Priority Stakeholders’ Perception: Social Responsibility Indicators

: This study aims to build a list of composite indicators by information that enable the assessment of philanthropic higher education organizations’ (PHEOs) social responsibility based on the interests of their stakeholders. A list of 88 social responsibility indicators was built based on a literature review and stakeholder interest to serve as a basis for the composite indicators. In order to identify and validate the indicators, ﬁeld research was carried out. Stakeholders from Brazil and the United Kingdom scored them from one (not important) to ﬁve (very important) for each indicator identiﬁed in the literature review. With 540 valid answers, they suggested inclusions and exclusions according to their interests. Next, a correlation analysis was performed to identify and eliminate redundant indicators. The principal component analysis extracted the composite indicators. The results point to 11 principal components that are conﬁgured as composite indicators to evaluate the performance of PHEOs social responsibility considering the stakeholder’s interests, a factor that differentiates this research from the literature revision done. Some of the composite indicators are close to the social responsibility categories reviewed in the literature. However, others show more speciﬁc and in-depth interests, especially regarding the stakeholders themselves. These composite indicators help managers establish disclosure policies whenever they are focused on seeking legitimacy in the social context of PHEOs. It also contributes to the advancement of theoretical knowledge, presenting composite indicators, from the stakeholder’s perspective, for the disclosure of social responsibility of PHEOs.


Introduction
Social responsibility is a key concept for management and sustainability of organizations. It involves all the organization's activities and the value chain, and considers the stakeholder perspective [1], while legitimizing the organization itself [2][3][4][5], it is a complex and constantly evolving concept [1,[6][7][8][9][10][11][12][13][14][15][16] that must be understood in the macroeconomic development. In other words, social responsibility means adopting behaviors that go beyond the indicative metrics of economic growth, namely gross domestic product (GDP) and/or per capita income. It means social welfare, which must also be measured by the other components comprising the human development index [17].
Thus, social responsibility can be classified into dimensions or categories (hereafter considered categories): Economic, financial, environmental, social, ethical, legal, products and services, strategic, governance, and stakeholders [3,18]. Therefore, the performance of organizations should not be analyzed solely based on economic and financial results. The analysis should include information covering the other categories of social responsibility different regions and from both countries. This relationship aggregates several indicators creating a principal component model and uses the strength of composite indicators to evaluate the performance of organizational social responsibility. Social responsibility indicators are a way of representing an informational reality about the organization, which in this case refers to PHOE. Its usefulness is related to the possibility of improving communication, thus reducing the information asymmetry between the organization and the stakeholders [18,40].
Finally, it is highlighted that this article reached priority stakeholders from two countries (Brazil and the United Kingdom) from different cultural backgrounds and development levels, which may mean the adoption of sustainability policies with different characteristics [41]. However, the PHEOs of the two countries are similar in the social constitution form and provision of educational services. All of them have electronic internet pages containing, among others, information about their social responsibility practices. Neither were identified in the literature reviewed, indicators created from the expectations of its stakeholders, for the disclosure of PHEOS of either country. In addition, this research did not seek consensus on the main indicators, but rather to collect all the information considered relevant by the priority stakeholders, identifying, and validating indicators that go further than the reviewed literature. Following, the indicators were clustered by means of the principal component, creating the composite indicators which also are configured as the contribution of this research to the disclosure of PHEOs.
From the results obtained indicators, suggested by the stakeholders, in the research form, were not recognized in the reviewed literature. As well as some composite indicators created which are different from the categories of social responsibility found in the literature. Some of them show specific and in-depth interests, especially regarding the stakeholders themselves. These composite indicators will serve as an alternative for the social responsibility categories, regarding the advancement of theme knowledge, and will, also, serve as a basis for the establishment of the PHEOs disclosure policies attending to the priority stakeholders' expectations.
Following, this article contains a literature review, methodological procedures, research results, final considerations, and references.

