Institutional isomorphism under the test of Non-financial Reporting Directive. Evidence from Italy and Spain

Abstract

Purpose

This paper aims to analyse the impact of the European Union (EU) Directive on the quality of sustainability reporting under the institutional theory lens. Specifically, the authors evaluate what kind of institutional pressure has the highest impact on the quality of corporate disclosure on sustainability issues.

Design/methodology/approach

The authors build a quality index based on the content analysis of sustainability information disclosed, before and after the transposition of the Directive, by Italian and Spanish companies belonging to different industries. The authors use an OLS regression model to analyse the effect of coercive, normative and mimetic forces on the quality of the sustainability reports.

Findings

The results highlight that normative and mimetic mechanisms positively affect the quality of sustainability reporting, whereas there is no evidence regarding coercive mechanisms, indicating that the new requirements do not provide a significant contribution to the development of better reporting practices, at least in the two analysed countries.

Originality/value

To the best of the authors’ knowledge, this is one of the few studies assessing the quality of sustainability reporting through an analysis involving the period before and after the implementation of the EU Directive. It enriches the literature on institutional theory by analysing how the different dimensions of isomorphism affect the quality of information disclosed by companies according to the EU requirements. It contributes to a better understanding of the impact of the non-financial information Directive, and the results of this paper can be relevant for regulators, practitioners and academia, especially in view of the adoption of the new Corporate Sustainability Reporting Directive proposal.

Citation

Posadas, S.C., Ruiz-Blanco, S., Fernandez-Feijoo, B. and Tarquinio, L. (2023), "Institutional isomorphism under the test of Non-financial Reporting Directive. Evidence from Italy and Spain", Meditari Accountancy Research, Vol. 31 No. 7, pp. 26-48. https://doi.org/10.1108/MEDAR-02-2022-1606

1. Introduction

Over the past few years, the European Union (EU) has actively promoted firm transparency concerning sustainability issues. Disclosure of social and environmental information supports the measurement, monitoring and management of corporate non-financial performance and contributes to identifying sustainability risks and increasing stakeholders’ trust [European Commission (EC), 2014]. The adoption of the European Directive 2014/95/EU represents a decisive step change in disclosing sustainability information. This EU Directive mandates each country to regulate the obligation of public interest entities to issue a new statement including, as a minimum, information related to “environmental, social and employee matters, respect for human rights, anti-corruption and bribery matters” [European Commission (EC), 2014, par. 1]. The adoption of the new regulation enforced the transition from voluntary to mandatory disclosure of non-financial information (NFI), traditionally reported in standalone non-financial reports (e.g. corporate social responsibility reports, environmental reports, social reports and others). Thereby, a radical change in European reporting practices was promoted, fuelling the debate on the impacts of mandatory sustainability information requirements (Dumitru et al., 2017; Doni et al., 2019; Mion and Loza Adaui, 2019; Caputo et al., 2020). Prior literature has widely debated the changes produced by mandatory disclosure on the quality of information with controversial results. Some studies stress the role of regulation in improving the quality of sustainability reports (Deegan, 2002; Venturelli et al., 2017; Mion and Loza Adaui, 2019). Others argue that regulation alone may not suffice to guarantee higher reporting levels (Bebbington et al., 2012; Luque-Vilchez and Larrinaga, 2016; La Torre et al., 2018).

This paper investigates the impact of the EU Directive on the quality of sustainability reports under the institutional theory lens. Indeed, institutional isomorphic processes significantly influence reporting practices adopted by companies in response to mandatory requirements (Dumitru et al., 2017; Carungu et al., 2020). To this end, we perform a comparative analysis of sustainability information [1] disclosed within different periods (i.e. pre and post-NFI Directive) by companies belonging to different industries in two European countries i.e. Italy and Spain. These two countries have deep historical-social connections and similar legal and economic systems. They are located in Southern and Latin Europe and have many elements in common related to the stakeholder orientation [both countries have a civil law legal system, and according to Ball et al. (2000) and Simnett et al. (2009), civil law countries are more stakeholder oriented] socio-cultural and political aspects (de Nardis, 2015; Tarquinio et al., 2018; Paolotti et al., 2019; Patuelli and Carungu, 2020). Our study adopts the institutional theory framework grounded on the assumption that organizations are affected by coercive, normative and mimetic forces resulting from the institutional context where they operate. These forces create and diffuse a common set of rules, values and beliefs institutionalized within organizations, eventually producing homogenization across corporate practices (Meyer and Rowan, 1977; DiMaggio and Powell, 1983).

Using a methodology adopted in the literature dealing with sustainability reporting (Romero et al., 2019; Michelon et al., 2015; Beretta and Bozzolan, 2004), we create a quality index based on the content analysis of sustainability information disclosed by Italian (years 2013, 2016 and 2017) and Spanish listed companies (years 2013, 2017 and 2018). We consider different years given the different timing of the EU Directive transposition into Italian and Spanish law (Posadas and Tarquinio, 2021). Our results highlight that normative and mimetic mechanisms positively affect the quality of sustainability information. On the contrary, we find no evidence regarding the effect of coercive mechanisms.

The remainder of the paper is organized as follows. Section 2 provides the theoretical and empirical framework. Section 3 describes the research method. Sections 4 and 5 present and discuss our findings. Section 6 provides the conclusions of our analysis.

