This paper aims to deepen knowledge of the principle of connectivity between financial and non-financial information defined in the Integrated Reporting (IR) conceptual framework. We have created a multidimensional connectivity index that is used to assess the level of connectivity in reporting. The results indicate that the levels achieved are low, suggesting that firms publishing their reports under the IR conceptual framework have not fully internalised this principle. Our results also show that coercive and normative institutional factors such as the mandatory nature of IR or the legal system, drive connectivity. Connectivity is also generally influenced by internal managerial decisions mainly related to the report's length and assurance. These findings may have practical implications for regulators and standard-setters, who should explicitly incorporate this principle into their pronouncements, providing guidelines to facilitate its practical implementation, and for preparers who should enhance the interactivity of reports.
The complexity and rapid pace of change in today's environment, characterised by financial, governance, ecological and health crises, have increased stakeholders' pressures on companies to disclose information about economic, social, or environmental issues affecting sustainability (Busco et al., 2019). Thus, there is a need to foster a culture of transparency and accountability (Tejedo-Romero et al., 2023). In response to these demands, there has been a significant rise in non-financial reporting accompanied by the proliferation of frameworks aimed at standardising the information to be disclosed (Dumay et al., 2016). Among them, in recent years, the Integrated Reporting (IR) disclosure initiative (Songini et al., 2023) has stood out, launched in 2013 by the International Integrated Reporting Council (IIRC), whose conceptual framework was revised in 2021 (IIRC, 2021). The conceptual framework of IR not only describes what content should be included in reporting, but also suggests the application of a set of principles to ensure the quality and transparency of information (Permatasari & Tjahjadi, 2023). Among these principles is the connectivity principle, which establishes the need to link the content of corporate information in order to provide stakeholders with a comprehensive and integrated picture of corporate performance (Caglio et al., 2020). The connectivity of information enables understanding of the interdependencies between financial and non-financial performance, hindering the assessment of future sustainability (Barth et al., 2017). According to Robles-Elorza et al. (2023), sustainability is an integrated issue that requires a balance between financial and non-financial performance, so in sustainability reports, both contents must be interconnected.
From an internal management perspective, the application of the connectivity principle promotes integrated thinking that provides a holistic view of the factors affecting the organisation's ability to create value over time (Eccles et al., 2015; Songini et al., 2023). Furthermore, the connectivity principle can also play a significant role in reducing information asymmetries in markets (Barth et al., 2017; Brennan & Merkl-Davies, 2018; Busco et al., 2019; Caglio et al., 2020; Dimes & de Villiers, 2024; Zhou et al., 2017), at a time when capital providers demand more information about companies' commitments to ESG (Economic, Social, and Governance) criteria. Promoting connectivity among contents enhances the quality of information provided to investors (Dimes & De Villiers, 2024). Disconnected information creates information asymmetries, making it difficult for investors to have a holistic view of organisations' performance and their contribution to sustainability. On the other hand, the more connected the contents of a report, the greater its comprehensibility, coherence, and conciseness (Busco et al., 2019). All these reasons justify the need to advance in the application of the principle of connectivity. Additionally, this may be an opportune moment to address the development of this principle, given the current significant regulatory and standard-setting momentum of international bodies, such as the International Financial Reporting Standard (IFRS) and the European Financial Reporting Advisory Group (EFRAG), which have already included information connectivity in their agendas (Giner et al., 2023).
However, to date, few studies have focused on examining the principle of connectivity (La Torre et al., 2019), and those that have not addressed it comprehensively. Specifically, some studies have analysed the interdependencies between the different capitals as defined in IR, including natural, social and relationship, human, intellectual, manufacturing, and financial capitals (e.g., Grassmann et al., 2019; Pistoni et al., 2018). However, no previous research has adopted a multidimensional approach to connectivity that encompasses all the key forms of connectivity as described in the IR conceptual framework (IIRC, 2021).
This paper aims to deepen understanding of the connectivity principle through the study of its practical implementation, answering the call for research on the practical application of IR principles and specifically on the connectivity principle (Grassmann et al., 2019; La Torre et al., 2019; Pigatto et al., 2023; Rinaldi et al., 2018; Simnett & Huggins, 2015).
We have investigated the existence of institutional factors that explain the levels of connectivity achieved. Although the effects of institutional isomorphism on the decision to adopt IR have already been analysed (e.g., Frías-Aceituno et al., 2013; Jensen & Berg, 2012; Kiliç et al., 2021), its influence on the application of IR principles is yet to be explored. The study of institutional factors is relevant because they elucidate the determinants of sustainability reporting practices, providing insights for regulators and standard setters to consider when shaping their policies and the content of their proposals.
To explore the practical application of the connectivity principle, a content analysis of the sustainability reports of 387 companies from 16 countries included in the IR database was conducted. In addition, an ordinal logistic regression model was used to examine the institutional factors determining the connectivity level achieved. The results show that the level of connectivity in IR is low and that, in applying the connectivity principle, companies respond more to coercive institutional forces (i.e., the mandatory nature of IR) or normative forces (i.e., the legal system). Additionally, it is generally influenced by internal management decisions, mainly those related to the length and assurance of the report.
The remainder of the paper is structured as follows. Section two reviews and discusses the literature on the principle of connectivity. Section three presents the potential explanatory variables of connectivity and formulates the hypotheses for the study. The fourth section describes the methodology employed. Section five presents the results, and section six discusses them. Finally, section seven contains the main conclusions and practical implications of the study.