Literature Review
Social responsibility performance is measured by the impact of organizations' activities on their stakeholders [42][43][44][45]. Stakeholders are the beneficiaries of organizational actions and have specific interests in them. They can affect and be affected by organizations. Therefore, they are those who evaluate their behavior, legitimizing them or not to remain in the communities in which they operate [2,4,5,18,[46][47][48][49].
However, serving all the interests of all stakeholders may not be possible or strategic for organizations. In this sense, organizations should identify the key audiences that should be addressed primarily [3,[50][51][52][53][54]. The priority stakeholders are the owners, collaborators, customers, suppliers, and the community where the organizations are inserted, that is, those that most directly affect or are affected by the organizations [3,48,54].
Social responsibility is an abstract attribute that cannot be directly observed in the products and/or services provided by organizations, which creates the need to disclose information about it to society. The informational efficiency on social responsibility requires the organization to establish strategies that focus on stakeholders. In other words, organizations should publish information that is of interest to key stakeholders, thereby establishing a competitive advantage [43,48,54,55].
Organizations use voluntary disclosure of social responsibility information to manage their performance. Such disclosure informs stakeholders of the impacts of the organization on the social system in which it operates [5,18,29,42,56]. There are several initiatives by entities to guide disclosure by organizations. These include the Sustainability Disclosure Database [57], Global Reporting Initiative (GRI) standards to general disclosure [58], ETHOS indicators for sustainable and responsible businesses [59], self-assessment policies and management systems for university social responsibility [60], International Standard of Social Accountability 8000 [61], Form 990 [62], Form 20-F [63], Brazilian Technical Accounting Standard 15-NBC T 15 [64], and guidance on corporate responsibility indicators in annual reports [65]. These initiatives were built upon listening to experts who are knowledgeable about social responsibility. They offer a general information perspective that attempts to embrace stakeholders' demands. However, the diversity of the informational interest of stakeholders requires that they be heard to identify their demands and enable disclosure to contribute to the efficiency of the informational policy established by the organization [18]. If, on the one hand, specialists present idealized indicators, given their capacity and knowledge resulting from their training, on the other hand, stakeholders present a pragmatic perspective on information that they consider relevant. From this perspective, this study is relevant.
It is worth noting that most empirical studies reviewed here use various ways to choose what to disclose. The most widely used methodology is the empirical analysis in disclosure media, mainly annual published reports of organizations [74][75][76][77][78][79][80][81][82][83][84][85][86][87][88][89], to name some of the most recent studies, and in annual university reports [90,91]. Undoubtedly, the organization's report is an important source as it allows the shareholder to litigate against the organization whenever a conflict of interest is established because of the materiality of information. However, the annual report is configured as one of the communication channels, which requires a relative technical knowledge for its understanding, a feature that is not common among stakeholders. This restricts this source as a basis for the construction of social responsibility information indicators.
Other methodologies used to construct an indicators list are empirical literature reviews [88,[92][93][94][95], specifically at universities [96]. As well as guidelines established by institutions organizations that recommend information on social responsibility that needs to be disclosed [97][98][99], specifically at universities [100,101]. However, both the empirical studies reviewed here, and the institutional initiatives start from an idealization about the interests demanded by stakeholders in terms of disclosure of information about social responsibility. In addition, the construction of indicators list is also identified through consultations with experts on the subject with the aid of statistical and econometric tools [102,103].
Virtually all these methodologies are based on a reference that is not configured in the manifestation of stakeholders since they were not asked about what the information would be, they would like to see disclosed, ending up ignoring their expectations. The exception has been the GRI, which listens to several stakeholders, but the organization establishes requirements for the reports of organizational sustainability, with specific rules for economic, environmental, and social aspects [58]. However, recently, the empirical literature reviews show that indicators considering stakeholders' interests and/or expectations have been created, but that it is still incipient [3,18,104] due to the difficulty of listening to the interests of stakeholders or the complexity of the concept attributed to social responsibility.
Thus, to define representative information indicators, in addition to the stakeholder's perspective, it is necessary to consider their adherence to some principles: Exact definition, easy interpretation, applicability, measurability, comparability, relevance, clarity, and representation of reality. They should reflect the abstract concept to be analyzed [20,[105][106][107][108][109][110][111][112]. In addition, the frequency of assessments should be defined to evaluate the performance of organizations. The costs of obtaining and monitoring information should not exceed the benefits generated [18,42]. On the other hand, performance evaluation should not consider the predominant approach of analyzing the most widespread and therefore comparable standardized economic and financial indicators alone [42,56].
Indicators contribute to the knowledge of reality through representations. They are tools for measuring and monitoring reality [18,40]. Therefore, there is a need to listen to the interests of stakeholders in the process of creating indicators of the social responsibility of organizations. These indicators may represent information that establishes a legitimate relationship vis-à-vis the organization's stakeholders in the community in which the organization operates. In addition, the composite indicator aims to "tell a story", promoting greater latitude discoveries of a given phenomenon [113] (p. 776). In view of this, composite indicators are popular in some areas, such as in economics, for example, GDP, what increases in interest in them and, therefore, in their creation, covering several themes [113,114], the which can be extended to the PHEOs social responsibility.

