Corporate Social Responsibility Engagement through Social Media. Evidence from the University of Salerno

: Over the last few decades, stakeholders’ growing attention towards social and environmental issues has challenged universities’ traditional accountability boundaries, imposing the adoption of innovative reporting tools that facilitate stakeholders’ engagement in Corporate Social Responsibility (CSR) practices and performances. Against this backdrop, online communication tools, such as websites and social media platforms, have gained momentum as a pivotal means to increase dialogue with the myriad of stakeholders, especially during the pandemic period, as it has dramatically reduced physical interactions. Based on these premises, this study aims to dive deep into the use of social media to communicate CSR strategies in the university context by exploring the case of the University of Salerno. To this end, all posts published by the University of Salerno’s ofﬁcial Twitter account from 2015 to 2021 have been extracted and analyzed. Accordingly, the degree of interactions with stakeholders and the communication direction and balance level have been examined based on Carroll’s pyramid. Findings show a higher level of engagement for CSR posts and, in particular, for the philanthropic dimension. Results also highlight that, during the COVID-19 pandemic, the University of Salerno has empowered CSR disclosure through Twitter. The methodology adopted could be replicated for other universities to understand better how public universities use social media to involve a broader range of stakeholders in their CSR practices.


Introduction
In current decades, sustainability and Corporate Social Responsibility (CSR) were deemed pivotal drivers for value creation for both private and public institutions (Bezani 2010;Siboni et al. 2013). With the United Nations (UN) Agenda 2030, public organizations have been prompted to pay particular attention to sustainable practices and policies due to their strategic role in the journey towards worldwide sustainable development. In particular, universities that contribute to creating and disseminating scientific knowledge are called upon to integrate CSR principles in all of their academic activities, including research teaching and third mission activities (Lozano 2011;Aversano et al. 2020a;Nicolò et al. 2021aNicolò et al. , 2021b. As a result, universities are asked to communicate their commitment to CSR and sustainability to their stakeholders (Del Sordo et al. 2016;Moggi 2019).
Accordingly, universities have started to disclose their CSR practices and performances using different communication tools, such as sustainability reports, CSR reports, alternative non-financial reports, corporate websites, press, and others (Nicolò et al. 2021a;Schroder 2021). However, these channels do not allow the opportunity to engage with internal and external stakeholders as they are based on a one-way interaction (Gori et al. 2020;Schroder 2021). Thus, social media (SM) has arisen as a new communication channel that employs "mobile and web-based technologies to create highly interactive platforms via which individuals and communities share, co-create, discuss, and modify user-generated (1) To investigate the extent to which UNISA is exploiting the potential of SM tools (i.e., Twitter) to disclose CSR information and engage with stakeholders; and (2) To inquire into the degree of interactions with stakeholders and the communication direction and balance between UNISA and its stakeholders.
To this end, a supervised content analysis of all posts that the UNISA official Twitter account published from 2015 to 2021 has been performed, using a coding framework based on Carroll's pyramid.
The remainder of the paper is structured as follows: Section 2 presents the literature review results focused on CSR and stakeholder engagement in the university context. Section 3 presents UNISA and describes the research methodology. The results are provided in Section 4. Lastly, discussion, preliminary conclusions, limitations, and future research directions are outlined.

Literature Review and Theoretical Background
In recent years, the concept that universities should implement responsible policies and actions towards stakeholders and society as a whole has been discussed at length (Aversano et al. 2020b). Carroll provided one of the most authoritative definitions of CSR, according to which "the social responsibility of business encompasses the economic, legal, ethical, and philanthropic expectations that society has of organisations at a given point in time" (Carroll 1979, p. 500). Drawing on this definition, several scholars have started to interpret the concept of CSR in the university context (e.g., Kotecha 2010;Dima et al. 2013;Parsons 2014). These contributions have emphasized the pivotal role of CSR in supporting the core functions of universities-teaching, research, and third mission activities-becoming an intrinsic aspect that cannot be considered independently (Parsons 2014). Other academics have highlighted the fact that social responsibility is already incorporated into universities' mandates (Vallaeys 2007;Hayter and Cahoy 2018). Both of these perspectives show that universities have a strategic role, not only for global economic growth but also in developing a sustainable and socially responsible awareness in the coming generations (Rodríguez Bolívar et al. 2013;Moggi 2019).