2. Theoretical and empirical framework

The institutional theory builds on the assumption that organizations are strongly influenced by the institutional context in which they operate. The institutional environment defines a common set of rules, values and beliefs that become institutionalized within organizations, leading to the adoption of homogeneous forms of behaviour (Meyer and Rowan, 1977; DiMaggio and Powell, 1983). Companies tend to align their policies and practices with those of other companies facing the same environmental conditions, thereby giving rise to the so-called phenomenon of “institutional isomorphism” (DiMaggio and Powell, 1983). This process is generally driven by the need to abide by stakeholders’ expectations and reach greater legitimacy (Scott, 2005). Legitimacy is the degree to which companies gain collective approval in society, representing an important means of stability and survival (Rahman et al., 2019; Latif et al., 2020). Companies can gain or ensure legitimacy due to three institutional pressures related to coercive, normative and mimetic isomorphic forces. Coercive pressure arises from the regulatory legal system and societal expectations; normative pressure stems from professional norms and values comprising a specific field; mimetic pressure is associated with the imitation of successful or leading companies in response to uncertainty (DiMaggio and Powell, 1983; Lindblom, 1993; Deegan, 2014; Ali et al., 2021).

Numerous studies show how different isomorphic mechanisms have shaped sustainability reporting. Shabana et al. (2017) suggest that three institutionalization stages characterize the adoption of corporate social responsibility (CSR) reporting: defensive reporting, driven by coercive isomorphism; proactive reporting, stemming from normative forces; and imitative diffusion, associated with mimetic behaviour. Some authors argue that the three isomorphic pressures act together on sustainability disclosure practices (Moseñe et al., 2013). Other authors sustain that some forms of isomorphism can exert a stronger influence than others. For instance, coercive isomorphism has often been identified as the most impactful factor guiding sustainability reporting (Pedersen et al., 2013; Dumitru et al., 2017; Latif et al., 2020). García-Sánchez et al. (2016), in an international study of CSR reporting, show that normative and coercive forces are the main drivers of corporate transparency. The influence exerted by normative and mimetic pressures is especially significant in disseminating environmental information by the world’s largest companies (Cubilla-Montilla et al., 2020). Moreover, Tiron-Tudor et al. (2019) find that mimetic pressure considerably affects levels of sustainability information reported by Romanian companies, whereas coercive isomorphism has no effect.

Therefore, we acknowledge that the study of changes in sustainability reporting quality during the transition from voluntary to mandatory disclosure must consider the role that these three forms of isomorphism play throughout this process.

2.1 Coercive isomorphism

Coercive pressure has a relevant impact on companies’ approach to sustainability reporting (Dumitru et al., 2017; Latif et al., 2020). Governments are considered one of the most influential stakeholders affecting corporate reporting practices due to their capacity to enforce regulations and impose sanctions (Rahman et al., 2019). Corporate response to governmental pressure results in compliance with legal requirements (La Torre et al., 2018). Governments can play an important role in a context where sustainability issues have become a very sensitive subject by introducing regulations to reflect societal expectations on corporate transparency concerning these matters (Depoers and Jérôme, 2020). In this sense, the adoption of the EU Directive is considered a coercive mechanism that, by making sustainability information mandatory, affects companies’ approach to sustainability reporting (Tiron-Tudor et al., 2019; Aureli et al., 2020; Carungu et al., 2020).

After analysing the effects of mandatory requirements on sustainability disclosure, prior literature has drawn different conclusions. A number of these studies have examined reporting practices adopted before the Directive to understand how companies reacted to the new requirements. Other papers document the good quality and completeness of the non-financial reports (Carini et al., 2018; Manes-Rossi et al., 2018). Yet others identify an important gap between voluntarily disclosed sustainability information and what the Directive requires (Matuszak and Różańska, 2017; Venturelli et al., 2019; Doni et al., 2019). At any rate, all the cases highlight the need for further improvement in disclosure quality and the potential function held by the Directive. The new regulation is assessed as a milestone able to improve corporate information systems by enhancing the accurateness, comprehensiveness and credibility of sustainability information (Venturelli et al., 2017; Doni et al., 2019; Caputo et al., 2020). It can also harmonize non-financial disclosure to ultimately lead to greater comparability (Dumitru et al., 2017; Matuszak and Różańska, 2017; Carini et al., 2018). However, Venturelli et al. (2019) argue that the Directive can represent an opportunity for improvement only in companies with limited experience in the field of sustainability reporting. Several drawbacks have been attributed to the mandatory requirements. Moreover, regulation could discourage reporting proactivity (Tarquinio et al., 2020) and induce companies to symbolically adopt responsible practices (Dumitru et al., 2017; Doni et al., 2019; Caputo et al., 2020).

2.2 Normative isomorphism

As highlighted before, coercive pressure prompts the harmonization of sustainability reporting. However, pressure from the national context could create divergences across the EU member states in disclosing such information (Aureli et al., 2020). National culture, norms and values surrounding the business environment are associated with normative forces that can produce different orientations towards sustainability disclosure (Martínez-Ferrero and García-Sánchez, 2017; Cubilla-Montilla et al., 2020). Normative mechanisms establish the role and responsibilities that companies are expected to assume in society by defining a social and moral code of conduct (Scott, 2013). These cultural rules are taken for granted within companies and embedded in their decision-making processes and strategies (Perrault Crawford and Clark Williams, 2010).

Normative pressure induces companies to go beyond legal requirements to shape the quantity and quality of sustainability information disclosed (Moseñe et al., 2013; Cubilla-Montilla et al., 2020). Indeed, it is assumed that information disclosed in response to moral obligations strives to be reliable and material (Perrault Crawford and Clark Williams, 2010).