2Conceptual frameworkStakeholder pressure for companies to demonstrate their sustainability and to become more transparent has led to mandatory non-financial reporting in many countries. Markets, capital providers, information preparers, standard setters, regulators and consumers, among others, have expressed the limitations of financial statements in reflecting the true value of companies (Tejedo-Romero et al., 2023) and have demanded that companies provide information on their long-term sustainability. In an attempt to ensure that the information disclosed is comparable and credible, we have witnessed a veritable regulatory tsunami and publication of conceptual frameworks and pronouncements in the sustainability reporting landscape for just over a decade. During this time, two major reporting initiatives have emerged: the Global Reporting Initiative (GRI) and the Integrated Reporting (IR) framework. Disclosure of this type of information was significantly boosted in 2009 with the publication of the King Report on Corporate Governance (King III) in South Africa, which requires all companies listed on the Johannesburg Stock Exchange to publish an integrated sustainability report. Later, in 2014, Directive 2014/95/EU was published, establishing its mandatory nature for the largest companies in the European Union. After some initial years of confusion and dispersion in the field of non-financial reporting, regulatory bodies and accounting standards setters have recently taken the first steps towards consolidating non-financial reporting frameworks. Thus, in June 2021, the leading promoter of Integrated Reporting, the IIRC, merged with the Sustainability Accounting Standards Board (SASB) to form the Value Reporting Foundation (VRF). Several months later, in November 2021, the IFRS Foundation (International Financial Reporting Standards) announced at COP26 the launch of the International Sustainability Standards Board (ISSB) to establish a global baseline for sustainability reporting. The ISSB aims to develop standards that result in comprehensive, high-quality global baseline sustainability disclosures focused on the needs of investors and financial markets. In March 2022, GRI and ISSB signed a collaboration agreement to coordinate their work programs. Additionally, the same year, the IFRS assumed responsibility for SASB standards and incorporated the VRF into its organization. Finally, in June 2023, the ISSB approved its first sustainability-related disclosure standards: IFRS S1 and IFRS S2.
In parallel, the European Commission, having recognised the need to improve the quality of standards, initiated a process to develop European Sustainability Reporting Standards (ESRS). This culminated in the approval of EU Directive 2022/2464 in 2022 and the publication of the standards proposed by EFRAG (European Financial Reporting Advisory Group) in December 2023. These standards will be mandatory for many European companies starting in 2024.
The ISSB, EFRAG and GRI are currently collaborating to develop a global proposal for sustainability reporting. However, each body maintains its own guidelines and prepares mapping resources to ensure the interoperability of disclosure requirements. This interoperability should not be confused with connectivity. It is an attempt to align disclosure requirements, reduce complexity, and avoid duplication of content rather than an improvement in information connectivity.
In their recent pronouncements, both EFRAG (2023) and the IFRS Foundation (2023) have emphasised the importance of interconnected financial and non-financial reporting ecosystems, aiming to serve the interests of capital markets, companies, and stakeholders in general. EFRAG (2023) states that presenting and developing non-financial sustainability information in close connection with financial information is vital for delivering a comprehensive, coherent, and integrated view of corporate reporting.
The main challenge for the practical application of the connectivity principle lies in the absence of a consensus definition, making it an abstract and vague concept (David & Giordano-Spring, 2022). To date, the literature that has studied the connectivity of information has done so from at least four angles: additive, textual, relational and holistic.
First, some authors propose an additive view of connectivity (Eccles & Krutz, 2010), which involves aggregating different content from various reports that are the basis for the preparation of an integrated report (financial statements, management commentary, and social and environmental, corporate governance, and intellectual capital reports). However, their view of connectivity as a mere summary of individual reports in a single document or links between different sections is insufficient for understanding the relationship between financial and non-financial performance.
Secondly, some authors have advanced beyond the previous approach and have examined connectivity by analysing textual semantics and its effects on language comprehension. For instance, Caglio et al. (2020) evaluate the readability of reports by assessing certain textual attributes, such as reading complexity, length and tone of discourse. Other authors have measured linguistic connectivity to gauge reporting comprehension (Brennan & Merkl-Davies, 2018; Grassmann et al., 2019). Connectivity has also been linked to tone in discourse construction and report narratives (Caglio et al., 2020; Melloni, 2015). Similarly, the role played by elements beyond language, such as the use of visual tools like photographs, in connectivity has been analysed (e.g., Eccles et al., 2015; Nicolò et al., 2021). Although these elements can facilitate the understanding of information, they alone do not guarantee a comprehensive and holistic view of the organisation (Barth et al., 2017).
A third group of research has taken a relational approach to studying connectivity, analysing the links between different social media and stakeholder communication platforms. This approach ensures that the various communication channels with stakeholders are consistent and interactive (Brennan & Merkl-Davies, 2018; Lodhia & Stone, 2017; Rivera-Arrubla et al., 2017; Masiero et al., 2020). However, this type of relational connectivity often exceeds the boundaries of the integrated report itself and relates to accessing other sources of information through social media and other digital tools.
Finally, IR advocates for a holistic approach to connectivity that promotes the existence of integrated thinking, which breaks down information silos and facilitates dialogue between functional units (Dimes & De Villiers, 2024; Stubbs & Higgins, 2014; Songini et al., 2023). This shift in mindset should provide external stakeholders with higher-quality reporting that represents the trade-offs between various contents and adopts a long-term approach to sustainable value creation.
The IR framework differs from other approaches to understanding connectivity (such as additive, textual, or relational) by emphasising the effective exploration of connections between different pieces of information. The IR approach is focused not only on what to connect (e.g., Busco et al., 2019; Zhou et al., 2017) but also on how to connect the information presented in the report to address the increased complexity of the content (WICI, 2013). Connectivity is facilitated when explicit links between the different information elements prevail in the report, i.e. when information is displayed as dynamic and interrelated rather than static and linear or as a mere sequence of overlapping and fragmented textual contents. These interdependencies are important as they provide a more comprehensive view of firms’ performance, help identify critical business elements (Giorgino et al., 2020), and assist in developing a consistent message (WICI, 2013).
This way of understanding connectivity has been synthesised by IR, which identifies the key forms of connectivity in reporting (IIRC, 2021):
- 1.
Connectivity between the contents, so as to provide a total picture that reflects the dynamic and systemic interactions of an organisation's activities as a whole.
- 2.
Connections between the past, present, and future.
- 3.
Interdependencies and trade-offs between capitals, including changes in their availability, quality and affordability that affect an organisation's ability to create value.
- 4.
Connections between financial and other information.
- 5.
Connectivity between quantitative and qualitative information. For that, it proposes to include key performance indicators as part of the narrative explanation.
- 6.
Connectivity and consistency between external accountability information and information for internal business management and the board.
- 7.
Linkages between the information in the integrated report and other communications and information from external sources.
This description seems to make it clear that there is no single path to achieve connectivity. However, measuring these seven key forms of connectivity in isolation is challenging since some of them are not mutually exclusive and can coexist simultaneously. For instance, financial information often tends to be both quantitative and backward-looking. Therefore, to facilitate the analysis, we suggest grouping all these forms of connectivity into three dimensions.