The relevance of CSR in the university context has also drawn attention among European and international policy-makers. In the last decades, numerous initiatives have been implemented (Tbilisi's declarations, Talloires Halifax and Lüneburg, the Copernicus Charta and RIO+ 20) and settled several international declarations aimed at integrating sustainability and CSR pillars in academic activities (e.g., United Nations (UN) Decade of Education for Sustainable Development 2005-2014; UNWorld Action Program Education for Sustainable Development) (Alonso-Almeida et al. 2015;Nicolò et al. 2021a). More recently, the UN Agenda 2030 for Sustainable Development launch has extended the concept of CSR in the university context to the broader concept of sustainability . In particular, the fourth Sustainable Development Goal (SDG) had the purpose to "ensure inclusive and quality education and promote lifelong learning opportunities for all" (Agenda 2030 2015, p. 14).
Following this political momentum, the growing awareness of the role of universities as vehicles for socio-economic development, as well as for driving towards the transition to a society based on sustainable development, has increased the stakeholders' needs and expectations of the community creating greater accountability and engagement needs (Van Weenen 2000;Nicolò et al. 2021a).
Accordingly, the theoretical perspective on which this work is drawn integrates two theories: the stakeholder theory and the legitimacy theory.
The stakeholder theory assumes that companies must create long-term value by establishing a relationship with their stakeholders (Carroll 2021). Freeman (1984) has defined stakeholders as a group or individual who can impact or be impacted by the organization's performance. In the higher education context, universities and stakeholders must cooperate to reach a mutual interest according to their shared values, and CSR is one of the values which has to inspire cooperation (Hörisch et al. 2014). The range of university stakeholders is particularly wide (e.g., territory, public administration, local enterprises, students, graduates, professors, researchers, administrative) and its variety makes the identification difficult (Sica et al. 2021). However, the university's context, within which higher education institutions are considered organizations with a public mandate, is a unique scenario where the stakeholders' classification requires a deeper investigation (Ferrero-Ferrero et al. 2018). Burrows (1999) has provided a stakeholders' classification in the context of higher education. More specifically, a four-dimension classification has been developed, distinguishing stakeholders according to their (i) location, (ii) involvement status, (iii) potential for cooperation and interest, and (iv) influence on the organization. However, stakeholder participation and involvement in the university's activities depend on each institution's governance and regulation (Ferrero-Ferrero et al. 2018). In this journey towards sustainable transition, universities are called to engage with their stakeholders to create a virtuous and strategic network and co-create value for society (Phillips 1997;Corazza and Saluto 2020).
Consequently, "the more an organisation engages with its stakeholders, the more accountable and responsible that organisation is towards these stakeholders" (Greenwood 2007, p. 1). Accordingly, universities are called to fulfil the public's request for accountability, providing all information related to their activities, policies, and strategies. Social media can help universities improve their stakeholders' engagement since it enables the interaction between parties in a two-way dialogue in which mutual expectations could be shared (Bellucci and Manetti 2017;Bellucci et al. 2019). Accordingly, scholars have defined SM as a powerful tool of dialogic communication, supporting the dialogical accounting system which shares information on the stakeholders' expectations (Bebbington et al. 2007). This issue makes SM strategic in the CSR field, where it is progressively essential that universities take stakeholders' engagement seriously (Thomson and Bebbington 2005).
In this perspective, the use of SM has transformed the reporting dynamics activating a dialogue that can be strategic to seek legitimacy. The legitimacy theory posits that the actions of organizations should comply with the system of values, norms, and expectations that is established within the social context in which they operate (Suchman 1995). Accordingly, universities are called to undertake activities to manage their CSR performance in order to obtain legitimacy from society and survive (Pellegrino and Lodhia 2012). A legitimacy gap emerges when a disparity emerges between two value systems (Yongvanich and Guthrie 2007). Thus, organizations are called to manage societal expectations and legitimacy gaps to survive (Nicolò et al. 2021b). This legitimization process is particularly vital for organizations, such as universities, which need social and political supports . Organizations that need a political and social endorsement require greater legitimacy than others (Lodhia et al. 2020;Vitolla et al. 2019a). In this perspective, scholars have demonstrated that "the legitimacy of higher education in society will increasingly be a direct function of the nature, quality and evolving ties with the stakeholder society" (Jongbloed et al. 2008, p. 307). Accordingly, the interlinkage between the stakeholder theory and the legitimacy theory in the university context is particularly evident. Given the intrinsic features of higher education institutions, the need to identify their stakeholders occurs not only at an institutional level but also at all levels, such as departmental levels (Trireksani et al. 2021). In fact, each university belongs to a disciplinary community that needs to establish partnerships and create a network among departments to support scientific development in several disciplinary fields (Jongbloed et al. 2008). In this perspective, the wide range of stakeholders that have to be involved in the value creation process has emerged. Accordingly, higher education institutions' value systems have to agree with society to avoid potential legitimacy gaps and meet stakeholders' expectations (Manes Rossi et al. 2018). In doing so, universities are called to engage with their stakeholders.
