From words to action: SDG reporting insights in the Italian private sector

The resolution adopted by the United Nations General Assembly on September 25, 2015, ratified the adoption of the 2030 Agenda, a broad and universal policy aimed at pursuing global sustainable development. The 2030 Agenda consists of 17 sustainable development goals (SDGs) with 169 associated targets, which are interrelated and indivisible. The SDGs follow up on the millennium development goals achievements and represent common goals that affect all countries and individuals. World leaders have committed to joint action to pursue global growth through win-win cooperation, which aims to become more significant benefits for today’s and future generations (United Nations General Assembly, 2015). Indeed, the 2030 Agenda is considered the most intriguing, ambitious and comprehensive project undertaken in sustainable development and environmental policy (Biermann et al., 2017).

The SDGs came into force on January 1, 2016, although they are not legally binding. Each state was allowed to decide how to implement them within its territory, considering its national context’s peculiarities and respecting its policies and priorities (United Nations General Assembly, 2015). For their effective achievement, however, the need for holistic community involvement has also been highlighted, entailing the collaboration of governments, companies, organizations and citizens to increase the sustainable actions implemented (Bebbington and Unerman, 2018; Caprani, 2016; Pizzi et al., 2020). States have also been required to periodically monitor and assess their progress in pursuing the SDGs (Rickels et al., 2016). To this end, accounting plays a crucial role (Bebbington and Unerman, 2018). Indeed, it allows reporting, monitoring and making information available to all stakeholders for their empowerment and informed participation in decision-making (Brown and Dillard, 2015; García‐Sánchez et al., 2022; Nicolò et al., 2024a; Raimo et al., 2020). Accounting is also essential to establish corrective actions for the implemented processes to make them more efficient and effectively contribute to the planet’s and people’s prosperity (Martínez-Córdoba et al., 2020).

However, the absence of regulation and common standards for preparing SDG reporting makes the disclosure process complex. Although previous studies show that, especially on the part of private companies, voluntary information on the pursuit of the 2030 Agenda is beginning to be included in nonfinancial reports (Pizzi et al., 2020; Rosati and Faria, 2019), the relatively vague, purely qualitative nature of some SDGs leaves considerable margin for interpretation in assessing the efficiency of implemented actions and measuring their results (Biermann et al., 2017). Adopting a common reporting framework could help in these measurability challenges (Rickels et al., 2016) and the homogeneity of SDG reports (Aguado-Correa et al., 2023; Pizzi et al., 2021). Those open issues highlight how there is still plenty of room for improvements in SDGs reporting activity (Hummel and Szekely, 2022).

Disclosing SDGs can be considered a novelty in the academic literature (Aguado-Correa et al., 2023). Indeed, previous studies on SDGs have highlighted the limited number of studies on companies’ commitment to achieving the SDGs (Damiano and Di Maria, 2024; Heras-Saizarbitoria et al., 2022), on integrating the SDGs into corporate activities and reporting (Curtó-Pagès et al., 2021; Hummel and Szekely, 2022; Rosati and Faria, 2019; Tsalis et al., 2020) and on the role of accounting in helping companies contribute to the achievement of the SDGs (Bebbington and Unerman, 2018). The present study aims to fit into this stream of literature by analyzing SDG disclosure in a specific context: private Italian companies. The study uses natural language processing techniques to reveal the disclosure of SDGs in nonfinancial reports of 208 companies, which is a common practice in analyzing large amounts of textual data (Culpi Mann et al., 2024; Kang et al., 2020). This approach is novel for the proposed topic and might provide valuable insights that are not possible with the use of a more traditional methodology.

Findings prove firms’ increased interest in voluntarily disclosing SDG information in their nonfinancial reports, even with solid differences in quantity and type of information reported on each SDG. This should prompt policymakers and regulators to structure a unique reporting framework for helping companies monitor and disclose this data. Furthermore, more knowledge on sustainability and global goals should be spread in companies’ management to improve the achievement of those objectives.

After this introduction, Section 2 includes previous literature on SDG reporting, while Section 3 describes the context of the study and the applied methodology. Furthermore, Section 4 presents the analysis results, which are discussed in Section 5. Finally, Section 6 concludes the paper and provides suggestions for future research.