The first refers to the interdependencies between the different types of capital that a firm uses and generates (key form number 3 in the IIRC's (2021) definition of connectivity previously indicated). The measurement of connectivity between capitals has been studied by Arul et al. (2021), Grassmann et al. (2019), Giorgino et al. (2020)and Grassmann et al. (2022), primarily through textual analysis. In our study, the analysis of this connectivity dimension implies that capitals are not described in isolation in the report, but rather, their interrelationships, trade-offs and impacts are shown, often using capital flow diagrams, infographics or resource maps. In addition, the capitals are connected to other essential contents of the report (e.g., strategies, action plans, material issues, or risk assessment) (Eccles et al., 2015; Pigatto et al., 2023).
The second dimension seeks to balance content of different nature, e.g., historical information with future prospects (Adams, 2015; Kiliç & Kuzey, 2018; Melloni, 2015; Menicucci, 2018). This type of connectivity is achieved when a firm's performance is accompanied by forward-looking forecasts and analysis, for example, citing past performance alongside future targets and plans. Furthermore, effective linkage requires combining financial and non-financial information, and qualitative and quantitative pieces of information in a way that is understandable and easy to interpret (IIRC, 2013; Melloni, 2015; Stubbs & Higgins, 2014). This may be possible, for example, by exposing trends in one's own performance over time with linkages to important changes in strategy or by comparing performance with that of other companies (WICI, 2013). This dimension aligns with key forms of connectivity in the IIRC's (2021) definition of connectivity numbered 2, 4, and 5.
Finally, our third dimension relates to connecting the information in the integrated report to other communications and information sources using technology, allowing users to customise report queries (IIRC, 2013; La Torre et al., 2019; WICI, 2013), for example, through links to external information or QR codes. This dimension corresponds to key forms of connectivity in the IIRC's (2021), numbered 6 and 7.
These three dimensions of connectivity (see definitions and examples in Table 1) provide the total picture referred to in number 1 of the IIRC's (2021) definition.
Dimensions of connectivity
However, to date, no studies have simultaneously analysed all these forms of connectivity. Previous research has typically focused on linguistic connections and the coherence of report narratives (e.g., Brennan & Merkl-Davies, 2018) or on the interdependencies between capitals (e.g., Grassmann et al., 2019; Giorgino et al., 2020; Grassmann et al., 2022). Fig. 1 illustrates how integrated thinking promoted by IR can help achieve a more holistic view of information in reporting, thanks to the application of the connectivity principle.
The levels of connectivity achieved by companies can be used to identify institutional drivers for the implementation of the principle. Previous literature has already focused on assessing the influence of contextual factors on the dissemination of non-financial information (e.g., Frías-Aceituno et al. 2013; Gallego-Álvarez & Quina-Custodio, 2016; García-Sánchez, Cuadrado-Ballesteros, & Frías-Aceituno, 2016; Lai et al. 2016). However, these factors have not yet been explicitly linked to the connectivity principle.
Institutional theory (DiMaggio & Powell, 1983) explains that organisations adopt sustainability reporting practices through a process of “isomorphism”, which is necessary for responding to pressures arising from their operational context. Sustainability reporting practices may be generated by a mimetic process derived from observing what is happening in their environment. They can also be influenced by the presence of norms and cultural values within the context in which companies operate. However, coercive isomorphism, involving the imposition of regulatory requirements, appears to be the most effective mechanism for promoting these practices. Mandating disclosure could have a significant impact in the coming years, given the growing trend of regulating the disclosure of non-financial information (Aguado-Correa et al., 2023).
3Hypotheses developmentInstitutional theory posits that organisations disclose information in response to environmental pressures and align their behaviour with generally accepted norms and values (Jensen & Berg, 2012; García-Sánchez et al., 2013). It seems reasonable that reporting is influenced by the institutional context in which firms operate (García-Sánchez & Noguera-Gámez, 2018).
From this point of view, for companies operating in countries where sustainability reporting is mandatory, so-called coercive isomorphism drives disclosure (Stubbs & Higgins, 2014). In these contexts, regulations seek to improve reporting quality by standardising the contents to be disclosed and promoting the application of sustainability reporting principles. Therefore, it would be logical to assume that companies produce more connected information in countries where sustainability reporting is mandatory. However, the literature has shown that entities do not always respond to pressures arising from mandatory requirements (Kiliç et al., 2021). To further explore the effects of regulation on the application of the connectivity principle, we have formulated the following hypothesis:
H1 Information connectivity is higher in companies where sustainability reporting is mandatory.
Some countries mandate the preparation of sustainability reports using the IR framework as a guide. However, the use of the guidelines is controversial, as they can be used to justify that the organisation is acting correctly and thus to legitimise its performance and gain reputational benefits (Lai et al., 2016). Although connectivity is one of the principles of the IR conceptual framework, so far, there is no evidence to demonstrate that mandatory IR is a driver of connectivity. It would make sense to assume that when IR is mandatory, there is a greater understanding and development of integrated thinking, which would promote information connectivity (Arul et al., 2021). However, Arul et al. (2021) show that in contexts where IR is voluntary, there is a heightened concern for establishing connections between capitals, raising questions about the effect of mandatory IR on connectivity. This led us to our second hypothesis:
H2 Information connectivity is higher in companies for which the integrated reporting framework is mandatory.
So-called normative isomorphism should mean that, when companies act under the norms and cultural values of the same context, their reporting practices should be similar. This context may be influenced by the administrative tradition of the country (Jensen & Berg, 2012). In administrative code law (i.e., stakeholder-oriented), a multiplicity of interests prevails, which typically entails more non-financial reporting and transparency, while in the common law context (i.e., shareholder-oriented), reporting is more focused on financial aspects and investor protection (Frías-Aceituno et al., 2013; Gallego-Álvarez & Quina-Custodio, 2016; García-Sánchez, Cuadrado-Ballesteros, & Frías-Aceituno, 2016; Kiliç et al., 2021). On the one hand, Jensen and Berg (2012) demonstrate that IR firms are more likely to originate from countries with higher investor protection (i.e. common law). This is because, in market-based economies, investors make decisions based on self-analysed information, which necessitates the disclosure of not only financial information but also other relevant non-financial aspects (Prado-Lorenzo et al., 2009). Consequently, there may be greater incentives for connectivity. In contrast, Frías-Aceituno et al. (2013) have found evidence suggesting that companies in code law countries are more interested in using IR. This is because the standardisation of information introduced by the guidelines facilitates decision-making for a broader range of stakeholders, rather than just serving the interests of investors. Companies would, therefore, be more inclined to promote connectivity. Specifically, Tirado-Valencia et al. (2021) find that the levels of integrated thinking are higher in code law countries, suggesting that companies located in countries with this administrative tradition would show higher levels of connectivity in their reporting. The lack of clear evidence regarding the influence of legal context on connectivity led us to test the following hypothesis:
H3 The level of information connectivity is associated with the legal origin of the country.