Scholars have shown that CSR and sustainability information disclosure to stakeholders could help organizations in the legitimation process (Pellegrino and Lodhia 2012;Ferrero-Ferrero et al. 2018;Vitolla et al. 2019b).
The CSR disclosure emerges as a mechanism that universities can undertake in order to reach this objective. The information that public institutions traditionally provided (exclusively based on financial issues) has mainly been reported in their annual reports, making it difficult to completely address the stakeholders' accountability needs (Castelo and Lima 2008). Several scholars have investigated the use of traditional non-financial reporting tools, such as sustainability reports, social reports, and integrated reporting in the university context (e.g., Fonseca et al. 2011;Ceulemans et al. 2015;Ferrero-Ferrero et al. 2018;Gamage and Sciulli 2017;Aversano et al. 2020aAversano et al. , 2020bNicolò et al. 2020;Adhikariparajuli et al. 2021;Trireksani et al. 2021). However, these tools do not provide prompt access to information to universities' stakeholders (Esposito et al. 2021a). During the last decades, the development of Information and Communication Technologies (ICT) has determined a shift in communication strategies and practices, allowing organizations to easily reach stakeholders and disclose CSR information. Therefore, universities have started to use ICT to improve transparency and accountability, thereby improving their legitimacy (Rodríguez Bolívar et al. 2013). Accordingly, some scholars have investigated how universities disclose their sustainability and CSR information through their websites (e.g., Manes Rossi et al. 2018;An et al. 2019;Brusca et al. 2020;Nicolò et al. 2021a). These contributions have emphasized that websites provide flexibility and speed to information processes. However, communication on websites is based on a one-way interaction process, according to which universities can disclose information but stakeholders are unable to interact with universities regarding specific information. The spread of SM to disclose information has not only supported universities in disclosing CSR information but also in engaging with their stakeholders. Consequently, higher education institutions must develop disclosure strategies for CSR in order to fulfil stakeholders' informational expectations and provide a socially responsible image (Rodríguez Bolívar et al. 2013;Kashmanian et al. 2011). To this end, universities have started to use new accounting instruments to disclose sustainability and CSR information (Moggi 2019).
In order to increase the dialogue with stakeholders, higher education institutions have started to use SM to activate two-way interactions between universities and their stakeholders. In keeping with these arguments, academics have started to explore CSR and sustainability disclosure in the university context through SM (e.g., Hamid et al. 2017;Ndou et al. 2018;Gori et al. 2020). In particular, Hamid et al. (2017) have investigated the role of Facebook to engage university students and staff on environmental sustainability from a worldwide perspective. They have highlighted that social media should be useful for conveying higher institutions' policies and supporting their sustainability transition. However, Ndou et al. (2018) and Gori et al. (2020) have focused their research on the European scenario. More specifically, Ndou et al. (2018) have explored intellectual capital disclosure from a Big Data perspective, focusing on an Albanian University. Investigating both Facebook's profile and the university's website, they have demonstrated that intellectual capital disclosure is mainly provided unintentionally. Instead, Gori et al. (2020) have explored the sustainability disclosure that the University of Florence (Italy) provided, analyzing posts published on their official Facebook page. Unlike previous research, this study has shown a low interaction level between the University of Florence and its followers, hampering SM platforms' "dialogic potential" (Gori et al. 2020, p. 1).
However, despite SM's relevance in disclosing CSR information and engaging stakeholders, the university context still remains unexplored.