Methodological Procedures
Many indicators are representative of organizations' social responsibility, as evidenced by the empirical literature review. Their importance does not allow them to be ignored in this research. Thus, the first stage was configured by the construction of a list containing 95 social responsibility indicators, from the literature review of several theoretical and empirical articles on the topic [58][59][60]62,63,65,74,[76][77][78][79][80]82,[85][86][87][88][89][90]92,93,95,96,98,101,115]. The indicators were selected considering they should be related to the identified social responsibility dimensions; present in the largest number of reviewed literatures; or considered important and or complementary for the disclosure of PHEOs.
In the second stage, by building the indicators list, field research was carried out with priority stakeholders of PHEOs to verify interests for information about social responsibility and the relevance of each indicator in the initial list. It is noteworthy that in Brazil there are 1004 PHEOs subdivided into 798 colleges, 137 university centers, and 69 universities [116]. In the United Kingdom, there are 143 universities with characteristics of organizations with non-economic purposes [117].
To identifies the interest of the PHEOs' priority stakeholders, a form was constructed where they were asked about the degree of importance of each indicator for the disclosure of PHEOs' social responsibility. That is, for each proposed indicator, respondents assigned a score ranging from one (not important) to five (very important). In addition, stakeholders suggested inclusions of information considered relevant and exclusions of non-relevant ones, from their perspective. The form was applied in person in some cases and sent to PHEO stakeholders via social networks in others.
In the United Kingdom, at the Ethics Committee recommendation, the forms were applied in person and through the corporate intranet, reaching Durham University employees (professors and administrative staff) and students. In Brazil, some forms were also applied in person at PHEOs in the south of the country, reaching collaborators, students, suppliers, and some community representatives. The in-person application occurred in 12 PHEOs and resulted in 116 valid responses. In addition, the forms were sent by emails from several universities, from different country regions to students and collaborators. WhatsApp, Facebook, and Instagram were also used to send the forms, reaching the community in general. For this form of application, the number of PHEOs reached was not controlled. From the forms sent, 540 valid responses were obtained, as shown in Table 1. This stage of the research was concluded with a list composed of 88 indicators of information representing social responsibility (Appendix A). Evaluations of indicators and stakeholder's information-enabled descriptive statistics analyses [118,119]. Correlation and principal component analyses were used to reduce the number of indicators proposed and to extract principal components (composite indicators) that facilitate the assessment of the social responsibility of PHEOs. In other words, a set of indicators was transformed into a new variable, reducing the overall number of indicators and, even so, maintaining the possibility of measuring the organization's social responsibility, through the composite indicators.
The first step of the principal component analysis for component extraction was to assess the suitability of indicators for the analysis. This evaluation was performed using two criteria: Kaiser-Meyer-Olkin measure of sampling adequacy and Bartlett's test [42,120]. Kaiser's sample adequacy measure, the eigenvalue, assesses the adequacy of indicators by comparing the paired and partial correlation coefficients and assigns values between zero and one to them [3,42,[121][122][123]. The Kaiser-Meyer-Olkin measure of sampling adequacy should have values above 0.60. However, it is preferable to be greater than 0.70 [124]. Bartlett's test is used to test whether the variance of variables is homogeneous or not [125].
Cronbach's alpha was also calculated. It is a measure of internal consistency of indicators and a reliability estimator of assessment, allowing determining the lower limit of internal consistency of a group of variables, thus validating the set of indicators of the sample and the components extracted. The suggested minimum threshold for this test is 0.7 [3,121,126,127]. Cronbach's alpha and the Kaiser-Meyer-Olkin measure of sampling adequacy must be directly related [42,121,123].
Subsequently, the principal component technique that organizes uncorrelated components based on their decreasing variance was used. The first was the component with the greatest variance, and the last was the component with the smallest variance. The principal component analysis attempts to reduce the number of variables to illustrate the variance of the original variables better. The calculation of the load by the principal component method is considered unambiguous. The increase in the number of components does not change the original components [120,128]. A rotation of components was performed to facilitate the attribution of indicators to the extracted principal components. The orthogonal technique, i.e., rectangular varimax, was chosen for the rotation. Using this technique, the components remain uncorrelated after rotation [129]. Analyzing each line, it is possible to define the components of each principal component according to the variables that are most strongly associated [120]. The number of principal components was determined using Kaiser-Meyer-Olkin measure of sampling adequacy, and only components with values of 0.7 or higher were selected. They were also validated by Cronbach's alpha, reaching values above 0.8. The calculations were performed using the software Eviews 9.5 and Excel 2013.