In addition, so-called mimetic isomorphism could mean that companies that operate in settings with similar environmental impacts will adopt similar standards and behaviours, such as integrated reporting (Kiliç et al., 2021). Firms operating in sensitive industries face greater pressure from their stakeholders to report the impacts they generate, especially in recent years, as issues related to climate change have intensified (Gallego-Álvarez & Quina-Custodio, 2016; Lai et al., 2016). Specifically, Busco et al. (2019) demonstrate that firms operating in more environmentally sensitive contexts provide more holistic and interconnected information, as they need to link their environmental impacts to financial performance and risks. Moreover, in the most sensitive industries, information asymmetries are greater (Martínez-Ferrero et al., 2016), thus making it imperative to pursue greater connectivity to mitigate them. All of this led to our fourth hypothesis:
H4 Information connectivity is higher in companies belonging to environmentally sensitive industries.
The sample selection process started with all of the firms included in the IIRC database (534 firms). Lai et al. (2016) demonstrate that organisations in this database are characterised by the quality of the information dissemination and its accessibility to a wide range of stakeholders (Melloni, 2015). The information in this database extends until 2020, precisely the year when the integration process of the IIRC into the Value Reporting Foundation (VRF) commenced. Therefore, we searched for reports from 2019 on these companies' websites, given that the data for 2020 is also atypical due to the significant influence of the COVID-19 pandemic. We excluded 95 firms whose reports were not available for this year and an additional four because, despite being in the IIRC database, the reports did not follow the IR framework. Lastly, we excluded reports from companies located in countries with fewer than five observations (48 companies). This decision was made because our study primarily focuses on institutional theory, and a limited number of observations in a country constrains our ability to draw conclusions regarding the isomorphism effect. Consequently, our final sample consists of 387 integrated reports for the year 2019.
Table 2 presents the distribution of the samples by country in Panel A and by sector according to the Global Industry Classification Standard (GICS) in Panel B.
Sample distribution
The dependent variable is the index of information connectivity (Total_connectivity). This index is calculated by summing the values from the three dimensions described above: 1) the interdependencies between capitals (Capital_interdependencies), 2) the connectivity between different types of information (Content_connections), and 3) the interactivity of the report (Interactivity). All dimensions have been weighted equally in the calculation of the total index, to avoid potential subjectivity that weighted measures might introduce (e.g., Pistoni et al., 2018; Pigatto et al., 2023; Tirado-Valencia et al., 2021).
For the coding of connectivity levels in each of the dimensions, a scoring system was developed based on the creation of a scale, a method commonly used in the field of sustainability reporting (Pistoni et al., 2018). The scale is assigned a value of 0 when there are no connections among the elements mentioned in each dimension, and the content is presented in an isolated and overly narrative manner. A score of 1 is given when there are some connections among the aspects within the dimension, but they are limited. A score of 2 is applied to the dimension when connections are frequent, and elements that enhance the understanding of interrelationships are present (e.g., tables, charts, graphs, diagrams). Finally, the scale is assigned a value of 3 when connections are very frequent and abundant elements facilitate the understanding of interdependencies, complemented by other visual aids (e.g., icons, pictograms, color codes).
The total connectivity index ranges between 0 and 9, calculated by summing the scores obtained in the three dimensions, as each dimension (i.e. Capital_interdependencies, Content_connections, and Interactivity) can take values between 0 and 3.
As for the independent variables, they are all related to the institutional explanatory factors of connectivity identified in the development of the hypotheses: Sustainability reporting mandated (SR Mandated), IR Mandated, Legal System and Sensitive Industry.
In addition, this study incorporates several control variables to examine the characteristics of reports that influence the application of the connectivity principle. First, two variables dependent on internal managerial decisions have been included. The first of these is the conciseness of the report. A very lengthy report can be used as a form of obfuscation (Arul et al., 2021) and can be difficult to interpret (Caglio et al., 2020; Melloni et al., 2017; Pistoni et al., 2018). However, it is possible to produce concise and useful reports when content connectivity is fostered (Barth et al., 2017; Grassmann et al., 2019). However, according to Rivera-Arrubla and Zorio-Grima (2016), the opposite relationship would also be possible, i.e. connectivity would be more evident in lengthy reports since companies consider that when the report is longer, it is more necessary to connect the contents to improve its comprehension.
Assurance is another variable that has often been associated with reporting quality (Boiral et al., 2019; Caglio et al., 2020; Maroun, 2019; Simnett & Huggins, 2015). Farooq et al. (2024) suggest that assurance providers should analyse the links between financial and non-financial data and information, highlighting the importance of ensuring connectivity in reports. In particular, Grassmann et al. (2022) show that there is a relationship between assurance and the level of connectivity between capitals. However, no study to date has examined the effect of assurance on other dimensions of connectivity.
Furthermore, this study incorporates some control variables related to firm incentives for disclosing voluntary information to mitigate information asymmetry issues, such as size, leverage and profitability. Size is often associated with greater pressure to address information asymmetry (e.g., Busco et al., 2019; Frías-Aceituno et al., 2013; Frías-Aceituno et al., 2014; García-Sánchez et al., 2013; García-Sánchez & Noguera-Gámez, 2018; Kiliç et al., 2021; Martínez-Ferrero et al., 2016; Prado-Lorenzo et al., 2009). Larger companies are more inclined to disclose more information because they face higher agency costs (García-Sánchez & Noguera-Gámez, 2018) and have more resources to enhance communication. Therefore, it would be reasonable to expect higher levels of integration and connectivity (Busco et al., 2019).