Aiming to fill this gap, the present research focused on one Italian university, providing preliminary insights from the CSR disclosure and the level of engagement and interactions among stakeholders through Twitter.

Context Description: The University of Salerno and ITS CSR practices
The present paper focuses on the UNISA, a leading Italian university seated in Fisciano, a small town in the province of Salerno. The University of Salerno has paid great attention to CSR and sustainability issues during the last decades, implementing several socioenvironmental tools (e.g., Social Reports; Gender Reports) and establishing initiatives and policies oriented towards sustainable development (e.g., Plan for Gender Equality (GEP); R&I Peers-Pilot experiences for improving gender equality in research organizations; social inclusion projects; courses based on renewable sustainability resources and CSR; etc.) (Sica et al. 2021).
As a member of the Network of Universities for Sustainable Development (Rete delle Università per lo Sviluppo Sostenibile, (RUS 2020)), UNISA has started to implement and communicate sustainability practices in order to co-create value for the territory and catalyze the path towards achieving the 17 UN SDGs. Moreover, establishing internal regulatory bodies for sustainability and CSR (e.g., Equal Opportunities Committee, Interdepartmental Observatory for Gender Studies and Equal Opportunities, Single Guarantee Committee, Ethical Commission, etc.) has bolstered the presence of CSR principles and values in university policies.
After the spread of COVID-19, UNISA has empowered socially responsible investments to support the territory and its community in fighting in a resilient way against the virus. Accordingly, UNISA has been called to develop new strategies to disclose and promote CSR practices, inside and outside the university, to inform and engage with all relevant internal and external stakeholders (Table 1). In this perspective, SM has emerged as a pivotal way to cope with all critical issues related to the physical distance imposed by the global pandemic and reach a considerable level of CSR public engagement.

Research Methods
Following previous studies on this theme (Pizzi et al. 2020;Schroder 2021), the present paper aims to investigate the level of CSR disclosure and its corresponding degree of stakeholder engagement of UNISA through Twitter.
This SM was chosen for several reasons: first, Twitter is one of the most used social networks in the world (Mergel 2013). Second, it promotes civic participation and interaction between public administration and society (Wukich and Mergel 2016). Third, Twitter allows an excellent open-source platform for scholars to collect and analyze text data (Panagiotopoulos et al. 2014). Moreover, despite the character limitation of Twitter posts, the shortness of the messages can reach a broader audience and companies can publish more frequently (Kim et al. 2014). Accordingly, Twitter has been considered a suitable social network to build engagement (Pizzi et al. 2020).
The paper has been developed implementing the research methodology described in Figure 1, performing a Content Analysis of both qualitative and quantitative natures. Through a data mining activity, all tweets published by the UNISA official Twitter account were extracted from 2015-the starting date of Twitter disclosure activity-to 17th of July 2021-the date on which the analysis was carried out. This period has been chosen because our purpose was to analyze the evolution of the UNISA Twitter communication pathway. However, our data do not consider posts published after the data extraction, providing only a partial view of the disclosure trend of 2021.
The data mining process was performed using "NCapture", an open-source extension of NVivo software based on Application Programming Interface (API), which allows access to the public accounts on Twitter (Reyes-Menendez et al. 2018). For each tweet, the data extracted included the publication date, the number of "likes", and the number of "retweets". This information allows us to investigate the number and type of interactions among stakeholders. More specifically, as suggested by Pizzi et al., "the 'favourites' analysis made it possible to evaluate the users' satisfaction with the specific proposed content, while the analysis of 'retweets' allowed us to assess the level of debate among stakeholders" (Pizzi et al. 2020, p. 4).
After the extraction, all data were cleaned and prepared for analysis. The data analysis has been structured in three sub-steps ( Figure 1): (1) dictionary-based content analysis using a supervised machine learning technique through NVivo software, (2) manual content analysis performed by two coders independently, and (3) results integration.
proposed content, while the analysis of 'retweets' allowed us to assess the level of debate among stakeholders" (Pizzi et al. 2020, p. 4).