Research Results
This presents the topic of the descriptive statistics, correlation, and principal component analyses-the extraction of the principal components that are configured as composite indicators.

Descriptive and Correlation Analyses
Initially, some descriptive statistical information is presented from stakeholders in Brazil and the United Kingdom who responded to the survey. Of the 540 respondents, 224 are men and 316 are women. Of which, 194 did not complete higher education. While, 182 completed higher educations, 89 have lato sensu specialization, and 75 have stricto sensu postgraduate studies. It is noteworthy that, considering all stakeholders classified as priorities, of equal importance for PHEOs disclosure policies, their participation, in the research sample, was not proportional. The distribution by stakeholder type, country, can be seen in Table 1.
Regarding the 88 indicators evaluated by the stakeholders, who assigned values between one and five, a summary of the descriptive statistics is presented in Table 2. The average score of all indicators together was: 4.03. The standard deviation shows that the variation between the assigned scores is low, which is also observed by verifying the minimum and maximum averages assigned by the stakeholders for the indicators' set.
The indicators considered more and less important by the stakeholders are presented in Table 3. It is noteworthy that the research did not aim to compare perceptions between types of stakeholders, therefore, the results are general. The correlation analysis was carried out to check correlated and redundant indicators to remove them from the list that served as the basis for building the composite indicators of organizational social responsibility. The value of the correlation coefficient expresses the intensity of relationships between indicators, making it possible to identify those that provide approximately similar information. Indicators showing a strong correlation (higher than 0.8) with statistical significance are presented in Table 4. Based on the correlation analysis results, seven redundant social responsibility indicators were removed from the basic set for carrying out the principal components analysis: Profile of officers and board members, as they were included in the employee profile; selection process of officers and board members, as they were included in the employee selection process indicator; remuneration of officers and board members, also, because they were included in the compensation of employees; and surplus on revenue, because it is considered the best indicator of equity [130]. Therefore, 81 indicators remained in the list used for principal components analysis. Although some of them also have a significant degree of correlation, they were considered complementary for the PHEOs disclosure.