The association between reporting and financial performance has been a long-debated issue in literature (e.g., Frías-Aceituno et al., 2013; Frías-Aceituno et al., 2014; García-Sánchez & Noguera-Gámez, 2018; Lai et al., 2016). When a company's financial situation deteriorates, there is a greater need for reporting efforts to reduce information asymmetries and mitigate the advantages of better-informed investors (García-Sánchez & Noguera-Gámez, 2018; Martínez-Ferrero et al., 2016). In such cases, it becomes more critical to establish connections between financial and non-financial performance, and a forward-looking orientation can demonstrate that a poor financial position does not jeopardise the long-term sustainability of the company (Jensen & Berg, 2012).
However, the literature also presents an alternative interpretation, as when the situation worsens, the company may reduce its integration of financial and non-financial information in an effort to maintain legitimacy (Frías-Aceituno et al., 2014; García-Sánchez et al., 2013; Prado-Lorenzo et al., 2009).
The list of variables, the criteria for their quantification, and their relationship to the hypotheses are summarised in Table 3.
Definition of independent variables
| Hypothesis | Variable | Definition |
|---|---|---|
| H1 | SR Mandated | Binary variable, 1= Sustainability reporting mandatory, 0= Not mandatory (source: Carrots & Sticks ESG & Sustainability Policy worldwide database) |
| H2 | IR Mandated | Binary variable, 1= IR framework mandatory, 0= Not mandatory (source: Carrots & Sticks ESG & Sustainability Policy worldwide database) |
| H3 | Legal System | Binary variable, 1= Code law, 0= Common law (source: University of Ottawa (http://www.juriglobe.ca/eng/sys-juri/index-alpha.php)) |
| H4 | Sensitive Industry | Binary variable, 1= Sensitive, 0= Not Sensitive (source: self-defined in analogy to Busco et al. (2019)) |
| Control variables | ||
| Firm-level | Length | Natural logarithm of the number of pages of the integrated report, excluding the pages containing the mandatory financial statements (if included in the integrated report) |
| Assurance | Binary variable, 1= Assured, 0= Not assured | |
| Size | The logarithm of total revenue | |
| Leverage | Total liabilities/total equity | |
| Return on Assets (ROA) | Net income/total assets |
We analysed the report content of the 387 companies in the sample to assess the levels of connectivity in the three defined dimensions. While acknowledging the subjective nature of content analysis measurements, this methodology is one of the most widely used for the analysis of non-financial reports (Dumay et al., 2016).
To ensure the reliability and internal consistency of the results, three researchers scored the level of connectivity in each of the dimensions using a subsample of six reports. The assigned scores were subsequently reconciled among the three researchers to prevent discrepancies (Krippendorff, 2013).
After consensus on the scores, we analysed the contents of the reports in the sample. This approach allowed us to first perform a descriptive analysis of the information to identify the frequencies of each dimension as well as calculate the total connectivity index. This analysis aims to determine whether preparers are producing their reports based on the previously defined dimensions of connectivity.
In addition, in this study we carried out a correlation analysis to evaluate the association between the variables incorporated into the model and used in the subsequent multivariate analysis.
Finally, we tested the hypotheses by performing a multivariate analysis using ordinal logistic regression. This choice was based on the ordered nature of the total connectivity index, and the Brant test's Chi-square score indicates no violation of the proportional odds assumption for ordinal logistic regression.
5Results5.1Descriptive statistics and correlationsBefore analysing the results, the validity of the scores was initially confirmed, assessing their consistency with the external ratings produced by EY South Africa (EY, 2020) for its 2020 Excellence Integrated Reporting Awards1. One of the aspects evaluated within this rating includes the connectivity of the reports. Since EY only gathers data from South African firms, we utilised a subsample of companies domiciled in South Africa (a total of 128 firms). This external rating has been employed in prior studies to demonstrate the consistency of their reporting quality metrics (Malola & Maroun, 2019; Zhou et al., 2017). Pearson Chi-square statistics confirm a statistically significant association (p<0.05) between our connectivity scores (total connectivity and for each dimension) and the external rating developed for the EY awards.
Regarding the descriptive statistics, the results show that the connectivity levels of the integrated reports of companies in our sample are low (see Table 4). Panel A of Table 4 shows that the average total connectivity index score is only 3.62 out of a maximum total score of 9.
Dependent variables: descriptive statistics and frequency (percentage)
As for the results by dimension, the highest average is found in content connectivity (1.42 out of a maximum of 3), as reports typically combine both qualitative and quantitative information. The average connectivity levels are even lower in the case of interdependencies between capitals (1.25 out of a maximum of 3) and information interactivity (0.95 of a maximum of 3).
The frequency of the scores in Panel A of Table 4 indicates that 37 % of the companies present a value of 0 for the interdependencies between capitals dimension and 36 % for interactivity. For these two dimensions, more than 60 % of the companies accumulate their scores in the lowest values (0-1). On the other hand, only 24 % of the analysed companies obtain the highest score in the dimension of interdependencies between capitals, 15 % in that of connections between contents and 8 % in that of interactivity. A total of 32 companies score zero on the total connectivity index (8.27 %) (see Panel B of Table 4). However, only 11 companies (2.84 %) achieve the highest possible score in all three dimensions. These results reveal that most of the companies in our sample lack interest in connectivity or do not understand the application of this principle.
The analysis of the correlation matrix included in Table 5 provides the first statistical approximation of the associations between the variables. The different dimensions of connectivity present significant positive correlations between them, although the values are not very high, except for capital interdependence and content connection (0.581; p<0.01).