After the extraction, all data were cleaned and prepared for analysis. The data analysis has been structured in three sub-steps ( Figure 1): (1) dictionary-based content analysis using a supervised machine learning technique through NVivo software, (2) manual content analysis performed by two coders independently, and (3) results integration. The authors have developed a thematic dictionary adapting the coding framework that Pizzi et al. (2020) and Nicolò et al. (2021a) proposed. In particular, using the CSR theoretical paradigm that Carroll (1991) formulated, four items have been identified (i.e., economic responsibility items, ethical responsibility items, philanthropic responsibility items, and legal responsibility items). For each item, a group of words referring to the university's context were selected in order to classify the four dimensions of CSR according to the higher education scenario ( Table 2). Tweets that did not fall into these categories have been classified as "NO-CSR" content. Since UNISA's tweets are in Italian, a group of words has been developed in Italian and then translated into English in order to avoid linguistic bias (Esposito et al. 2021b).  The authors have developed a thematic dictionary adapting the coding framework that Pizzi et al. (2020) and Nicolò et al. (2021a) proposed. In particular, using the CSR theoretical paradigm that Carroll (1991) formulated, four items have been identified (i.e., economic responsibility items, ethical responsibility items, philanthropic responsibility items, and legal responsibility items). For each item, a group of words referring to the university's context were selected in order to classify the four dimensions of CSR according to the higher education scenario ( Table 2). Tweets that did not fall into these categories have been classified as "NO-CSR" content. Since UNISA's tweets are in Italian, a group of words has been developed in Italian and then translated into English in order to avoid linguistic bias (Esposito et al. 2021b). The Supervised Machine Learning technique adopted allows the automatic quantification of the occurrences using the NVivo software. Moreover, in order to check whether the messages have an informing or interacting nature (Morsing 2006;Schroder 2021), two Adm. Sci. 2021, 11, 147 8 of 15 independent contractors have carried out manual content analysis, adopting an empirically grounded approach (Miles and Huberman 1994). This approach permits carrying out a deeper qualitative analysis without the data precoding that is performed by the software. In particular, the coders have classified the tweets as informing if they expressed (a) activities, (b) performance, (c) initiatives, or (d) teaching offers of UNISA, while interacting if they outlined a commitment to stakeholders (e.g., students, local authorities, NGOs, banks, employees, etc.) (Schroder 2021). In addition, this research has investigated the level of stakeholder engagement by analyzing the communication direction of each message. More specifically, UNISA's messages that triggered a comment by a Twitter user were classified as a "two-way communication" and alternatively were coded as "one-way communication" (Schroder 2021). Since the content analysis is susceptible to subjective interpretation, Krippendorff's alpha index (α) was calculated in order to assess the inter-coder reliability (Krippendorff 2018). Krippendorff's alpha coefficient-calculated on the 30 July 2021 on the first 25% of tweets-is equal to 0.87. This value is considered acceptable since it is within the theoretical range between 1.00 (exactly equal) and 0.00 (completely different) (Hayes and Krippendorf 2007). The results of the two-step analysis have been integrated and synergistically examined.
The following section will present and discuss the results obtained from the analysis described above.

Results
Table 3 summarizes the descriptive statistics of "CSR" and "NO-CSR" tweets published by the UNISA official account from 2015 to 2021. Results show that 81.15% of UNISA social communication is classified as a non-CSR message, while only 18.85% focus on CSR content. However, in line with Gori et al. (2020), we found that the CSR tweets have a higher level of engagement than the non-CSR messages (i.e., likes for CSR tweets: mean = 68.5; st. dev. = 69.2; retweets for CSR tweets: mean = 2.75; st. dev = 2.01). The low value of retweets (i.e., CSR retweets: mean = 2.75, st. dev. = 2.01; non-CSR retweets: mean = 2.62, st. dev = 1.98) for both CSR and non-CSR posts suggests that UNISA's followers interact with all contents, mainly providing feedbacks (likes) and not sharing posts (retweets). Table 3 shows the classification of UNISA's tweets.   Nicolò et al. (2021a). In fact, the traditional university communication approach, based on formal one-sided communication, has collapsed with the rise of the COVID-19 pandemic, which has imposed the need to converse with the university's stakeholders. It is also interesting to stress that the highest number of "CSR-tweets" was published in 2020. This evidence confirms that, in an emergency state, the university as a public entity is called to disclose CSR practices and policies in order to obtain legitimacy from its primary stakeholders and support the restart of the whole economy and education system. The results provided for 2021 are partially considered and discussed since the tweets extracted are limited to the date the analysis was performed (i.e., 17 July 2021). university's stakeholders. It is also interesting to stress that the highest number of "CSRtweets" was published in 2020. This evidence confirms that, in an emergency state, the