Literature has extensively identified different country-specific factors as key determinants of sustainability disclosure. Legal systems and law enforcement mechanisms applied in a country represent important indicators of stakeholder orientation. Stakeholder-oriented countries are deemed to be more concerned with sustainability issues and thus tend to report these issues (Frias-Aceituno et al., 2014; Ioannou and Serafeim, 2017). Similar results are found with reference to the level of cultural development. Companies operating in culturally developed countries present greater levels of sustainability commitment, CSR disclosure and assurance demand (García-Sánchez et al., 2016; Martínez-Ferrero and García-Sánchez, 2017). The presence of a strong CSR tradition and the resulting levels of CSR expertise also explain national differences in adopting sustainability information reporting practices (Mion and Loza Adaui, 2019; Venturelli et al., 2019; Aureli et al., 2020; Cubilla-Montilla et al., 2020). Furthermore, countries with vast experience in sustainability reporting are expected to be more subject to normative isomorphism. Dumitru et al. (2017) observe a positive influence of expertise on sustainability information quality. Likewise, the enactment of pre-existing national regulations is a critical factor in the company’s response to the new EU requirements on NFI (Aureli et al., 2020). Specifically, previous legislation has been associated with better disclosure quality (Venturelli et al., 2019).

As highlighted in several studies (de Nardis, 2015; Tarquinio et al., 2018; Paolotti et al., 2019; Patuelli and Carungu, 2020), Italy and Spain have the same legal system (positive law), they belong to the same territorial context (Mediterranean countries) and, even if there are some variations among them, they have a similar cultural orientation (according to Hofstede’s national culture dimensions). Moreover, they are both characterized by a CSR reporting rate higher than the global average (KPMG, 2015, 2017, 2020). Despite these considerations, there are differences that may arise from key elements underlying the implementation process of the Directive.

2.3 Mimetic isomorphism

The introduction of the EU Directive has generated some uncertainty among companies about how to put the new requirements into practice (Szabó and Sørensen, 2015). In response, companies generally align their practices with those of successful companies, thus, engaging in imitative behaviours. Uncertainty is seen as the most powerful force for mimetic isomorphism (DiMaggio and Powell, 1983), especially in the case of first-time reporters (Pedersen et al., 2013). Another element that produces imitation is the level of company exposure to the public, that is, their visibility and vulnerability to the environment in which they operate. Higher levels of business exposure can cause legitimacy problems motivating companies to interact better with their stakeholders and enhance their reporting practices following leading companies in the field (Shabana et al., 2017). Moreover, Latif et al. (2020) suggested that companies respond to mimetic pressure to reach a superior performance and thus gain a competitive advantage in the market.

Literature documents that companies are typically affected by mimetic pressure resulting from the industry to which they belong (Moseñe et al., 2013; Fernandez-Feijoo et al., 2014; Depoers and Jérôme, 2020). The industry is taken as a reference group because companies with similar nature and structure share comparable objectives, address similar uncertainties and institutional pressures, engage with similar stakeholders and are committed to the same regulation (Depoers and Jérôme, 2020; Cubilla-Montilla et al., 2020). In this sense, research shows that the company sector is an important determinant of sustainability disclosure performance (Dumitru et al., 2017; Caputo et al., 2020; Raucci and Tarquinio, 2020). Several authors argue that companies belonging to environmentally sensitive industries – namely, industries with a major impact on the environment – are more likely to disclose information on their sustainability performance. It is documented that they also face stronger pressure from stakeholders (Patten, 2002; Cho and Patten, 2007; Dumitru et al., 2017; Sierra-Garcia et al., 2018; Venturelli et al., 2019; Carungu et al., 2020).

3. Methodology

3.1 Research question

Given the role of the three isomorphisms forces in driving NFI reporting, and the open debate on their level of influence, this study aims to answer the following research question:

RQ1.

Which form of institutional isomorphism has the highest impact on the quality of sustainability information disclosed in non-financial reports?

To shed light on this question, we analyse sustainability information from different periods (i.e. pre and post-NFI Directive); from two countries with similar culture and stakeholder orientation in which sustainability reporting has developed in different ways, and from companies belonging to different industries. Italy and Spain have often been compared in the literature due to their similar culture (Patuelli and Carungu, 2020; Paolotti et al., 2019; Tarquinio et al., 2018). The interest in the comparison between Italy and Spain has arisen because, despite their similarities, the two selected countries were characterized by different backgrounds relating to sustainability reporting. Consequently, this comparison offers an interesting scenario to study the role of these characteristics in the impact produced by the Directive and the three isomorphic pressures.

Studies on the impact of legislative changes in Italy are more documented than those concerning the Spanish case. Italian companies reported a reduced number of sustainability indicators reported over the years (Loprevite et al., 2020; Raucci and Tarquinio, 2020; Raucci et al., 2020; Tarquinio et al., 2020). Most of the studies compare pre and post-NFI Directive in which the latter shows higher quality levels (Mion and Loza Adaui, 2019; Leopizzi et al., 2019; Caputo et al., 2020). On the contrary, other research found no significant impact of the EU requirements on sustainability information levels (Nicolò et al., 2020; Carungu et al., 2020). Studies conducted on the Spanish case have mainly explored sustainability information disclosed by companies before complying with the new requirements (Luque-Vilchez and Larrinaga, 2016; Manes-Rossi et al., 2018; García Sánchez et al., 2019). Analyses before and after the implementation of the Directive have assessed effects produced on the location of sustainability information (Sierra-Garcia et al., 2018; García-Benau et al., 2022) and on the impression management techniques applied by Spanish companies in their sustainability reports (García-Sánchez and Araújo-Bernardo, 2020).