Correlation matrix
| Variables | (1) | (2) | (3) | (4) | (5) | (6) | (7) | (8) | (9) | (10) | (11) | (12) | (13) |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| (1) Total connectivity | 1.000 | ||||||||||||
| (2) Capital interdependencies | 0.839* | 1.000 | |||||||||||
| (3) Content connections | 0.826* | 0.581* | 1.000 | ||||||||||
| (4) Interactivity | 0.680* | 0.305* | 0.376* | 1.000 | |||||||||
| (5) SR mandated | -0.062 | -0.125 | -0.024 | 0.025 | 1.000 | ||||||||
| (6) IR mandated | 0.167* | 0.247* | 0.091 | 0.024 | 0.258* | 1.000 | |||||||
| (7) Legal system | -0.106 | -0.203* | -0.016 | 0.003 | 0.310* | -0.593* | 1.000 | ||||||
| (8) Sensitive industry | 0.126 | 0.062 | 0.122 | 0.124 | 0.028 | 0.066 | -0.037 | 1.000 | |||||
| (9) Length | 0.271* | 0.224* | 0.226* | 0.185* | -0.046 | 0.058 | -0.126 | 0.075 | 1.000 | ||||
| (10) Assurance | 0.206* | 0.187* | 0.160* | 0.132* | -0.064 | -0.086 | 0.086 | 0.189* | 0.287* | 1.000 | |||
| (11) Size | -0.101 | -0.206* | -0.043 | 0.045 | 0.111 | -0.402* | 0.466* | 0.094 | 0.122 | 0.069 | 1.000 | ||
| (12) Leverage | -0.080 | -0.048 | -0.070 | -0.075 | 0.012 | 0.010 | 0.021 | -0.096 | 0.008 | -0.008 | 0.014 | 1.000 | |
| (13) ROA | 0.054 | 0.054 | 0.074 | -0.004 | -0.019 | 0.033 | 0.045 | 0.026 | -0.015 | -0.073 | 0.133* | 0.019 | 1.000 |
As shown in Table 5, pairwise Spearman correlations reveal statistically significant associations between the dependent variables and some of the explanatory variables. The results indicate that the mandatory use of the IR conceptual framework, as well as the length and assurance of the report, have a significant positive association with the level of the total connectivity index and that of some of the dimensions.
The correlation matrix also reveals some significant correlations among the independent variables, but none approach the critical threshold of 0.7, indicating the absence of multicollinearity issues (Lehmann et al., 1998).
5.2Results of the multivariate analysisThe results of the ordinal logistic regression performed to examine the influence of the explanatory factors on the connectivity levels are presented in Table 6. Based on institutional theory, we find evidence suggesting that the coercive and normative pressures exert a greater influence on the level of connectivity than the mimetic aspect. Surprisingly, the influence of the mandatory SR variable turns out to be contrary to what was expected in H1. In countries where sustainability reporting is mandatory, reports show lower levels of connectivity (Odds ratio (OR) 0.521; p<0.05).
Regression results
| Variables | Model (1) Total connectivity | |
|---|---|---|
| Coeff. | OR | |
| SR Mandated | -0.652⁎⁎ | 0.521⁎⁎ |
| (0.326) | (0.170) | |
| IR Mandated | 0.891⁎⁎⁎ | 2.437⁎⁎⁎ |
| (0.285) | (0.694) | |
| Legal System | 0.552⁎⁎ | 1.737⁎⁎ |
| (0.269) | (0.467) | |
| Sensitive Industry | 0.266 | 1.305 |
| (0.255) | (0.332) | |
| Length | 0.946⁎⁎⁎ | 2.575⁎⁎⁎ |
| (0.201) | (0.518) | |
| Assurance | 0.582⁎⁎⁎ | 1.790⁎⁎⁎ |
| (0.210) | (0.375) | |
| Size | -0.103* | 0.903* |
| (0.053) | (0.048) | |
| Leverage | -0.011 | 0.989 |
| (0.010) | (0.010) | |
| ROA | 0.042⁎⁎⁎ | 1.042⁎⁎⁎ |
| (0.007) | (0.008) | |
| Observations | 387 | |
| Pseudo R2 | 0.039 | |
| Log Lik | -817.7 | |
Robust standard errors in parentheses
However, in line with H2, the results from Table 6 demonstrate a positive relationship between the IR mandated variable and the index of total connectivity (OR 2.437; p<0.01), underscoring the significance of including this principle explicitly in the IR framework for report connectivity.
Regarding the institutional factors related to normative isomorphism, the results confirm our H3 and suggest that companies located in code law countries are more likely to report higher levels of total connectivity (OR 1.737; p<0.05).
As for the institutional factor related to mimetic isomorphism, the results show that the environmental sensitivity of the sector in which the company operates does not seem to be a crucial factor in determining the level of connectivity. The results do not provide evidence to support H4.
With regard to the variables related to internal managerial decisions, report length shows a strong positive association with the connectivity level (OR 2.575; p<0.01). There is also evidence of the effect of assurance on the application of the connectivity principle. Firms that have their reports assured are more likely to exhibit higher levels of connectivity (OR 1.790; p<0.01).
Finally, the findings reveal a significant positive association between ROA (Return on Assets) and the connectivity level (OR 1.042; p<0.01). We also identified a weak negative relationship between firm size and the connectivity level (OR 0.903; p<0.1). This may be because larger companies are more complex, making it more challenging to establish interdependencies among their content. Consequently, greater connectivity may not be exclusive to the largest organisations in the sample.
5.3Robustness analysisAdditional robustness checks are conducted to ensure the reliability of the results and mitigate potential biases, particularly those associated with industry classifications and the mandatory nature of the Integrated Reporting (IR) conceptual framework in the country (see Table 7).