Although companies mainly pursue the aim of maximizing shareholder value (Martínez-Córdoba et al., 2020), in recent years, there has also been an increase in their interest in sustainability issues, particularly social and environmental sustainability (Raimo et al., 2022; Salvi et al., 2024). In addition to their sense of responsibility for the planet (Blowfield, 2012), they also perceive the corporate risk of operating in an environment where hunger, poverty, inequality and environmental stress thrive. Thus, companies have multiple interests in pursuing the 17 SDGs (Erin et al., 2022), and the call for their engagement in this area has now become imperative (Curtó-Pagès et al., 2021). This is because companies play a critical role in sustainable development (Bowen et al., 2017; Damiano and Di Maria, 2024; Elalfy et al., 2020; Haffar and Searcy, 2018; Heras-Saizarbitoria et al., 2022; Mio et al., 2020; Scheyvens et al., 2016). In particular, they can contribute to achieving SDGs by providing productive employment and decent work, ensuring responsible consumption and production, promoting sustainable economic growth and suggesting policy implications (Mengistu and Panizzolo, 2023).

Perceiving external pressure regarding their role in sustainable development (Nicolò et al., 2024b; Van Der Waal and Thijssens, 2020), companies seek to conform to market rules, sustaining their operational legitimacy and enhancing their corporate image (Garzoni et al., 2023; Nicolò et al., 2023b; Raimo et al., 2021). The role played by stakeholders in requesting the information is crucial for increasing their transparency and incentivizing sustainable value creation (Aguado-Correa et al., 2023; Brown and Dillard, 2015). However, excessive pressure could lead to incurring some risks, such as companies’ declaration of their commitment to the pursuit of the SDGs only for purposes of legitimacy or impression management (Silva, 2021), leading to the phenomenon of SDG-washing (Heras-Saizarbitoria et al., 2022). In these cases, according to legitimacy theory, companies would adopt behaviors responsive to stakeholder expectations, disclosing only the positive aspects concerning the SDGs (Gómez-Carrasco et al., 2021; Rickels et al., 2016), focusing on targets relevant to their value chain operations and remaining silent or avoiding measuring and operationalizing the SDGs when it is not convenient (Heras-Saizarbitoria et al., 2022; Van Der Waal and Thijssens, 2020). Those companies report for legitimacy reasons, even though they do not fully pursue these goals in practice (Elalfy et al., 2020), leading to an increase in symbolic disclosure aimed primarily at avoiding negative repercussions, seeking to prove to be compliant with laws and regulations and to meet internal and external stakeholders’ requests (Nicolò et al., 2023a, 2024c; Zampone et al., 2024), rather than to report on their commitment and efficiency (Heras-Saizarbitoria et al., 2022).

To avoid instances of SDG-washing and to better assess the credibility of disclosures, previous studies have emphasized the need to implement more efficient verification mechanisms (Heras-Saizarbitoria et al., 2022) and pointed out that greater stakeholder involvement in the verification of SDG disclosures would likely have a positive effect in increasing corporate accountability (Damiano and Di Maria, 2024; Elalfy et al., 2020). In addition, it would be necessary to prepare a universally accepted standard for reporting on the pursuit of the SDGs (Reyers et al., 2017), with more detailed outlines focusing on goals and performance (Johnsson et al., 2020), as at present completeness, materiality and accuracy are still absent (Heras-Saizarbitoria et al., 2022). To date, there are different approaches to disclosure. The most widely adopted reporting guideline is the Global Reporting Initiative (GRI), specifically the GRI 300 Environmental Standards, as its structure provides for the measurement and assessment of performance in compliance with certain SDGs (Elalfy et al., 2020; Ordonez-Ponce and Khare, 2021) and has been indicated internationally as a valid framework for their operationalization (Aguado-Correa et al., 2023; Erin et al., 2022). Over the years, the GRI has also attempted to improve its guidelines and established a joint initiative with the United Nations Global Compact (UNGC) to empower businesses to act and make the achievements of the SDGs a reality (Perello-Marin et al., 2022; Tsalis et al., 2020). Using this framework would ensure international comparability and consistency of reports among organizations (Erin et al., 2022). Unfortunately, a significant discrepancy still exists in SDG reporting (Miola and Schiltz, 2019).