The comparison between Italy and Spain might also provide interesting insights concerning the effects of normative pressure on the quality of sustainability reporting. As aforementioned, both countries present a similar culture and stakeholder orientation and a similar legal system. However, they are characterized by different levels of experience in the area of sustainability information reporting. Spain has had and continues to have a remarkable and positive evolution. Before the 2013 European Directive, Spain promoted sustainability reporting among large companies (e.g. the 2008 Consejo Estatal de la Responsabilidad Social Empresarial [CERSE] and the 2011 Spanish Sustainable Economic Law [SEL]). According to KPMG (2020), out of the 52 surveyed countries, Spain is in the top 10 countries, acknowledging climate change as a financial risk and connecting business activity to Sustainability Development Goals (SDG). However, Italian companies have shown little commitment to voluntarily disclose sustainability information (Costa and Agostini, 2016; Venturelli et al., 2017; Carini et al., 2018; Doni et al., 2019). For instance, before the EU regulation, only banking foundations, some social cooperatives and social enterprises were required to draw up social reports (Tarquinio et al., 2020). Nevertheless, Italian companies have significantly increased their sustainability reporting practices due to the impact of Legislative Decree 254/2016, transposing the European Directive (KPMG, 2020). Both countries are in a good disposition towards sustainability reporting, thereby offering us an interesting scenario for study.

Finally, Italy and Spain have a similar industry structure, where sectors such as manufacturing and tourism linked to culture and heritage are driving forces for the economy (Lazzeretti et al., 2012; Lazzeretti et al., 2016; Fondazione Symbola and Unioncamere, 2019). Countries with industry similarities are subject to similar pressures from stakeholders (Bilal and Hosli, 2000). Consequently, the analysis of the Italian and Spanish contexts provides adequate grounding to investigate how mimetic forces affect the disclosure of sustainability information.

3.2 Sample

We selected listed companies in the main stock markets of both countries for each of the years under analysis: the 40 companies belonging to the Italian FTSE MIB and the 35 companies from the Spanish IBEX35. These two indexes consist of the most liquid and highly capitalized stocks listed on the Italian and Spanish Stock Exchange. Consequently, the comparison was made among the most representative Italian and Spanish companies. To answer our research question, we need to identify three periods that would allow us to verify the possible regulatory effects produced on companies by Directive. Therefore, we chose the pre-European Directive and two periods, pre- and post-national regulation. We decided to analyse 2013 as the pre-Directive year for both countries (the Directive was adopted by the European Parliament and the Council of the European Union in 2014); considering that Spanish legislation was delayed due to internal political issues, we considered 2017 as the pre-national-legislation year for Spain and 2016 for Italy. Consequently, we chose 2017 for Italy and 2018 for Spain as post-national-legislation years (these years were, respectively, the first year in which for Italian and Spanish companies came in to force the new EU requirements).

3.3 Procedure

The dependent variable, the Quality index, is based on Romero et al.’s (2019), Michelon et al. (2015) and Beretta and Bozzolan (2004). We downloaded the companies’ reports related to sustainability. For 2013, we looked for sustainability information that could be within an integrated report (IR), in a standalone report, or as a part of the annual report (AR). We collected one document per company. In the case of multiple corporate reports drawn up, we analysed the sustainability report or the IR better than the AR. We followed the same procedure for the pre-national-legislation year (2016 for Italy and 2017 for Spain). For the post-national-legislation year, we first searched for the mandatory non-financial report. If it was missing, we followed the former procedure.

Once we had downloaded the reports, we used content analysis for their dissection using “RapidMiner” to separate the linguistic sentences and counted the concepts previously selected to calculate the quality index of the information disclosed. Appendix exhibits the list of core concepts in English, although the respective versions in Spanish and Italian were adopted for the processing. We completed Romero et al.’s (2019) concepts with others considering the European Directive. We then calculated the quality index as described in the variable section.

The statistical procedure began by assessing the sample adequacy. We continued with the descriptive and factor analysis to identify the principal components used as control variables in our proposed model. Finally, we ran a regression to obtain evidence to respond to our research questions.

3.4 Variables

3.4.1 Dependent variable.

The main dependent variable is the Quality index, a measure of the quality of information disclosed obtained from the content analysis of the sustainability information gathered. The quality index ranges from 0 (lowest quality) to 1 (highest quality). Our methodology considers not only the absolute number of core concepts included in the report but also their relative importance with respect to the whole disclosure. Following Romero et al. (2019), we define three components: relative quantity, density and accuracy.

Relative quantity presents the frequency of the concepts disclosed considering the industry and the size of the company. It is calculated as the standardized residuals of the regression used as a proxy for the disclosure, measured as total sentences, including core concepts. The model is: D^i=β0+∑j=1kβJINDJ+βk+1 LNSize, according to Beretta and Bozzolan (2004). Standardized residuals are calculated as the difference between Di-D^I, being D_i, the observed disclosure and D^i the estimated disclosure, both for company i. The variable is expressed in a continuum. Size is measured as the logarithm of total assets. The industry is classified into six groups (consumer goods; basic materials; oil and energy; consumer services; financial and real estate services; and technology and telecommunications).

Density informs on the level of compactness of the report. It is measured as the number of sentences in which a core concept is included and the total number of sentences in the report. Its value varies between 0 and 1. Higher values indicate higher density.

Accuracy indicates the level of precision, rewarding numerical and monetary sentences versus general literature sentences. Following Michelon et al. (2015), sentences with core concepts with monetary values weigh 3, numerical values weigh 2 and textual sentences weigh 1. Accuracy is calculated as the ratio between the total addition of the weighted sentences and the total number of sentences in the report. Its value ranges from 0 to 3. Higher values indicate higher accuracy.

To avoid subjectivity, we do not include the fourth Michelon et al.’s component, managerial orientation. The three components were standardized (OECD, 2008) to range from 0 (those observations with the minimum value) to 1 (those observations with the maximum value). Our quality measure for each observation is the arithmetic media of the outputs of the standardization process.

3.4.2 Independent variables.

To test our hypotheses, we use the following independent variables that measure the effect of three isomorphic mechanisms:

1. Coercive isomorphism, as highlighted before, refers to the legal regulatory system within which companies operate, and it is measured with a dichotomous variable. A code of 1 is given if the report corresponds to a post-national-legislation year; a code of 0 otherwise.