Robustness analysis
| Model (2) Excluding South Africa | Model (3) Non-financial firms | |||
|---|---|---|---|---|
| Total connectivity | Total connectivity | |||
| Variables | Coeff. | OR | Coeff. | OR |
| SR Mandated | -0.656* | 0.519* | -0.488 | 0.614 |
| (0.342) | (0.177) | (0.371) | (0.228) | |
| IR Mandated | 0.931⁎⁎⁎ | 2.537⁎⁎⁎ | ||
| (0.315) | (0.799) | |||
| Legal System | 0.638⁎⁎ | 1.893⁎⁎ | 0.678⁎⁎ | 1.971⁎⁎ |
| (0.294) | (0.556) | (0.286) | (0.563) | |
| Sensitive Industry | 0.281 | 1.324 | 0.222 | 1.248 |
| (0.310) | (0.411) | (0.266) | (0.332) | |
| Length | 0.953⁎⁎⁎ | 2.593⁎⁎⁎ | 1.018⁎⁎⁎ | 2.767⁎⁎⁎ |
| (0.218) | (0.567) | (0.236) | (0.653) | |
| Assurance | 0.487⁎⁎ | 1.627⁎⁎ | 0.488⁎⁎ | 1.629⁎⁎ |
| (0.241) | (0.392) | (0.241) | (0.392) | |
| Size | -0.157⁎⁎ | 0.854⁎⁎ | -0.097 | 0.907 |
| (0.072) | (0.0611) | (0.067) | (0.061) | |
| Leverage | -0.005 | 0.995 | 0.004 | 1.004 |
| (0.004) | (0.004) | (0.024) | (0.024) | |
| ROA | 0.047⁎⁎⁎ | 1.048⁎⁎⁎ | 0.039⁎⁎⁎ | 1.039⁎⁎⁎ |
| (0.009) | (0.010) | (0.008) | (0.009) | |
| Observations | 259 | 308 | ||
| Pseudo R2 | 0.040 | 0.039 | ||
| Log Lik | -528.4 | -651.1 | ||
Robust standard errors in parentheses
First, the ordinal logistic regression model is re-estimated, excluding South African firms (128 in total) from the sample, to address potential biases stemming from the mandatory nature of the Integrated Reporting (IR) framework in this country. Given that the principle of connectivity is part of this conceptual framework, it is reasonable to assume that these companies are more dedicated to this principle (Grassmann et al., 2022). Thus, the subsample used consists of 259 firms from countries other than South Africa, where IR is entirely voluntary.
The results of this robustness test are provided in Model 2 of Table 7. The subsample analysis for companies that adopt the IR framework voluntarily confirms that our results are robust, as they demonstrate that, consistent with the results of the main regression (see Table 6), in the absence of coercive pressures due to the mandatory nature of the IR conceptual framework, the legal system, length, assurance, and ROA are significantly and positively associated with the connectivity level.
Second, the analysis is repeated for non-financial firms (308 in total) to account for potential bias resulting from considering financial and non-financial companies jointly, as they exhibit differences in their information disclosure regulation (Grassmann et al., 2022). Furthermore, the financial sector reports more integrated reports than other industries, which could make them more inclined to apply the connectivity principle (Vitolla et al., 2020). Accordingly, the baseline regression model is run only for non-financial firms. The results in Table 7 for this model (Model 3) align with those obtained in the main regression reported in Table 6, as they reveal that the connectivity level is positively associated with IR mandated, the legal system, length, assurance and ROA.
6DiscussionThe aim of this study is to deepen the understanding of the principle of connectivity and analyse its implementation, expanding the research on the principles of the IR conceptual framework (e.g., Ruiz-Lozano & Tirado-Valencia, 2016). A multidimensional model has been proposed to capture the diversity of elements that constitute connectivity, allowing us to discern which aspects of connectivity have been considered by companies and which could be enhanced.
Our results provide evidence that companies disclosing their information under the IR conceptual framework have not fully embraced the principle of connectivity. This finding aligns with the conclusions of Grassmann et al. (2019), who also observed a low level of connectivity in reports. This result can be attributed to the fact that many companies still maintain fragmented approaches in report preparation, emphasising siloed thinking rather than integrated thinking (Tirado-Valencia et al., 2021).
The reports analysed in our study exhibit characteristics more in line with what Eccles et al. (2015) refer to as combined reports than integrated reports. They also align with what Busco et al. (2019) described as conservative or minimalist levels of integration, characterised by routine and imitation but with limited commitment to transparency and accountability. This approach could hinder stakeholders' understanding of financial and non-financial performance (Robles-Elorza et al., 2023).
Despite the emphasis placed by the IR conceptual framework on the importance of demonstrating interdependencies between capitals and describing the interrelationships between factors influencing a firm's ability to create value, companies continue to report on their capitals in a disconnected manner (Giner et al., 2023). Our results reinforce the conclusions of Arul et al. (2021), Grassmann et al. (2019), Giorgino et al. (2020) and Grassmann et al. (2022) regarding the need for stronger links between financial and non-financial capitals, allowing stakeholders to understand the interdependencies between them for assessing the sustainability of companies.
Regarding the connections between quantitative and qualitative aspects, behaviour patterns are quite diverse, but generally involve a combination of both types of content. However, it is challenging to comprehend how certain aspects influence others. Qualitative contents are 'added' to quantitative ones but are not integrated, as no information is provided about their interdependencies. Furthermore, we observed that, regarding the connection between the past and the future, reports often lack a forward-looking perspective, in line with the findings of Menicucci (2018) and Tirado-Valencia et al. (2021).
The low levels of connectivity in the interactivity dimension suggest that companies are very close to what Masiero et al. (2020) call monolithic connectivity, which is textual in nature but lacks interaction between the company and the recipients of the information. Companies are missing out on the opportunities offered by new communication technologies (Cea Moure, 2019).
In addition, our paper extends previous studies on the determinants of IR (e.g., Frías-Aceituno et al., 2013; Frías-Aceituno et al., 2014; García-Sánchez et al., 2013; García-Sánchez & Noguera-Gámez, 2018; Jensen & Berg, 2012; Kiliç et al., 2021) and finds some evidence that coercive pressures, such as mandatory enforcement of the IR framework, and normative forces such as the legal system, may drive connectivity.
With respect to the impact of mandatory sustainability reporting, the result is contrary to what was expected. Some authors have already suggested that coercive forces can have contradictory effects (Kiliç et al., 2021). This could be explained by the fact that in current regulations, connectivity is not included as a reporting principle, so companies are not highly motivated to implement it. Moreover, in cases where reporting is not obligatory, a substitution effect becomes evident as firms voluntarily adopt the principle. Conversely, the mandatory nature of Integrated Reporting (IR) significantly influences connectivity since this principle is explicitly part of its conceptual framework, making companies better understand the importance of its implementation.
The findings show that, in line with previous literature, the legal system, as a normative pressure, influences the level of reporting connectivity (e.g., Frías-Aceituno et al., 2013; Gallego-Álvarez & Quina-Custodio, 2016; García-Sánchez, Cuadrado-Ballesteros, & Frías-Aceituno, 2016; Kiliç et al., 2021). We found evidence that companies domiciled in stakeholder-oriented countries (i.e., code law) are more committed to connectivity. This suggests that they are striving to enhance the quality of information directed at a multitude of stakeholders and not just responding to the interests of investors. This could be explained by the fact that in this cultural context, companies exhibit a higher degree of integrated thinking (Tirado-Valencia et al., 2021).