Composite Indicators Extraction
Initially, it is necessary to demonstrate the individual indicators list of social responsibility adequacy carried out using Kaiser-Meyer-Olkin measure of sampling adequacy, Cronbach's alpha, and Bartlett's test.
Kaiser-Meyer-Olkin measure of sampling adequacy indicate the individual suitability of the components. For all components, the results of the eigenvalues were higher than 0.8, confirming that they are adequate to compose the indicators list. The Kaiser-Meyer-Olkin measure of sampling adequacy, with all the components, shows this suitability by the eigenvalue, reaching 0.9457. The Cronbach's alpha, which reached 0.9849, indicates the internal consistency of the indicators list. It is noteworthy that the values are significant mainly due to the number of the form respondents, i.e., 540. They provide consistency to the results. In addition to the Kaiser-Meyer-Olkin measure of sampling adequacy and Cronbach's alpha statistics, by Bartlett's test, the null hypothesis that the components are not interdependent is rejected and, therefore, the basic requirements for using principal component analysis are satisfied, as shown in Table 5. Having fulfilled the requirements to continue the analyzes, 81 components were extracted which together explain the total variance of the original variables. The first 14 components with an eigenvalue higher than one, account for 72.69% of the variation. The varimax rotation solution meets the objective of the applied principal component analysis, reducing the number of components to 11, and improving results interpretation. Figure 1 presents with the scree plot of the eigenvalue after the varimax rotation.  The following are the individual components (81 indicators) that integrate each principal component, hereinafter called composite indicator. Statistics are also presented. They confirm the validity and the suitability of the results: Cronbach's alpha and Kaiser-Meyer-Olkin measure of sampling adequacy.
The first composite indicator, presented in Table 6, consists of 14 indicators representing economic and financial information, or that result in the composition of the value of organizations. Thus, this composite indicator can be called economic and financial information. Economic and financial responsibilities have been the subject of academic discussion since 1770 in the works of Adam Smith [71]. For social responsibility, economic sustainability must come first. Only after this responsibility is met should organizations address the other interests of their stakeholders [6,8,10,11]. It appears that the composite indicator was very close to what the literature has been presenting as an economic and financial category [15,93,131]. This demonstrates that the stakeholders' perspectives are already recognized by the literature in this regard.
The second composite indicator, presented in Table 7, aggregates three indicators. They include information about the members of PHEOs boards and can, therefore, be called representations in the PHEO boards. Some studies analyze the impacts of boards on organizational activities, identifying that their composition reflects on the social behavior of organizations [132][133][134]. This composite indicator demonstrates that stakeholders understand the importance of counselors in the PHEOs governance system, just as with corporations. In the reviewed studies, information about the boards is contained in the category of organization's governance [3,59,134].
The third composite indicator, presented in Table 8, is composed of eight indicators. They express the need for information about the environment and the investments that are made for its preservation. Therefore, it can be called environmental information. Organizations should disclose information about their environmental impacting activities relevant to the analysis of organizational sustainability [29,58,67,89,95,135,136], including universities [137]. The environmental category of the organization's responsibility has been studied at least since the 1970s. In this sense, the PHEOs stakeholders demonstrate to be in line with what the literature has been presenting [29,66,68,86,97,131,135,136,138,139].
The fourth composite indicator, presented in Table 9, is composed of seven indicators. They express issues related to the social involvement of organizations. It can be called interaction and social investments. Table 9. Interaction and social investments.

Interaction and social investments
Interaction with the community Investments in philanthropy Social contribution report Social investments Social responsibility policies Support policies for social projects University extension projects The social responsibility of any type of organization must encompass social programs and policies, interacting with and targeting the general well-being of the community. In the reviewed studies, these indicators are analyzed within the social category [6][7][8][9][10][11]59,68,137,138]. However, by this research results, the stakeholders demonstrate to seek more specific and detailed information about the interaction with the community, the policies and the investments types made to demonstrate the organization's social return [2,4,5,43,44]. In this sense, the composite indicator highlights this expectation from stakeholders [42,113,114].
The fifth composite indicator, presented in Table 10, aggregates nine indicators. They express the organization's strategic and governance information. It can be called PHEOs governance information. Governance information is part of the information that should be disclosed about social responsibility. This is because a qualified governance system indicates that organizations act with social responsibility [3,60,140]. PHEO stakeholders also feel this need, demonstrated by the formation of this composite indicator, in line with what the literature has already presented as a social responsibility category [3,42,113,114].
The sixth composite indicator, presented in Table 11, aggregates seven indicators. They represent information about the relationship of PHEOs with their collaborators. It can be called work policies. Collaborators are also the primary stakeholders of PHEOs. Organizations have assumed responsibilities and commitments towards them that need to be met [3,48,54,58,59,96]. After all, the values received from the organization may affect not only employees but also their dependents. The sustainability of the organization represents the responsibility in the social environment in which it operates [10] (p. 5). The formation of this composite indicator demonstrates the specificity of interests with each stakeholder type separately, differently from what has been advocated by the reviewed literature [3,58,59,66,68,75,139,141,142].
The seventh composite indicator, presented in Table 12, contains eight information indicators involving the students of PHEOs. It can be called student care. Students are also PHEOs priority stakeholders [3,48,54]. Therefore, disseminating specific information regarding the attention given to them is expected by the stakeholders, as observed by the formation of this composite indicator [17,42] and it becomes strategic for organizations and represents a socially responsible behavior [55].
The eighth composite indicator, presented in Table 13, is formed by six indicators. They express general information about primary stakeholders, and therefore, can be called stakeholders profile. Stakeholder information has been considered important for designating social responsibility since the 1980s [72], but more strongly since 1990 [66,68,139]. The result of this composite indicator is consistent with the stakeholder's category presented in the literature review made [3,58,59,66,68,75,139,141,142]. This shows that stakeholders, to legitimize organizations [2][3][4][5]96], in addition to specific information about each stakeholder type, also want general information.
The ninth composite indicator, presented in Table 14, comprises six indicators. They represent diverse information about the behavior of organizations, which can, therefore, be called: Organizational history and behavior.