university as a public entity is called to disclose CSR practices and policies in order to obtain legitimacy from its primary stakeholders and support the restart of the whole economy and education system. The results provided for 2021 are partially considered and discussed since the tweets extracted are limited to the date the analysis was performed (i.e., 17 July 2021). All of the tweets classified as CSR content have been codified following the theoretical paradigm that Carroll proposed. The codification process has provided the following results. Table 4 shows the descriptive statistics for each CSR dimension (i.e., philanthropic, ethical, legal, economic). Our findings suggest that UNISA has oriented its communication process predominantly to philanthropic issues (i.e., 50.48%). In contrast, only 12.13% is based on economic content. These results are in line with previous studies on CSR disclosure (e.g., Pizzi et al. 2020;Schroder 2021), according to which there is a need to communicate issues that overcome the economic dimension, going deeper into socially responsible actions and policies. However, the average rate of likes shows that philanthropic and legal issues generate a similar level of engagement (i.e., mean of likes for philanthropic posts = 25. All of the tweets classified as CSR content have been codified following the theoretical paradigm that Carroll proposed. The codification process has provided the following results. Table 4 shows the descriptive statistics for each CSR dimension (i.e., philanthropic, ethical, legal, economic). Our findings suggest that UNISA has oriented its communication process predominantly to philanthropic issues (i.e., 50.48%). In contrast, only 12.13% is based on economic content. These results are in line with previous studies on CSR disclosure (e.g., Pizzi et al. 2020;Schroder 2021), according to which there is a need to communicate issues that overcome the economic dimension, going deeper into socially responsible actions and policies. However, the average rate of likes shows that philanthropic and legal issues generate a similar level of engagement (i.e., mean of likes for philanthropic posts = 25.3; mean of likes for legal posts = 25.3). The lowest engagement level is detected in the economic dimension (i.e., mean of likes for economic posts = 15.2). Nevertheless, posts on ethical issues provide a higher average number of retweets (i.e., 3.32).
Considering the direction and balance of the UNISA communication activity on Twitter, Table 5 shows the annual distribution from 2015 to 2021. Overall, the highest percentage of UNISA posts is classified as two-way communication (i.e., 87.65%). Moreover, in this case, 2020 represents a breaking point for the university in the communication and engaging process (i.e., 293 posts). This classification allows us to measure the level of stakeholder engagement that the SM communication strategy generated. As described in Figure 3, UNISA has always reached a higher level of engagement. However, during the last three years, it has empowered the intensity of its communication, obtaining a growing level of involvement by its followers and representing both internal and external stakeholders. Moreover, in this case, 2020 represents a breaking point for the university in the communication and engaging process (i.e., 293 posts). This classification allows us to measure the level of stakeholder engagement that the SM communication strategy generated. As described in Figure 3, UNISA has always reached a higher level of engagement. However, during the last three years, it has empowered the intensity of its communication, obtaining a growing level of involvement by its followers and representing both internal and external stakeholders.  Concerning the balance of communication, Table 5 shows that 66.66% of the tweets published by the UNISA have informative content, while only the 33.34% have the purpose of interacting with their followers. These results are also confirmed for the "CSRtweets", analyzed in Table 6. Accordingly, the 89.81% of the CSR posts have been classified as two-way communication while the 73.30% as informative messages. Concerning the balance of communication, Table 5 shows that 66.66% of the tweets published by the UNISA have informative content, while only the 33.34% have the purpose of interacting with their followers. These results are also confirmed for the "CSR-tweets", analyzed in Table 6. Accordingly, the 89.81% of the CSR posts have been classified as two-way communication while the 73.30% as informative messages.

Discussion and Conclusions
Nowadays, growing attention in enhancing the social engagement of higher education institutions and the consequential information disclosure about social responsibility issues is becoming a vital aspect of universities' accountability. In particular, during the COVID-19 pandemic, university systems were forced to revise their traditional paradigm in order to radically face the emergency. Attuned, universities have started to invest in intangible resources to continue their activities, even at a distance. Among them, investments in SM communication emerge as one of the most revolutionary options. In particular, the preliminary results obtained from our case analysis have shown that UNISA has provided a higher level of disclosure through SM and, in turn, has obtained a growing percentage of engagement and interaction compared to the pre-COVID-19 period. In this context, analyzing the impact of the pandemic on accountability practices represents a challenge in higher education institutions. Thus, our findings should be interpreted in light of these considerations.