Moreover, the accounting literature pinpoints how indicators currently used for reporting often fail to assess the level of actual SDG achievement (Arena et al., 2023; Rickels et al., 2016). In addition, it is recognized that the measurement of individual SDGs is still complex because of the presence of strong interlinkages between them, which current evaluation methods struggle to detect (Miola and Schiltz, 2019). Furthermore, SDGs seek a solution to wicked problems, which, according to the literature, are defined as problems “for which there can be no final solution since any resolution generates further issues, and where the solutions are not true or false or good or bad, but the best that can be done at the time” (Brown et al., 2010, p. 10). They thus require complex moral and systems dynamics to be understood so that measurement of the pursuit of these goals can be effectively performed (Bebbington and Larrinaga, 2014).

Furthermore, from the previous literature, the existence of companies’ prioritization of some goals over others emerges. Indeed, several studies show that some SDGs are mentioned more frequently in reports, demonstrating firms’ more significant interest in these issues (Forestier and Kim, 2020). Although this may seem a negative aspect, in cases where prioritization is reasoned and conscious, it demonstrates the presence of responsible internal reflection and the integration of the goals into corporate culture (Heras-Saizarbitoria et al., 2022).

Given the scarce previous studies on integrating the SDGs into corporate operations and reporting (Curtó-Pagès et al., 2021; Hummel and Szekely, 2022; Rosati and Faria, 2019; Tsalis et al., 2020), we have the purpose of contributing to this stream of literature by analyzing the content of SDG disclosure. Italy was chosen as the context of the study because of its acknowledgment by previous research as the European country with the highest level of SDG reporting (Bose and Khan, 2022). In detail, the current research aimed at analyzing the information included in nonfinancial reports produced in Italy in order to understand if Italian companies already include data on SDG measurement or achievement in their reports, highlighting the most and least disclosed SDGs and attempting to identify the reasons behind these decisions. Moreover, we outlined the information mainly discussed by Italian companies for the most disclosed SDGs, considering their operating industry sector. The research questions are formulated as follows:

Investigating the reasons for prioritizing certain SDGs in those companies may provide suggestions that help researchers and policymakers better understand the interventions and actions needed to improve the achievement of the 2030 Agenda (Mio et al., 2020).

In Italy, Legislative Decree No. 254 of December 30, 2016 – implementing Directive 2014/95/EU for the disclosure of nonfinancial information – introduced, in Article 2, the obligation to publish a statement (individual or consolidated) of nonfinancial disclosure (NFD) in the head of relevant public interest entities (EIPR), as defined in Legislative Decree No. 39 of January 27, 2010. In addition, Article 7 of Legislative Decree No. 254/2016 states that entities other than EIPR may also voluntarily publish a nonfinancial report, marking the report with a “compliance statement” if it is prepared under the provisions of the same decree. Annually, and following Article 3 of the Consob Regulation (Commissione Nazionale per le Società e la Borsa, which is the regulatory body of the Italian financial market) of January 19, 2018, implementing Legislative Decree No. 254/2016, Consob publishes on its website the list of entities that have issued the nonfinancial statement and, in case they have prepared it voluntarily, that have deposited the statement at the Companies Register and have notified Consob. The list is prepared by distinguishing entities that fall into the categories of companies with shares listed in Italy, other obligated EIPR and voluntary NFD. Moreover, Consob indicates the Framework used to prepare the report for each company. Concerning 2022, there were 208 Italian companies on the Consob list, and all of them used the GRI Framework. Two reports from the list were excluded from the analysis due to being written in English. Therefore, the resulting data set consisted of 206.

The collocation analysis method was used to examine the context of SDG-related terms in the reports collected. Collocations are “arbitrary recurrent word combinations” (Benson, 1989, p. 3). In particular, bi-gram analysis was implemented, meaning the focus was on a set of 2 words commonly used together throughout the text corpora. We indicated the most common use of certain word combinations based on the initial dictionary of keywords indication to understand the disclosure of the SDGs, similar to the existent approach of collocation analysis in corporate reporting (Myšková and Hájek, 2018). LogDice – a statistical measure of word association used in natural language processing – was calculated to measure the significance of collocation usage in the reports (Gablasova et al., 2017; Rychlý, 2008). This method allows the identification of critical collocations and thematic patterns in large text data sets and minimizes random word pairings by considering the exclusivity between words based on the possibility of the two words co-occurrence compared to their general frequency in the data set (Gablasova et al., 2017; Öksüz et al., 2021). This approach allowed us to detect patterns in language used by the companies when discussing their sustainability and SDG-related initiatives.