2. Normative isomorphism is related to the company’s culture, norms and values derived from the national environment. In accordance with prior literature, we consider that the enactment of pre-existing national regulations can be considered a relevant determinant of the company’s responses to the requirements on NFI. Therefore, the code of 1 indicates that the report belongs to a Spanish company, and the code of 0 indicates that it belongs to an Italian company.

3. Mimetic isomorphism explains why companies adopt referenced behaviours of other companies and industries modelling themselves to such behaviours (DiMaggio and Powell, 1983). In the field of sustainability reporting, literature documents that companies are typically affected by mimetic pressure resulting from the industry to which they belong (Moseñe et al., 2013; Depoers and Jérôme, 2020). The approach to sustainability reporting also varies if the company operates in a sensitive or non-sensitive sector from a socio-environmental point of view. Therefore, according to prior relevant research (Patten, 2002; Cho and Patten, 2007), we considered as environmentally sensitive industries: pharmaceutical, chemical, mining, metals, pulp and papers, transportation, petroleum and utilities. Consequently, we use a binary variable that takes the value of 1 if the company reporting belongs to an environmentally sensitive industry and 0 otherwise.

3.4.3 Control variables.

Previous literature has signalled an extensive set of variables that might be of interest as control variables in similar studies. Some of them are financial variables (Ortas et al., 2015); others are linked to board characteristics (Manning et al., 2019; Moses et al., 2020). We can also find references to specific characteristics of the companies (Hąbek and Wolniak, 2016), report features (Romero et al., 2019) or variables describing the firms’ social behaviour. Previous literature has signalled all these variables as useful for similar analytical procedures. To apprehend their possible contribution to the proposed model, we decided to consider them by means of a factorial analysis. This statistical tool allows us to identify the core dimension of their contribution without deteriorating the validity of our results.

3.5 Model

We tested our hypotheses by running a linear regression of ordinary least squares to explain the relationships between the dependent variables and the set of explanatory variables. The model is stated as follows:

Quality index=β0+β1×Normative isomorphism+β2×Coercive isomorphism+ β3×Mimetic  isomorphism+(∑βi×Factorial  analysis  components)+ε

4. Results

We first analysed the adequacy of the sample. To identify potential outliers, we calculated the typical score associated with each dependent variable value and removed four observations that had associated typical values over 3 (Aggarwal, 2013). Our final sample consists of 221 observations.

The descriptive for the explanatory variables (Table 1) shows that 33.5% of the sample corresponds to a post-national-legislation year (Coercive variable), 46.6% are reports drawn up by Spanish companies (Normative variable) and 38.9% are reports issued by companies belonging to environmentally sensitive industries (Mimetic variable).

Table 2 summarizes the descriptive for the continuous and categorical control variables. Specifically, we consider the total assets (Size), its indebtedness (Leverage) and its profitability (ROA). From the Morgan Stanley International ESG, we use the environmental and social indices (MSCI-E and MSCI-S, respectively). In terms of the board, we include the size (Board size), percentage of women (%WOB) and independent members (Board Ind.).

Table 3 provides the descriptive for the dichotomous control variables. These variables refer to the Dow Jones Sustainability Index (DJSI) company membership as well as a few reporting characteristics like the adoption of sustainability reports (SR), the use of the Global Reporting Initiative Guidelines or Standards (GRI), the presence of independent assurance (Assurance), the disclosure of information related to the SDGs and information on taxes paid (Taxes).

In line with our aim to incorporate all attributes that previous literature point to as contributors to the quality of sustainability reporting, we carried out a principal component analysis to reduce the number of variables and create the corresponding factors. We apply Varimax with Kaiser Normalization as the rotation method. The components are shown in Table 4. Kaiser–Meyer–Olkin Measure of Sampling Adequacy is close to 0.6, and the Bartlett’s test of sphericity has a p-value of 0.000, indicating the validity of the factor analysis. We renamed the components considering the variables included in them: Company profile (ROA, Leverage, Size and Board size), Responsible behaviour (Taxes and SDGs), Reputation (MSCI-E, DJSI and MSCI-S), Reporting (GRI, Assurance and SR) and Board quality (%WOB and Board Ind.).

Table 5 presents the results of our regression model with the explanatory variables representing the institutional mechanisms and the five components previously identified as control variables. Variance inflation factor (VIF) values for all independent and control variables are below 10, indicating the absence of multicollinearity problems (O’Brien, 2007). The quality index is mainly affected by normative mechanisms (Std coef. = 0.251 and p-value = 0.000) followed by mimetic (Std coef. = 0.166 and p-value = 0.005). We found no evidence regarding the effect of coercive mechanisms.

Companies showing higher values in Reputation and Reporting standardization present higher reporting quality. These results may be linked to mimetic isomorphism, as companies may be willing to stand out their piers by means of improving their reputation and communicating with their stakeholders. Board quality and Company profile present a significant and negative association with the quality index. Both factors might be related to normative isomorphism. Board quality is shaped by custom, tradition and norms, clearly affecting sustainability reporting. The “going beyond” legal requirements that normative isomorphism represents is condicionated by the company capacity to actually perform and disclose high quality sustainability information. We find no association regarding Responsible behaviour.

