The Non-Financial Reporting Harmonization in Europe: Evolutionary Pathways Related to the Transposition of the Directive 95/2014/EU within the Italian Context

Since 2016, the European context has been characterised by the introduction of the Directive 95/2014/EU on mandatory non-financial reporting. The Directive has been transposed in the Italian context through the Legislative Decree 254/2016. However, following evidence previously collected from other jurisdictions, a high degree of scepticism by academics still characterises the debate on the effectiveness of this form of regulation. In fact, the literature highlights that the effects of a mandatory approach to non-financial reporting are limited to an increase in the annual quantity of information provided and not by an effective organizational change. Therefore, the contribution of our paper is twofold: a gap analysis between the overall degree of non-financial reporting quality between the years 2015 and 2017 and an empirical analysis, factors of which have impacted on the quality of non-financial reports prepared in 2017.


Introduction
The EU Directive 2014/95 has significantly raised and spread awareness of Corporate Social Responsibility issues within the European Community.
This directive was transposed into Italian law by Legislative Decree no. 254 on 30th December, 2016, obliging Public Interest Entities (PIE) with more than 500 employees to disclose non-financial and diversity information in their financial statements or in separate, independent documents. More specifically, this information dealt with environmental, social, personal, human rights and anti-corruption issues.
The aim of this study is to understand how Italy is responding to this regulatory directive and, thus, contributing to the specific literature on the impact of regulation on the quality of non-financial disclosure [1][2][3][4][5][6]. In order to understand the effect of the Directive on the quality of disclosure, the level of compliance with non-financial disclosure by companies subject to the Legislative Decree 254/16 was evaluated, and a comparative analysis of the information reported before and after the introduction of the decree was performed.
To this end, the application of a specific scoring system is proposed, based on the analysis of specific items including: materiality, the representation of a business model, the communication of sustainability policy, the main risk factors and indicator systems, as well as the more specific requirements of the decree pertaining to environmental, social, personal, human rights and anti-corruption activities.

1.
The definition of company size was adopted verbatim in most countries (19/30 including Italy) and adapted to the reference context in others (including the UK).

2.
The definition of a Public Interest Entity was adapted to the reference context in the majority of countries (24/30 including Italy).

3.
Report topics and content were fully adopted from the norms of the Directive in most countries (22/30 including Italy).

4.
Report framework was fully adopted in 22 of 30 countries, but not in Italy. 5.
The placement of non-financial disclosure was adapted to context and not adopted verbatim in the majority of countries (23/30) including Italy..
Non-compliance penalties were adapted to the context in most European countries. 8.
Diversity reporting was adopted verbatim in most countries (19/30).
This analysis shows how the Directive was changed and adapted to the country context both in Italy and elsewhere. In particular, as well as a clearly indicating the type of information to be disclosed, some sections of the Legislative Decree no. 254/2016 make specific reference to the principles of relevance or significance to the well-known Community principle of comply or explain and to the principle of data comparability.
The information to be disclosed is divided into topic (environmental, social and personal, human rights, active and passive anti-corruption and bribery matters), scope (business management and organisational model, company policies, results and indicators on non-financial loans) and type (use of energy and water resources, greenhouse gas and pollutant emissions, impact of risk factors on the environment, health and safety, social aspects of employee management, gender equality, fulfilment of international conventions and interaction with trade unions, respect for human rights and measures adopted to prevent their violation or discriminatory actions or attitudes, active and passive anti-corruption measures adopted). Regardless of the reporting standard adopted, article 3, comma 2 of the decree establishes the minimum content of non-financial disclosure. Based on this compulsory content, this study evaluated the qualitative level of non-financial disclosure available, and what determines the quality of the report.