The collected reports were converted into text format in RStudio for further analysis. The texts had been preprocessed: numbers, punctuation and extra spaces were removed; package “stopwords” was used to exclude words in Italian without any semantic meaning; all words were stemmed (reduced to the root without suffixes and prefixes, also known as lemma) and tokenized (each word turned into a single unit of data). The resulting database (for text data also known as corpora) consisted of 8,571,170 words with a dictionary of 83,611 unique lemmas. Based on the SDGs’ names and their descriptions by the Italian Alliance for Sustainable Development[¹], we developed three keywords for each of the SDGs (Table 1), following a similar pattern of developing keywords for SDGs by prior studies (Brugmann et al., 2019), which have also been stemmed and tokenized for further analysis.

Next, bigrams were indicated (85,190 total for the corpora), and their frequencies were calculated. The resulting collocations list was filtered so that one of the words was from the pre-identified SDGs keywords list. The logDice function was created and then calculated based on the frequencies of each bigram and the words they were made of. To analyze the results, the bigrams were filtered to be mentioned at least five times throughout the data set, and the logDice measures more than 3.

To explore whether the frequency of SDG disclosure varied significantly across different industries, ANOVA test was performed (Table 2). The test was applied to the top three most disclosed SDGs – 7, 8 and 12 – across the companies in our data set. The p-values for each SDG were calculated to determine whether there were statistically significant differences between the frequency of SDG mentions across industries.

To identify the basic coverage of each SDG through all the reports, the aggregated average frequency share of each SDG per report was calculated. Figure 1 indicates the share of collocations over the total number of words in preprocessed text divided by the total number of reports.

The ANOVA test results (Table 2) showed statistically significant differences in SDG disclosure across industries with a p-value of 0.00659 for SDG 8, 3.05e−08 for SDG 12 and <2e−16 for SDG 7. These findings underscore the importance of industry-specific factors in shaping SDG disclosure, supporting the argument for tailored reporting frameworks that reflect the unique priorities of each sector.

Table 3 illustrates the top 10 results of collocation analysis with the highest logDice measures of keywords for the top three most used SDGs across the data set: 7, 8 and 12. Columns “word. x” and “word. y” represent the collocation itself, one being the SDG keyword. The words in the table are stemmed – cut to the root of the word. Columns “n.xy,” “n.x” and “n.y” illustrate the frequencies for the words used across the data set together, the word “x” and the word “y,” respectively. Column “logDice” contains the calculated logDice measure. For further analysis, only collocations with interpretable semantic meaning were used, and those not having any were omitted.

Table 4 shows the average share of the top three most used SDGs per report across the different industries available within our data set. The calculated average percentage per report is based on the number of collocations associated with each SDG divided by the total number of words in the preprocessed text. These collocations are filtered based on their occurrence at least five times throughout the data set and having a logDice score higher than 3.

Concerning SDG 7, “Ensure access to affordable, reliable, sustainable and modern energy for all,” companies mainly disclose their aim of lowering CO₂ production and mitigating the impacts of the energy crisis. They pursue these objectives by generating electricity from diversified renewable sources instead of traditional sources and promoting energy efficiency initiatives, clean electrification and renewable energy use, both within the organization and along the supply chain. In addition, numerical data on electricity consumed, produced and distributed, as well as the production of electricity and thermal energy from renewable sources – such as cogeneration, hydroelectric and photovoltaic plants – are disclosed in the analyzed reports. Indeed, companies are investing in modernizing their plants and networks to make them more resilient and reliable in producing cleaner and more affordable energy. Furthermore, some companies are championing the emergence of renewable energy communities, as they believe that their deployment will help accelerate energy use from renewable sources, and they are encouraging research into new solutions to increase the efficiency of existing systems and combat energy poverty. Based on the results, we can indicate that SDG 7 is more commonly disclosed in the energy sector, while the banks and the aerospace industries have the lowest frequency of coverage in our data set.