5. Discussion

Coercive isomorphism involves pressures put on an organization from other organizations on which it depends (DiMaggio and Powell, 1983). Several authors observed that regulations adopted by the European Union exercise coercive pressures over its Members. Indeed, EU regulations are perceived by companies as a corporate duty to comply with. Therefore, given its mandatory nature, the introduction of the EU Directive is considered a form of coercive mechanism that could affect companies’ approach to reporting (Tiron-Tudor et al., 2019; Aureli et al., 2020; Carungu et al., 2020). Our analysis shows that coercive mechanisms resulting from the introduction of the EU Directive make no impact on sustainability reporting quality. Indeed, we found no statistically significant association between reporting quality and the introduction of the EU Directive. Mandatory disclosure “sometimes appears as a kind of magical minimalism that delivers significant rewards at little cost” (Estlund, 2010). However, according to our results, mandatory disclosure of NFI does not imply better disclosure levels. Predictably, the EU requirements determine an increase in the number of sustainability reports produced by companies; however, a quantitative increase is not necessarily linked to an increase in the quality of information (Luque-Vilchez and Larrinaga, 2016; Dumitru et al., 2017; Andrades Peña and Larrán Jorge, 2019). Harmonization processes induced by regulation might reduce flexibility and penalize the disclosure of business-specific sustainability aspects (Allini and Manes-Rossi, 2007; Venturelli et al., 2017). Moreover, governmental imposition of reporting rules might result in a corporate tendency to limit information to mandatory requirements instead of developing better reporting practices. In this sense, sustainability disclosures are only a response to regulatory pressures rather than part of a process to enhance transparency (Caputo et al., 2020; Tarquinio et al., 2020). An in-depth analysis using country subsamples shows no association neither for Italy (p = 0.323) nor for Spain (p = 0.515). In line with this result, Perrault Crawford and Clark Williams (2010) and Venturelli et al. (2019) assert that mandatory reporting has a limited impact if applied to companies previously engaged in reporting sustainability issues. Our results are consistent with the rationale that regulation is insufficient to ensure a better quality of information (Chauvey et al., 2015; Costa and Agostini, 2016; Luque-Vilchez and Larrinaga, 2016; Dumitru et al., 2017; Carungu et al., 2020).

We also find that normative isomorphism is significantly and positively associated with reporting quality. Contextual factors, such as country- and company-specific factors, are also key determinants of sustainability information disclosure (Lucchese, 2020; Korca and Costa, 2021). We analyse normative mechanisms by considering the national context in the companies’ orientation towards sustainability reporting. Differences between Spanish and Italian settings may explain the effect of normative pressure on the quality of sustainability reporting. Specifically, these two normative contexts are characterized by different maturity and background in this kind of reporting. Spain is widely recognized for its commitment to sustainability disclosure with its worldwide leading position in terms of both quantity and quality of sustainability reporting. Indeed, as underlined in several studies, Spanish companies achieve high scores in various sustainability indexes (KPMG, 2015, 2017, 2020; Romero et al., 2019; Sierra-Garcia et al., 2018; Vaz Ogando et al., 2018; Mio et al., 2021). Before the transposition of the EU Directive, Spain was already committed to stakeholder accountability and corporate transparency and many large companies published sustainability reports. (CERSE, in 2008; SEL, in 2011; Mio et al., 2021). The Spanish strategy on corporate social responsibility practices 2014/2020 (Spanish Ministry of Employment and Social Security, 2014) represented another important initiative leading towards an authentic sustainability culture among Spanish companies.

On the contrary, before the EU Directive, sustainability reporting in Italy was mandatory only in certain types of not-for-profit organizations (Tarquinio et al., 2020) without providing specific tools to create adequate social dialogue. Previous research describes an important reporting gap among Italian companies, which are scarcely inclined to disclose social and environmental information voluntarily (Costa and Agostini, 2016; Venturelli et al., 2017; Carini et al., 2018; Doni et al., 2019). Consequently, Italy had less experience in this field and, with the adoption of the Non-financial Reporting Directive seize the opportunity to develop reporting on environmental, social and governance aspects (Jeffwitz and Gregor, 2017).

Literature observes that companies operating in environments with stronger traditions in sustainability reporting produce more quality disclosures. Specifically, reporting practices spurred by normative mechanisms akin to those of the Spanish context provide timely, relevant and reliable information that are more suitable to stakeholders’ needs (Perrault Crawford and Clark Williams, 2010). Our results confirm previous cross-country studies, according to which long experience in the sustainability reporting field influences reporting quality (Dumitru et al., 2017; Venturelli et al., 2019).

Another interesting outcome of our analysis shows that mimetic mechanisms also affect sustainability reporting quality. In fact, the significant and positive relationship between the Mimetic variable and the quality index indicates that companies operating in environmentally sensitive industries produce higher reporting quality. This result is consistent with other studies showing that these companies attract greater public visibility, experience more pressure and criticism from stakeholders and face higher levels of political risk (Fernandez-Feijoo et al., 2014; Tarquinio et al., 2018). Consequently, they tend to disclose better-quality information to contrast potential concerns regarding the effects of their activities, thereby gaining or repairing their legitimacy (Dyduch and Krasodomska, 2017; Manes-Rossi et al., 2018). Analysing the mimetic effect in detail among the two countries, we observe that it is significant in the subsample of Italian companies (std. coef. 0.283, p-value = 0.001) but not in the Spanish subsample, p-value = 0.363). This divergence may be due to the fact that mimetic mechanisms are especially useful in the initial stages of the institutionalization process in which Italy is immersed. On the contrary, the Spanish setting is more mature, and sustainability reporting broadly extends to different industries, not only among environmentally sensitive companies.

Results on control variables are broadly consistent with the existing literature. We find that reputation influences the quality of sustainability reports. Indeed, efficient reporting practices contribute to preserving company membership in these indexes, thereby enhancing their legitimacy and producing useful reputational effects (Bebbington et al., 2008; Bonsón and Bednárová, 2015). This result may also be explained by a “me-too” effect, according to which companies in the same industry compete to be listed in the same indexes and thus mitigate reputational risks (de la Cuesta and Valor, 2013).