The Quality of Non-Financial Reporting: Voluntary Versus Mandatory
Over the years the debate in the literature on non-financial disclosure has focussed on the possibility of introducing a voluntary and/or mandatory element [7][8][9][10][11][12][13].
With regard to mandatory reporting, the idea that regulation could improve the quality and comparability of non-financial reporting was initially confirmed in the literature [4]. Therefore, some believed that regulation would be preferable to voluntary disclosure which may be incomplete and lack accuracy, neutrality, objectivity and comparability [1,2]. To this end, over the years, some European countries (Spain, France, Portugal, Finland, Denmark and Sweden) have introduced mandatory reporting on environmental and social issues, as previously stated.
Some comparative studies have shown that in countries with regulation in place, such as France, there is a higher quality of disclosure compared to countries where there is no regulation, such as the United States of America [3].
Moreover, according to some studies, the imposition of specific rules and reporting models by governments could result in the short term standardization of practice and a consequent increase in the number of reports containing non-financial information [6], while providing benchmarking and best practices [5].
On the other hand, a quantitative increase would not necessarily be associated with a qualitative increase in information. According to some authors, adopting a standardised framework would penalise the use of sector specific indicators. Some empirical studies have shown that regulation does not always improve the quality of non-financial reporting in financial statements [7,8,10] or that this alone can guarantee a better level of disclosure [9,10,12]. Finally, other studies [14] showed contrasting results for China and South Africa.
Advocates of voluntary reporting attach strong strategic value to the development of CSR, which spreads as a result of the value attributed to it by the company and managers, and where there is a positive relationship between non-financial reporting and company share value [15]. Therefore, the development of proactive CSR practices and the consequent (voluntary) non-financial reporting help to generate trust among investors [14,16] and improve company reputation [17]. However, the definitions of CSR in the literature [18,19] that clearly demonstrate its voluntary nature, as well as the spread of sustainable reporting at a global level in contexts with almost no normative provisions, are testimony to the effectiveness of voluntary reporting. However, as with regulation, it would seem that scientific debate on the voluntary adoption of non-financial disclosure has not yet reached a consensus.
Therefore, in order to provide an initial analysis of the impact of normativity on the quality of reporting, this study compared the qualitative level of non-financial reporting (which is highly criticised by supporters of mandatory reporting as it is considered to be too random in a context with no regulations) in a voluntary context (i.e., Italy in 2015) with a successive mandatory stage (i.e., 2017, first time adoption of Legislative decree 254/2016 in Italy).
Furthermore, as with the debate on mandatory and/or voluntary disclosure as a determining factor in the quality of reporting, other variables, which, at various levels, can potentially influence the qualitative level of reporting, have been considered in the literature. As to the factors closely related to disclosure, the literature shows how the adoption of consolidated standard setters [20][21][22] the provision of sustainability statements or integrated reports [23,24], assurances by independent third parties [25,26] and the quantity of information reported [27,28] are all factors that can positively affect the quality of reporting. Moreover, within this category, it should be noted that company size, i.e., all economic and financial factors related to the company, can influence the quality of disclosure.
The second part of this study evaluates factors that may have influenced the qualitative level of disclosure in the study sample using an OLS regression analysis.

Sample and Methodology
The aim of the present study is to assess the degree of adherence to the Directive 95/2014/EU by Italian Public Interest Entities listed on the Milan stock exchange in 2015 and 2017. The decision to evaluate the qualitative level of disclosure in these two periods was based on the different legal contexts that characterised both years. In fact, it should be stressed that both years relate, respectively, to the year prior to the transposition of the Directive in Italy and to the first year of its coming into force.
The study sample included 223 companies that, on 31 December 2015, were considered as large undertakings as defined in the Legislative Decree 254/2016, and therefore, subject to disclosure. However, for the purposes of analysis, only 147 companies that, by the study date, had published their non-financial disclosure for 2017 were considered ( Table 1). The ICB (Industry Classification Benchmark) was used for classification. Moreover, in order to better understand the study sample, it was divided by geography and experience. Two types of analysis were performed: a statistically descriptive one and an inferential one. The former is based on content analysis [29] of the company non-financial disclosure. More specifically, the disclosures conforming to Leg. Decree 254/2016 were analysed. To this end, it should be noted that the law permitted those companies that were within the requirements of the norms to submit their non-financial disclosure in a specific section of the Management Report or, alternatively, in independent documents such as the sustainability report or the integrated report. Content analysis was performed manually by researchers in order to understand the different aspects relating to the transposition of the directive in Italy and to interpret the content of the information reported. This methodology was in line with that of previous studies on the effects of transposition and/or implementation of the directive within EU member states [30][31][32][33]. This approach is preferable in accounting studies, when compared to the research on indicators (e.g., the Global Reporting Indicator) results can be useful both for understanding the whole study sample and to contribute to the scientific debate [34,35]. According to previous studies [31], a content classification system that expresses Non-Financial Scores as percentages was used. Evaluation of Non-Financial Scores was performed by numerical quantification of the level of transparency of specific features of non-financial disclosures. These referred to all information related to the business model, policy, sustainability risk management, Key Performance Indicators (KPIs) and diversity policy. Each element was awarded a score from 0 to 1 where 0 = no element; 0.5 = partial adherence; 1 = all information produced. Finally, a weighted average of the scores was calculated.
An inferential statistical analysis uses an ordinary least squares (OLS) regression analysis to evaluate the factors that, at various levels, can influence the qualitative level of disclosure. This methodology is in line with the previous literature on the impact of national and community normativity on the quality of reporting by companies [36,37]. Furthermore, to consider an adequate number of variables within our analysis, we have followed prior literature that suggests a minimum number of event-per-variable (EVP) equal to 10 [38]. The EVP consists of the ratio between the number of the variables considered within the study and the total number of observations. In this sense, we have considered a number of variables equal to 13 in order to describe the determinants of the NF_Score for the 147 observations.
The total NF_Score have been used as main dependent variable (Table 2). Furthermore, according to prior studies, we have performed separately the analysis of the five subitems that composed the NF_Score. In this sense, following the methodological approach proposed by Venturelli in 2017 [31], the six dependent variables (DV) tested in our analysis are: NF_Score, Business model, Policy, Sustainability risks, KPI and diversity.
The independent variables used were those from the previous literature [20][21][22][23][24][25][26][27][28], i.e., the third part independent assurance provided by the Big 4 accounting firms, the use of standalone reports, the number of pages, the adoption of integrated reports and the number of years since the first non-financial report. Moreover, in order to understand the impact of company structure on the willingness of management to produce non-financial reporting, some control variables relative to the economic-financial performance for the year 2017 were considered. In particular, the average number of employees, the total assets declared in the annual report, revenues, revenues per share and the debt to equity index. Finally, to understand the impact of external factors on the quality of reporting sectorial and country's effects were also considered.
According to this evidence, the general functional form of the OLS models tested in our analysis is the following:

Results and Discussion
The analysis of the non-financial declarations showed that, for 2017, businesses had a mixed approach to the first experience of non-financial reporting (Table 3). From some of the main factors analysed relating to the quality of reporting, it emerged that companies listed on the Milan stock exchange adopted, within legal limits, some significant regulations. While the legislator allowed different forms and standards of reporting to be used, the companies in the study sample mainly opted to integrate the information into the management report or, as with international companies, in the annual report. Moreover, as to the amount of detail included, most adopted the reporting standard of the Global Reporting Initiative according to the "In accordance-Core". The "comply or explain" system was also used, especially with regard to water resources and greenhouse gas emissions. The obvious difference in the number of pages demonstrates how companies adopted different ways to report information. Finally, the data on assurance highlights how the Italian market is mainly the prerogative of the Big 4. The NF_score for 2017 showed an average framework adherence of 69.43% (Table 4). This is higher compared to previous samples analysed using the same methodology [31]. The best performances were in the Business Model, as shown by the minimum and median values. However, it should be noted that the study sample strictly adhered to the requirements of KPIs, despite having a lower minimum value. Finally, the standard deviation values are worth noting. The nature of this indicator demonstrates how the average qualitative level of reporting tends towards the same/average value. This implies that regulation strongly favoured the standardization of reporting produced by the companies. This is in line with previous studies on accounting standardization and, more specifically, with studies on the shift from voluntary to mandatory disclosure [19,20]. However, it should be reiterated that the standardization of documents does not necessarily lead to the standardisation of practices by companies. In fact, this would require an integration of legal norms, relating to the area in which a company operates, with the ethical and moral norms relating to the attitude of management [7]. The comparative analysis of the years 2015 and 2017 (Table 5) shows a net improvement. This is due to the physiological increase in the number of reports that included non-financial information and to the more detailed and consolidated framework used by companies [39]. During the period studied, an average increase of 21.248% in the quality of reporting was observed. This was due to the increased awareness of companies in sectors that were hitherto less inclined towards corporate social responsibility. The greatest increase was noted in the "Basic Material" sector that includes companies operating in controversial environments that tend to be characterised by a lower willingness towards non-financial disclosure [39]. The lower increase in the quality of reporting with regard to Oil and Gas is due to the fact that, in the period prior to the introduction of the directive, these companies had already demonstrated a high level of reporting given that this sector was already strongly oriented towards CSR [40,41].
In the second stage of the study, a correlation analysis (Table 6) was performed to detect any possible collinearity between the variables analysed. However, none was found, and neither was there any causal relationship between variables. The analysis showed, however, that the IIRC variable was positively influenced by a company's previous experience in sustainability reporting and that, in the introductory phase of the directive, it was mainly adopted by these companies. It was also observed that the use of non-financial reporting was strongly correlated to the sector in which a company operates. In particular, the banking and insurance sectors mainly published disclosure separately from financial reporting. The correlation analysis showed no specific evidence related to the country in which the companies operate. The OLS analysis (Table 7) performed on the NF_Score highlighted the different factors that impacted on the qualitative performance of the study sample. The variables for the quality of non-financial reporting showed that IIRC, EXP and PAGE influenced compliance. The results for IIRC confirmed the existing literature and showed it to be a factor that can influence the pursuit of greater transparency in disclosure [42]. In fact, contrary to the GRI Standards, the IIRC consists in a set of principles inspired by the need to disclose the different relationship between types of capital [42]. Furthermore, the results related to the central role played by prior experience highlight how the adaption to the new legal requirement for these firms has been less complex due to the existence of prior reporting systems. The causality relationship between PAGE and NF_Score is particularly relevant to our study as it is linked to the quality and quantity of information reported and is dealt with separately in the literature [43,44]. Specifically, managers could disclose a large quantity of information without an effective orientation to sustainable paradigms through the adoption of greenwashing mechanisms. As to the size of companies, the variables LOG ASSET, TURNOVER and EMPLOYEES all positively influence the quality of disclosure. This confirms the findings in previous studies that willingness to non-financial disclosure tends to be positively correlated to firm size [45,46]. In fact, despite having a set of common rules about non-financial reporting, firms characterised by a high degree of financial resources could be more able to disclose their non-financial contribution through more sophisticated reports. Finally, the SECTOR variable negatively influenced the NF_Score. It seems that non-financial disclosure is more effective in companies that are not involved in banking and insurance, thus confirming the previous studies on CSR in the financial sector [47]. It has been suggested in the literature that non-financial firms are more interested in disclosing their contribution to society because they are perceived as being more controversial than financial firms.
The 4 OLS analysis performed on the subitems, revealed that IIRC was the most common variable and positively influenced the business model, sustainability risk and diversity. The EXP variable, which is important for policy and diversity, was also significant. However, it should be noted that dedicated reporting negatively influences the disclosure of sustainability risks. This could be related to the possibility for managers to exclude from their non-financial declarations part of the required information in according to the comply-or-explain principle. Finally, the COUNTRY variable was not statistically significant in this analysis, demonstrating that the social and cultural variables for the areas in which a company operates do not impact on non-financial disclosure. In this sense, the adaptation to the Decree by MNEs has not been more difficult than for firms that operate only on the Italian market.
The results show a significant standardization and similarity in company reporting. This supports the hypothesis that the regulatory role of the legislator could reduce the qualitative gap between the non-financial reports prepared by the European PIEs [2].