Previous studies on CSR in the university scenario (e.g., Lozano 2011; Hayter and Cahoy 2018) have highlighted that universities primarily promote socially responsible policies and have been called to integrate CSR into their activities. However, our results show that UNISA has not predominantly adopted a communication strategy based on CSR content. This result implies that CSR disclosure should not be considered an ordinary practice in the UNISA' SM communication practice.
Meanwhile, in line with Aversano et al. (2020a), the engagement level of UNISA's official Twitter account shows a discrete ability to involve its stakeholders, especially during the pandemic. These findings are also in line with Gori et al. (2020), according to which the University of Florence has based its communication strategy predominantly on community engagement. These results should be justified because SMs are considered suitable tools for creating engagement between organizations and people (Ndou et al. 2018). Unlike previous studies, which have highlighted that universities publish more information on the ethical dimensions (Lozano 2011) or environmental (Gamage and Sciulli 2017; Lopatta and Jaeschke 2014), our results have shown that philanthropic issues are the issues that UNISA most disclosed. Indeed, economic issues are predominantly disclosed in traditional financial reports, while environmental issues are not particularly addressed because they are likely to be disseminated on websites or because universities avoid providing such information to prevent social pressures on their activities. Following our coding framework, the philanthropic dimension is focused on occupational issues and education and training issues.
Moreover, health and social inclusion are considered aspects of the philanthropic dimension. Furthermore, this result should be interpreted in light of the pandemic scenario since the emergency state has impacted the need to disclose information related to health and territory.
Concerning the direction of UNISA's communication via Twitter, our results have demonstrated that the highest percentage of the published posts involved two-way communication. However, our study has not explored the university's answers to its followers' reactions. This limitation places our results in contrast with previous studies on SM disclosure in different sectors (e.g., Manetti and Bellucci 2016;Carrasco et al. 2020;Gori et al. 2020), according to which SM is partially exploited, using a traditional one-way communication approach.
In conclusion, our preliminary findings highlight that SM as new disclosure tools can reveal a great deal of unknown information about CSR practices and policies in the universities relevant to creating a dialogue with students, researchers, professors, employees, local institutions, and society at large. Following previous studies (e.g., Epstein and Widener 2010;Aversano et al. 2020a), the stakeholder dialogue emerges as a crucial aspect that allows universities to include their perspectives in the management and strategies of universities. Accordingly, our study also confirms the primary role of universities in guiding sustainable development and the improvement of society. This research may have implications for both university managers and academics. Managers can empower CSR communication through SM in order to disseminate the universities' attention to its stakeholders. Universities should share their knowledge on disclosing CSR practices and performances and increase engagement with their relevant stakeholders to improve the current non-financial reporting level. Moreover, it would be useful to publish posts with interacting content that stimulate the digital debate to empower the engagement level with universities' stakeholders. The paper has also revealed theoretical implications. In particular, even though the literature on this topic is scant, SM are considered the best tools to establish a dialogue with stakeholders.
Furthermore, the need to increase the recognition of the benefits deriving from the CSR disclosure in the university context should motivate scholars to go deeper in providing useful suggestions to public managers in order to orient universities' communication strategies. Moreover, scholars can investigate the CSR disclosure from SM in the light of different theories, such as the institutional theory, comparing the applicability of the stakeholder theory and legitimacy theory within the university context. Lastly, academics can replicate this analysis, adapting our coding framework to the peculiarities of different sectors.
However, this research is not without limitations. As previously highlighted, our analysis does not examine the UNISA answer to stakeholder reactions. Accordingly, a future study could go deeper in the communication direction, investigating this aspect. Moreover, the analysis only focuses on UNISA's Twitter account. Future works should investigate other SM, such as Facebook, Instagram, and LinkedIn, comparing the results obtained from these different communication channels. In addition, the results should be compared with information provided on the websites and the UNISA sustainability report.
Lastly, multiple case studies should be conducted to assess CSR disclosure and stakeholder engagement among different national, European, and international universities.