are doing.
Moreover, the compliance analysis results demonstrate an adequate level of disclosure in both years. Table 4 illustrates the compliance analysis per the entire sample analysed for the two years. The first and third rows outline the minimum, maximum and average number of elements disclosed, out of the 31 foreseen by the checklist. The second and fourth rows expose the minimum, maximum and mean of the average of the value. The last two rows compute the evolution from 2016 to 2017 in order to outline the changes brought by the entrance into force of the EUD. On average, SOEs disclosed around 21 elements per each year, while the minimum number of elements observed a decrease from 2016 to 2017, as in the first year, the lowest number of elements disclosed were ten and in the following year eight. The maximum number of disclosed elements remained stable, as a number of 27 elements were observed. Table 3. GRI framework adoption.
Overall, for both years, the IRDS score signals a fair level of disclosure in compliance with the IRF provided by SOEs. Such results are in line with the strategic legitimacy perspective, highlighting how SOEs, by improving their disclosure scores, are exploiting the potentialities of IR to improve communication toward stakeholders, justifying their pivotal role in society and demonstrating their behaviours are consistent with accepted social norms, values and expectations [17,36,50].
For both years, the highest results were achieved by two different financial sector organisations in 2016 and 2017. Moreover, the lowest number of elements disclosed belongs to an organisation operating in the field of "Consumer Goods" for 2016 and the "Financial" sector in 2017. Table 5 presents the compliance results for each category, providing a detailed image of the changes the occurred for the previous year of the entrance into force of the EUD and after it. The results are grouped according to the categories developed by Zhou et al. [25]. The most disclosed contents for 2016 are: "Organisational Overview" (0.78), "Governance" (0.64), "Opportunities and Risk" (0.68) and "Business Model" (0.68). In 2017, even if an increased disclosure is observed, the most disclosed elements do not change as the most appreciated contents remain firmly the same of the previous year: "Organisational Overview" (0.86), "Governance" (0.72), "Opportunities and Risk" (0.83) and "Business Model" (0.71). These results are consistent with those observed in previous similar studies conducted in the private sector [38,47] and permit the following considerations. Moreover, the less disclosed categories, for both years, were "Performance and Outcomes", "Future Outlook" and "Other Elements". To this extent, it is possible to note a negative pattern, reinforced by similar results that emerged in other investigations, as the "Performance and Outcome" category was also found among the lower scores in studies focused on a European sample [11] and an investigation focused on an eastern European economy [39]. Moreover, Kilic and Kuzey [38] and Pistoni et al. [47] also highlight the lack of disclosure in the "Future Outlook" category. Similar scores were also noted in categories where the disclosure levels are low, thus the concept of mimetic isomorphism [58] behaviour is further highlighted.
Firstly, the most disclosed categories include information which, in a certain way, is already required, although in a non-binding way, by management commentary settled by international accounting standard setters (e.g., IASB, 2010). Secondly, the least disclosed categories include information challenging to quantify, particularly commercially sensitive information, to the point that organisations tend to limit its disclosure to avoid harm both in terms of competitive advantage and reputational damage [38]. Accordingly, it seems that contrary to IRF expectations, a backwards-looking disclosure aptitude persists. In this respect, it is worth noting that the approach adopted by the investigated organisations does not follow the recommendations of the IRF regarding future-looking disclosure [16]. This negatively affects investors' resource allocation and stakeholders' decision making processes as backwards-looking information does not allow accurate estimates of organisations' financial and non-financial performance. In this vein, by observing similar behaviours related to the "Future Outlook" category, it is possible to observe an additional mimetic behaviour by SOEs [58].
However, it is worth noting that some categories that are mainly foreseen by non-financial reporting highlight satisfactory results such as "Governance" and "Opportunities and Risks", demonstrating an increasing of disclosure brought by the new concepts included in the IRF.
Results also suggest that no particular variations exist between the levels of IR disclosure provided by sampled SOEs pre and post EUD introduction. Despite the mandatory character of the Directive, the IRDS does not show a significant increase from the previous year, as the overall difference is merely symbolic (0.01). The shift can also be observed in the change that occurred in terms of contents.
This result implies that a process of institutionalisation of SOEs' reporting exists due more to the acceptance of the IR as a consolidated reporting practice able to convey legitimacy rather than to the introduction of EUD [37]. Thus, the isomorphism, which occurs, is more a result of normative and mimetic pressures rather than coercive ones. From this perspective, Table 6 sheds light on the industry average levels of disclosure. Accordingly, in 2016 "Financial" organisations obtained the highest scores, followed in order by "Basic Materials", "Consumer Services" and "Utilities". The high disclosure level observed is related to the need for transparency toward stakeholders, as the activities in which these institutions are involved are considered to be particularly hazardous, and therefore a higher level of disclosure in order to ensure transparency is expected [46]. The lowest score was obtained by "Consumer Goods" organisations.