As for mimetic forces, we cannot confirm the previous findings by Busco et al. (2019), which suggested that companies with higher environmental pollution levels disclose more integrated information.
Beyond institutional isomorphism factors, we have also observed how other factors dependent on preparers' preferences and internal managerial decisions, such as report length and assurance, drive the principle of connectivity. Regarding the report's length, Barth et al. (2017), Caglio et al. (2020) and Grassmann et al. (2019) have highlighted the positive effects of conciseness on content integration. Our results contradict these findings and are consistent with Rivera-Arrubla and Zorio-Grima (2016): greater connectivity efforts are made in longer reports. It does not seem that connectivity is being used to shorten the length of reports but to improve their comprehension when a large quantity of information is disclosed.
Our study also provides new evidence regarding the role of assurance in reporting, demonstrating its positive effect on content integration (Caglio et al. 2020; Grassmann et al., 2022). The results corroborate the positive effects of assurance on the application of non-financial reporting principles, not only on the content of the information verified, in line with the suggestions of Farooq et al. (2024).
Finally, we found a positive association between a firm's financial performance (i.e., ROA) and connectivity, consistent with the findings of Grassmann et al. (2019). Although this relationship between ROA and connectivity is somewhat insignificant, it could indicate that more profitable companies are employing connectivity as a tool to reduce information asymmetry and agency costs. This result could also suggest that when companies are more profitable, they are more concerned with demonstrating the links between financial performance and social performance along the lines suggested by Robles-Elorza et al. (2023). On the other hand, contrary to previous literature (e.g., Grassmann et al., 2019), size and leverage do not emerge as significant drivers of connectivity. Therefore, our results regarding the potential effect of connectivity on reducing agency problems are only partially conclusive.
7ConclusionThis study aims to deepen understanding of the connectivity principle embedded in the IR conceptual framework. A better understanding of the principle of connectivity can help management to have a more integrated view of performance, overcoming siloed thinking. In addition, when companies adhere to this principle, report recipients can link financial and non-financial performance (IIRC, 2021) and have a more comprehensive view of a company's sustainability.
Conversely, the absence of connectivity could negatively impact stakeholder's ability to assess the value creation process as a whole. Financial reporting primarily emphasises short-term results, but this information alone may be limited in identifying long-term sustainability issues. Hence, it is important to establish linkages with non-financial information, which often offers a forward-looking perspective with a longer time horizon. Connectivity ensures that financial information is related to sustainability disclosures, reducing information asymmetries and mitigating potential future risks. Our findings confirm that companies have yet to embrace these interdependencies fully, and there is still a long way to go in implementing the connectivity principle. The linkages should go beyond document-to-document interconnectedness and must effectively integrate financial and non-financial issues, encompassing the different dimensions of connectivity.
This study has theoretical implications as we have attempted to conceptualize connectivity by defining three dimensions that help better understand the scope of this principle. As suggested by IR, connectivity is a construct that goes beyond links between report contents and encompasses interdependencies between capitals, links between information of different nature (quantitative, qualitative, past, future, financial, non-financial), and the interactivity of the report.
The results of this research also have practical implications. Firstly for preparers, who should enhance the interactivity of reports to avoid documents with overly long narratives and no apparent linkages. We recommend they capitalise on the opportunities presented by information and communication technologies to improve this interactivity. In addition, we have provided evidence that external assurance of reports positively affects connectivity and we suggest that preparers consider having their reports externally assured.
Likewise, our results have implications for regulators. The low connectivity scores achieved indicate that companies have few incentives to enhance connectivity. We recommend that regulators emphasise the need to make these connections explicit and provide practical guidance or illustrative examples for achieving connectivity, as our findings suggest the importance of explicitly incorporating this principle into their pronouncements. It is necessary to go beyond legislation that merely checks boxes and does not regulate the application of reporting principles. The inclusion of principles could help mitigate the drawbacks of the "one-size-fits-all" approach (Busco et al., 2019). In other words, rather than specifying detailed content, regulations should establish fundamental guiding principles that offer flexibility in reporting while also providing a solid foundation for high-quality reporting. These recommendations may be particularly necessary in the European context, especially given the timing of EU Directive 2022/2464 on the disclosure of sustainability information, which has reinforced the mandatory nature of reporting and is developing principles for the practical application of EFRAG's ESRS guidelines. Furthermore, our findings support the idea that EU companies should be required to assure their reports to enhance the quality of reporting. It would be desirable for EU regulators to develop a standard for assurance on sustainability reports that promotes greater connectivity.
We would therefore recommend that regulators and standard-setters explicitly incorporate the principle of connectivity in their proposals, similar to what has been done with other principles, such as materiality, and suggest practical guidelines for operationalisation. Specifically, in the European context, the recent ESRS standards issued in July 2023 by EFRAG should address the connectivity between disclosure requirements and promote the application of this principle.
However, our results should be considered in light of the following limitations. The fact that all firms in the sample are part of the IIRC database could affect the results as these companies have proven to produce high quality IR reports. Therefore, the levels of connectivity achieved in non-financial reports following other frameworks could be even lower than the ones obtained in this study, which reinforces the need for further progress in the implementation of this principle. There is also some subjectivity in measuring the connectivity level, as identifying the factors that determine connectivity are not clearly defined and they are difficult to assess in the reports. This challenge arises because connectivity remains an abstract concept, despite our attempts to systematise it. Additionally, we cannot be sure that connectivity improves stakeholders' understanding of the reports, as they were not consulted in our research. Therefore, future research should aim to gather their perceptions of the improvement in the quality of information resulting from the application of the connectivity principle.
CRediT authorship contribution statementPilar Tirado-Valencia: Writing – review & editing, Writing – original draft, Supervision, Methodology, Investigation, Formal analysis, Conceptualization. Marta de Vicente-Lama: Writing – review & editing, Writing – original draft, Supervision, Methodology, Investigation, Formal analysis, Conceptualization. Magdalena Cordobés-Madueño: Writing – review & editing, Writing – original draft, Supervision, Methodology, Investigation, Formal analysis, Conceptualization. Mercedes Ruiz-Lozano: Writing – review & editing, Writing – original draft, Supervision, Methodology, Investigation, Formal analysis, Conceptualization.