Conclusions
The European context has been characterised by the increased attention of regulators to sustainable development. The Directive 95/2014/EU covered a central role within this debate due to its potential role in supporting the adoption of more sustainable business models by the PIEs. Our results highlight that the overall degree of NF_Score has been equal to 69,432%. In this sense, our results confirm the existence of a qualitative increase of the non-financial declarations prepared in according to the Legislative Decree 254/2016. However, similarly to other contexts such as South Africa, the approach of some of these firms could be based on a "tick-box" method in order to be compliant with the law [48].
The analysis of the determinants of the NF_Score has provided some insights about the role covered by the firms' characteristics. The main insight is related to the negative effects connected to the sector of activity. In this sense, the provisions of a set of common rules for firms that operate in different sectors could represents a physiological limit for an effective harmonization due to their different non-financial impacts. Furthermore, the central role covered by the IIRC could represent a possible development for the policy makers to increase the overall quality of the non-financial declarations. Finally, the result related to the prior experience suggests that the early adopters have been abler to adapt their non-financial declarations to the new legal requirement. The theoretical implications of our paper are represented by the opportunity for accounting scholars to provide further details about the shift from voluntary to mandatory non-financial reporting. In fact, our results suggest how the adaption of the early adopter to the legal requirements of the Directive 95/2014/EU has been simpler than for late adopters. Furthermore, the data related to the IIRC confirms how adopting this tool is an effective sign of orientation to CSR.
The managerial implication of our paper is represented by the opportunity for managers to use our data in order to identify possible reporting strategies in order to engage in a more effective way with their stakeholders. In fact, our results have highlighted the existence of some factors that impact positively on the quality of the non-financial declarations prepared in according to the Directive 95/2014/EU.
The political implication of our paper is that insights are provided on the difficulty for first adopters to disclose their non-financial information. In fact, our results highlight how early adopters are more oriented to disclose their non-financial information than the late adopters. The limitation of our paper is represented by the short period of analysis. In this sense, the absence of data has excluded the possibility to perform more sophisticated research methods, such as the panel data analysis. Furthermore, the adoption of cross-country analyses could favour greater understanding of the non-financial reporting dynamics at a country level.