A shift in the leading disclosure sectors can be observed, as the outlook changes in 2017. "Financial" organisations lose their position of highest disclosure to companies belonging to the "Consumer Services" sector. Thus, the landscape of top disclosure is composed as follows: "Consumer Services" (0.72), "Financial" (0.71), "Utilities" (0.69), "Basic Materials" (0.67). The lowest disclosure level is still observed in the case of the "Consumer Goods" organisations, although it increases from 0.33 to 0.45.
In this regard, an institutional isomorphism occurs as companies are conforming corporate behaviour in order to gain full acceptance [17,53]. Accordingly, coercive [58] isomorphism could have been expected. However, in the investigated period, when the EUD entered into force, the character of the isomorphism followed by organisations is mimetic, as a result of the IR adoption process [50], rather than coercive. Our results highlight a similar situation pre-post EUD, where the mandatory requirement did not improve the disclosure scores [39]. As demonstrated in Table 5, industries followed a similar pattern of disclosure level among the two analysed years, demonstrating a more institutionalised mimetic approach to the use of the IR. In addition, it is possible to observe disclosure decreasing for organisations involved in the "Basic Materials" sector, which registered a lower score from the previous year, even though the EUD had entered into force, while the "Consumer Service" sector improved its scores, obtaining the highest scores. To this extent, under a stakeholder theory perspective, it is possible to assume that the increasing scrutiny by civil society, promoted by the EUD, enhanced stakeholders' expectations and, in turn, organisations facing a higher number of stakeholders improved their disclosure scores.

Conclusions
The recent introduction of directive 95/2014/EU gave new impetus to non-financial issues, requiring entities of public interest with more than 500 employees to disclose a minimum content of information about environmental, social and employee matters, respect for human rights, anti-corruption and bribery matters. Integrated reporting, as an innovative communication form combining financial and non-financial information in a single document, is considered as one of the most suitable tools to comply with the directive. Although thought primarily to address for-profit companies, the IR is emerging as a robust accountability and transparency tool also in the public sector where stakeholders' pressures about non-financial information accountability have dramatically increased.
Grounded on these premises, the present paper provides two-fold contributions to the existing literature. Firstly, it examines the state of IR disclosure, in the context of SOEs also assessing the GRI framework adoption and exploring first SDGs mention in IR. Secondly, it examines the effects of mandatory non-financial disclosure requirements (EUD) on the level of IR disclosure provided by SOEs.
The study provides empirical evidence that SOEs show a fair level of IR disclosure in compliance with the IRF for both years. This sheds light on the usefulness of the IR in increasing SOEs within their context, demonstrating convergence with the systems of social values, norms and expectations. In spite of the aims of the paper, the analysis of the reports exposed a wide tendency toward the inclusion of the SDGs in the IR, adding to the literature a primary trace for the nexus between IR and SDGs.
Moreover, the findings reveal a certain degree of isomorphism following the introduction of the EUD. Despite the mandatory requirements of the EUD, the level of disclosure provided by SOEs through their IR remained fairly stable. This signals that EUD coercive pressures did not exert a decisive influence on the level of disclosure and that the IR has become an institutionalised practice, regardless of the EUD following a more mimetic scheme, as also evidenced by results of the analysis for industries. In this sense, it is possible to consider that the introduction of a mandatory regulation did not spur non-financial disclosure by SOEs in a significant manner; thus, coercive isomorphism was not observed.
This research may have implications for academics who can replicate the study adapting the same framework in other international contexts and for regulators and policymakers who can evaluate the potential effectiveness of the EUD in stimulating non-financial disclosure in the SOEs context and the role of the IR in complying with it.
The study also has some limitations. Firstly, it examines only a limited number of organisations due to the reduced number of SOEs preparing IRs. Secondly, it analyses only one year before and one year after the introduction of the EUD. Thirdly, it employs an unweighted disclosure index to analyse the level of IR disclosure provided in compliance with the IRF. Fourthly, it does not offer an in-depth analysis of how companies are including SDGs in reports as it is not the focus of the paper.
Future studies may expand the analysis considering a comparison between the private and the public sector, examining further years before and after the establishment of the EUD. Moreover, the development of a weighted disclosure index may be suitable to capture the quality of information. Additionally, future studies could focus on a larger SOEs sample derived from another non-financial reporting framework (e.g., the Global Reporting Initiative). Moreover, a more in-depth analysis of SDGs inclusion patterns in IR may be interesting to draw further insights.
Author Contributions: All authors wrote the paper, but their primary individual contributions were as follows: Section 1 is to be ascribed to G.N., Section 2 is to be ascribed to G.N., Section 3 is to be ascribed to A.T.-T. and G.Z., Section 4 is to be ascribed to G.N. and G.Z., and Section 5 is to be ascribed to A.T.-T. and G.N. All authors have read and agreed to the published version of the manuscript.