Disclosure on SDG 8, “Promote sustained, inclusive and sustainable economic growth, full and productive employment and decent work for all,” primarily concerns employees’ working health and safety in their place of work. The policies adopted by companies in this area aim to generate trust and increase people’s commitment to their work, ensuring a peaceful and inclusive work environment. The companies analyzed show a commitment to preventing discrimination, from the selection process to the definition of remuneration, from opportunities for professional growth to internal mobility, and they base their career management on fairness, sustainability and selectivity. Moreover, work employers demonstrate that they recognize their responsibility in talent development, employee training and skill development, and they aim to provide a path for professional growth within the company. The reports also provide detailed information on the number of dependent workers, job classification, type of working relationship, employment dynamics (entrances/exits), female employment, work accidents, hours worked and corporate inclusion policies undertaken. This information is reported rather consistently in reports from all sectors, with greater frequency in companies operating in the technology and automotive industries.

Regarding SDG 12, “Ensure sustainable consumption and production patterns,” companies mainly report their contribution to the circular economy, setting out their targets for reducing energy consumption and increasing the use of renewable energy. In particular, they report on energy produced and consumed, distinguishing the type of energy, where possible. They also set out data on the production and management of urban, hazardous and nonhazardous waste, specifying the quantities destined for disposal. In this case, the results show that the sector in which this information is most reported is chemical, while little relevance is given to this data in the banks and aerospace sectors.

The conducted collocation analysis identified which SDGs are most disclosed in the reports of Italian private companies. Findings demonstrate a prioritization of issues related to work and economic growth (SDG 8), responsible consumption and production (SDG 12) and energy (SDG 7). This prioritization likely stems from the direct relevance of these goals to companies’ core operations, where businesses can more easily quantify their impact. The result on SDG 8 is aligned with what Mengistu and Panizzolo (2023) have identified as the main contribution of businesses to the pursuit of the SDGs: to achieve maximum profit and ensure sustainable economic growth, enabling going-concern activities while ensuring the well-being and personal and professional growth of employees. SDGs 12 and 7 are instead related to environmental disclosure, demonstrating companies’ commitment to supporting the circular economy and the production and use of renewable energy, positively impacting sustainability. This result also aligns with the reporting framework used by the companies under study, the GRI Standards, which offer a particular focus on environmental goals (Elalfy et al., 2020; Ordonez-Ponce and Khare, 2021).

On the other hand, the study reveals less-reported SDGs that concern hunger (SDG 2), poverty (SDG 1) and peace, justice and institutions (SDG 16), respectively. These are issues that can be considered wicked problems (Brown et al., 2010), difficult to effectively measure (Bebbington and Larrinaga, 2014; Biermann et al., 2017; Damiano and Di Maria, 2024) and perceived by companies as more distant, on which they think having less impact because they are not intrinsic to their business activities. Moreover, in the Italian context – one of the largest economies worldwide – firms may perceive these issues as less characterizing the context in which they operate and, consequently, perceived as more distant and not to be prioritized. The analyzed companies carry out the prioritization process in a reasoned manner, showing their interest in the topics in which they think they can have a more significant impact (Heras-Saizarbitoria et al., 2022).

Overall, all companies tend to report information related to their commitment to pursuing the SDGs in their nonfinancial reports. This is even more appreciable given the absence of a regulatory reporting requirement in this area (Curtó-Pagès et al., 2021). This, therefore, demonstrates the increased recognition of the relevance of these issues by Italian companies and their engagement in pursuing these goals. Nonetheless, there are significant differences in the amount of information disclosed. This result is even more evident when comparing the results obtained in the different sectors: despite the use of the same reporting framework, disclosed information is dissimilar, showing how discrepancies in SDG reporting among organizations still exist (Erin et al., 2022; Miola and Schiltz, 2019).

However, what can be observed from the analysis of the information reported by companies is that, for the most disclosed SDGs, companies are trying to include – other than qualitative information about their goals and strategies for pursuing SDGs – also quantitative data and numerical indicators that actually measure their performance in the sustainability sphere. This contributes to preparing more substantive and easily verifiable reports, decreasing rhetoric and positively affecting corporate accountability (Damiano and Di Maria, 2024; Elalfy et al., 2020). However, it is important to consider potential biases in self-reported data, particularly in voluntary sustainability reporting. Companies may overemphasize positive aspects of their SDG contributions, selectively disclosing areas where they perform well while underreporting on goals where progress is lacking or harder to quantify.