Our study also reveals a positive association between the quality index and the Reporting construct. This result indicates that using the GRI, the external assurance of information and the disclosure of a sustainability report constitute a powerful tool for enhancing the quality level of sustainability reports. The presence of these factors generally denotes interest in demonstrating commitment and reinforcing stakeholder trust. Therefore, companies adopting these instruments are more concerned about the content and quality of information (Hąbek, 2017; Sethi et al., 2017; Orazalin and Mahmood, 2018).

Our findings show that Company profile construct has a significant and negative relationship with the quality of sustainability information. Previous studies find that variables such as leverage, profitability and size can be negatively associated with levels of disclosure (Purushothaman et al., 2000; Ho and Taylor, 2007; Muttakin and Khan, 2014; Romero et al., 2019). According to Purushothaman et al. (2000), highly leveraged companies are prone to establish closer relationships with creditors and communicate more directly with them, thus limiting reporting practices. Dyduch and Krasodomska (2017) argued that less profitable companies might be better committed to sustainability reporting because they aim to overshadow their lesser financial performance by showing their contribution to society. In line with Romero et al. (2019), our results suggest that the largest companies in the sample may use other mechanisms to legitimize their commitment to sustainability in addition to reporting.

We also find a negative relationship between the quality index and Board quality. This result suggests that the impact of women and independent members on reporting practices should also be examined in view of their background, competencies and the tasks they perform on the board. For example, Amorelli and García-Sánchez (2020) find that women with leadership positions adopt behaviours linked to male stereotypes showing less sensitivity towards social and environmental issues. Moreover, Michelon and Parbonetti (2012) argue that independent directors positively affect sustainability disclosure only if they are influential community members (i.e. retired politicians, academics, members of social organizations, etc.).

Finally, we find no relationship between the reporting quality and Responsible behaviour construct. We verify Responsible behaviour by focusing on two reporting practices that are gaining increasing attention in the field of sustainability reporting. However, this result can be due to the fact that tax and SDG disclosure still represent emerging practices, and their adoption remains limited (Hardeck and Kirn, 2016; De la Cuesta-González and Pardo, 2019; Rosati and Faria, 2019; Curtó-Pagès et al., 2021; Pizzi et al., 2021). In fact, only 27.1% and 25.3% of the reports in our sample include information on SDGs and taxation, respectively (Table 3).

6. Conclusions

European Directive 2014/95/EU represents one of the most significant regulatory developments in sustainability disclosure. This Directive has led to a new European scenario for corporate reporting in which the voluntary part of sustainability information becomes mandatory. This study aims to answer our RQ1.

The research focuses on two European countries, Italy and Spain, having similar culture and stakeholder orientation but a different development of sustainability reporting. We use institutional theory as the conceptual framework to examine how the institutional environment influences corporate disclosure practices. Specifically, we analyse the role of coercive, normative and mimetic isomorphic pressures in shaping sustainability reporting quality.

According to our results, normative and mimetic institutional mechanisms are the most effective way to improve the quality of sustainability information disclosed in non-financial reports. Similarly to Bebbington et al. (2012) and Luque-Vilchez and Larrinaga (2016), we find no evidence regarding the effect of coercive mechanisms on sustainability information quality. Although the legislative provision for the preparation of the sustainability report has resulted in an increase in the number of companies that are obliged to draw up the sustainability report, this increase in quantity of reports has not been accompanied by an improvement in the quality of the reporting. Specifically, Spain and Italy had already developed a high-level tradition of sustainability reporting, meaning that the NFI Directive made no additional contribution to the prior reporting situation in these countries. These results are in line with those of Caputo et al. (2020) and Mio et al. (2021). For instance, the Directive requirements are considered minimal and unsupported by specific rules on collecting, analysing and reporting data. Moreover, it cannot ensure comparability, completeness and clarity of sustainability information disclosed, even recognized by the regulator (EC, 2020). The achievement of the desired effects depends more on informal than on formal norms. Formal norms are those introduced by regulation. Informal norms are dictated by the generally accepted principles and practices that underpin the voluntary adoption of sustainability reporting and are socially recognized as legitimate (Bebbington et al., 2012). This highlights the key role of contextual factors in company response to regulatory requirements (Korca and Costa, 2021). Informal norms resulting from the environment in which companies operate not only should accompany changes in the law but also prevail over the law in shaping reporting practices (La Torre et al., 2018). In this vein, our analysis shows that normative and mimetic forces positively affect the quality of sustainability information. Consequently, the role of the national context and the company sector as major drivers of disclosure performance is confirmed. Our findings suggest that companies operating in environments with a strong tradition in sustainability reporting produce more quality disclosures. This is the case of Spain, where national regulators created institutions aiming to boost corporate communication on CSR issues and foster social dialogue. The creation of this normative climate has served to consolidate an authentic sustainability culture among Spanish companies. Therefore, at least in this country, the Directive has been a step back in CSR communication.

This research strengthens both the theoretical and empirical literature on sustainability information. This is one of the few studies (Mion and Loza Adaui, 2019) assessing the quality of sustainability information through an analysis involving the period before and after the implementation of the EU Directive by means of a three-year longitudinal and cross-country analysis. Therefore, it contributes to show that the compulsory provision of sustainability reports does not always generate improvements in terms of the quality of these reports. In addition, it enriches the literature on the adoption of the NFI Directive, focusing on two countries on which, till now, research has not been specifically focused.