The literature debate on SDG reporting is still in its infancy (Aguado-Correa et al., 2023). The obtained results aim to contribute to this conversation by studying SDG reporting activity in the Italian private sector using collocation analysis for 206 nonfinancial reports published in 2022. Research has found that the most disclosed SDGs in nonfinancial reports of Italian companies (RQ1) are those that companies can easily influence through their business activities, including increasing economic growth and employee well-being (SDG 8) or decreasing negative environmental impact (SDGs 7 and 12). At the same time, they focus less on wicked problems, which are more challenging to measure and on which they believe to have less effect (SDGs 1, 2 and 16). On the other hand, information that Italian companies include concerning the most disclosed SDGs in their nonfinancial reports (RQ2) is more quantitative and aims to measure their impact on pursuing the SDGs. This data is complemented by qualitative information related to the description of adopted corporate policies and actions, enabling them to seek the 2030 Agenda actively.

This research thus contributes to the discussion regarding the preparation and the content disclosed in voluntary reports inherent to the SDGs, highlighting companies’ efforts to measure their sustainable impact (Rickels et al., 2016). Compared to previous studies, this paper adopts a different innovative methodological approach in the field that enables going more in-depth with the analysis of SDGs reporting. Indeed, the results highlight the presence of this type of information, specify the content of the disclosure and point out the underlying rationale for SDG prioritization. The latter does not necessarily turn out to be a negative aspect, a symbol of misconduct or a partial interest in sustainable development (Heras-Saizarbitoria et al., 2022). Still, it represents a necessary company’s choice to make strategic decisions based on its own characteristics and potential.

Overall, we agree that there is still ample space for improvement in the quality and content of SDG reporting (Hummel and Szekely, 2022). Indeed, systematization of the reporting structure, with specific indicators to be included and requirements to be submitted, would make it easier for enterprises to prepare their reports and for stakeholders to compare the results of different organizations. Moreover, to develop this structured report, companies would be required to improve their internal control activities, and this would also positively affect corporate accountability and the pursuit of SDGs. This practically implies the need to urge policymakers and regulators to prepare a unified SDG reporting framework (Reyers et al., 2017). Given, however, the peculiarities of different industries – also highlighted by the results obtained in this study – concerning the information that companies consider valuable and meaningful to report, it would be appropriate, as GRI has already done on nonfinancial reporting, to prepare sector-specific standards for SDGs reporting as well.

Furthermore, more training would be needed to educate businesses on the role they can play in all 17 SDGs. Providing practical examples of actions that could be implemented could encourage their adoption by more firms. The preparation and dissemination of more SDG reports could also facilitate this process, as identifying and disseminating best practices knowledge would be more accessible. This would then promote improved performance in achieving all the goals set by the 2030 Agenda.

The present study undoubtedly has some limitations. The first relates to the context of the study: only nonfinancial reports from Italian companies were analyzed, focusing only on those that sent their reports to Consob. Furthermore, only nonfinancial reports were considered, while additional information regarding the SDGs published in other documents or channels was not included. Consequently, further studies could also consider these channels to get a holistic view of the SDGs reporting of these companies and expand the sample by considering a worldwide setting.

On the other hand, a second limitation concerns the time frame of the analysis, in that only reports related to 2022 were analyzed. To overcome this limitation, a longitudinal study could be carried out to highlight the trend over the years of SDG reporting, any increase in disclosed information and the evaluation of its quality. This evolution could be affected by the issue of the SDGs reporting framework by GRI and UNGC. Further research could then be conducted after its approval to assess its efficiency.

Finally, the results obtained from the collocation analysis provide partial insight into the process of SDGs prioritization within companies. The methodology chosen allows the study of a large sample of data (206 reports and 8,571,170 words total in this study) but lacks the investigation results out of the predefined keywords for each SDG. This means the results of this study may miss some information regarding SDG disclosure if companies used words different from the ones we have defined. Interviews with the companies’ management would be necessary to understand how the SDGs prioritization process takes place in more detail. This would then allow the results obtained through the analysis of nonfinancial reports to be confirmed and strengthened.