Only a limited number of studies have investigated the role of institutional factors in corporate response to EU requirements. This study contributes to the literature on institutional theory by showing how the different dimensions of isomorphism affect reporting practices adopted by companies considering the EU requirements. Our findings showed that legislative change not necessarily supports the desiderated improvement of sustainability reporting quality. Therefore, the results of this paper might also have practical implications for regulators and policymakers. Understanding the determinants and pressures that influence reporting practices may help adopt new laws or revise existing regulation on sustainability disclosure, and this is relevant, especially considering the new Corporate Sustainability Reporting Directive (CSRD) (EU, 2021) that will expand significantly the mandatory adoption of sustainability reporting.

This paper also has some limitations. Firstly, our analysis is limited to the Italian and Spanish cases; thus, the impact of country-specific characteristics on sustainability reporting could be further researched by including other European countries in the analysis. Secondly, even though the major effects of a new regulation should be manifested after a number of years, our research only considers the first year of the EU Directive adoption. Therefore, it would be helpful to encompass a broader timeframe to better appreciate the evolutionary path of mandatory sustainability information reporting. As underlined before European Commission has recently issued a CCSRD proposal that would amend the existing requirements of Directive 2014/95/EU and extend its scope to all large companies and listed SMEs (EC, 2021). Consequently, future research could verify the effects of the Directive revision, assessing whether institutional factors impact differently on companies of these two countries.

Table 1.

Explanatory variables

Variable	Value	N	%
Coercive	Yes	74	33.5
	No	174	66.5
	Total	221	100
Normative	Yes	103	46.6
	No	118	53.4
	Total	221	100
Mimetic	Yes	86	38.9
	No	135	61.1
	Total	221	100

Table 2.

Continuous and categorical control variables

Variable	N	Min.	Max.	Average	SD
Size	221	12.680	21.100	16.974	1.725
Leverage	221	0.127	0.990	0.712	0.205
ROA	221	−12.950	26.450	4.542	6.033
MSCI-E	221	0.700	10.000	5.868	2.055
MSCI-S	221	0.000	9.500	5.003	1.439
Board size	221	1	24	12.57	3.619
%WOB	221	0.000	0.611	0.239	0.127
Board ind.	221	0.000	0.940	0.523	0.180

Table 3.

Dichotomous control variables

Variable	Value	N	%
DJSI	Yes	79	35.7
	No	142	64.3
	Total	221	100.0
SR	Yes	111	50.2
	No	110	49.8
	Total	221	100.0
GRI	0	50	22.6
	1	171	77.4
	Total	221	100.0
Assurance	Yes	29	13.1
	No	192	86.9
	Total	221	100.0
SDGs	Yes	60	27.1
	No	161	72.9
	Total	221	100.0
Taxes	Yes	56	25.3
	No	165	74.7
	Total	221	100.0

Table 4.

Factor analysis

Variable	Company profile	Responsible behaviour	Reputation	Reporting	Board quality
ROA	−0.837	−0.011	0.055	−0.118	0.143
Leverage	0.780	−0.051	−0.030	0.003	0.225
Size	0.762	−0.072	0.309	0.021	0.193
Board size	0.414	0.075	0.336	−0.308	0.072
Taxes	−0.031	0.930	−0.016	−0.001	−0.073
SDGs	−0.014	0.928	−0.055	0.015	−0.062
MSCI-E	0.111	0.058	0.723	0.040	0.053
DJSI	0.133	−0.063	0.641	0.386	0.039
MSCI-S	−0.055	−0.104	0.542	0.097	−0.248
GRI	0.046	−0.004	0.205	0.780	−0.014
Assurance	−0.047	0.101	0.188	0.624	0.300
SR	0.226	−0.304	−0.271	0.492	−0.473
%WOB	0.087	−0.149	−0.105	−0.014	0.715
Board ind.	0.152	−0.025	−0.022	0.180	0.667
KMO Measure of Sampling Adequacy			0.596
Bartlett’s test of sphericity		Apr.Chi-Sq.	658.960
		gl	91
		Sig.	0.000
Rotation converged in 7 iterations

Note: Italic values indicate the variables included in the principal components

Table 5.

Regression analysis

Variables	Std coef. Beta	Std Error	t	Sig.	VIF
Constant		0.006	62.192	0.000
Coercive	0.001	0.007	0.022	0.983	1.188
Normative	0.251	0.007	4.142	0.000	1.177
Mimetic	0.166	0.007	2.832	0.005	1.110
Company profile	−0.113	0.003	−1.977	0.049	1.050
Responsible behaviour	−0.028	0.003	−0.497	0.620	1.052
Reputation	0.217	0.003	3.641	0.000	1.146
Reporting	0.269	0.003	4.648	0.000	1.079
Board quality	−0.161	0.003	−2.691	0.008	1.148
Adj. R Square	0.315
Durbin-Watson	1.606
F (Sig.)	13.672 (0.000)
N	221

Note: Italic values indicate the statistically significant variables

Table A1.

Core concepts

Environment	Social
Environment	Employment	Discrimination	Local communities
Material	Labour	Freedom of association	Compliance
Energy	Labour relations	Collective bargaining	Customer health
Water	Labour management	Trade union rights	Customer safety
Biodiversity	Employee	Child labour	Labelling
Emission	Working conditions	Forced labour	Marketing Communication
Effluent	Employee health	Compulsory labour	Customer
Waste	Employee safety	Security	Customer privacy
Pollution	Training	Indigenous rights	Customer compliance
Climate	Education	Human rights	Supply chain
Transport	Diversity	Public policy	Social assessment
Environmental impact	Diversity policy	Corruption	Social risk
Environmental policies	Equal opportunity	Anti‐competitive behaviour
Environmental assessment	Equal remuneration	Society
Environmental risks	Labour practices	Social impact
Environmental initiatives	Employee grievance	Social dialogue

Note

1. The term “sustainability information” is used as an umbrella term that encompasses the different types of information analysed in our study (information reported within sustainability reports, integrated reports, annual reports and mandatory non-financial reports).