Composite Indicator Individual Indicator
Organizational history and behavior

Donors (major) Information on gender at work Information on minorities at work Information on outsourced services
Information on race at work Organization social history Regarding the commitments made, there is a need for respect for human rights, the establishment of labor policies, and conduction of activities that build the history of the organization. These aspects would legitimize the organization with the stakeholders [3,5,58]. However, in the reviewed literature the organization's history was not included in any social responsibility category. As for other information related to stakeholders, were included in the stakeholder's category [3,58,59,66,68,75,139,141,142] but they were not addressed as organizational behavior, which is done with the creation of this composite indicator. Thus, once again the importance of composite indicators is demonstrated [3,42,113,114].
The tenth composite indicator, presented in Table 15, comprises seven indicators. They express concerns about the ethical and moral behavior of organizations. It can be called ethical behavior of the PHEOs. Table 15. Ethical behavior of the philanthropic higher education organizations (PHEOs).

Composite Indicator Individual Indicator
Ethical behavior of the PHEOs Access to the consumer law Accidents at work Conduct code Employee assistance and benefits Ethical commitments Ethics committee Health and safety at work Ethics represents the respect of organizations for the principles, values, uses, customs, and culture of the community in which it operates. That is, organizations should operate as expected from them [3,6-10, [34][35][36][37][38][39]. With the creation and consistency of this composite indicator, it is possible to perceive the consonance of the expectations of the stakeholders with what the reviewed literature has been presenting.
The eleventh and last identified composite indicator, presented in Table 16, comprises six indicators. They express the importance of the support that PHEOs give to their stakeholders, the research, and the course evaluation policies. Thus, this component can be called stakeholder satisfaction and support policies. Table 16. Stakeholders satisfaction and support policies.

Composite Indicator Individual Indicator
Stakeholders satisfaction and support policies Course evaluation policy (government) Course evaluation policy (PHEO) Psychological support to students Student satisfaction survey Support to government campaigns Support to government projects There are several stakeholders who should have their interests met and to whom organizations should be accountable for regarding their activities [46][47][48][49]51,54]. Therefore, it is part of the organizational social responsibility to establish policies to meet these interests and measure the degree of stakeholder satisfaction [3,18]. The formation of this composite indicator shows that stakeholders want to identify the organization's involvement with the community [5], observing what it does and what is the feedback from stakeholders contemplated by the distributed social benefits. Thus, in possession of this information, they would be prepared and or willing to legitimize the organization in the community where it operates.

Composite Indicators Summary
In summary, the results of this research allow us to infer 11 composite indicators representing social responsibility considered important by the surveyed stakeholders, as shown in Table 17. It is noteworthy that the composite indicators are presented by the internal consistency verified from Kaiser-Meyer-Olkin measure of sampling adequacy and accompanied by Cronbach's alpha. Both measurements validate the composite indicators extracted by the principal component analysis. Some composite indicators are similar to the social responsibility categories presented in the literature review. Economic and financial information is equivalent to the economic and financial category presented in the literature since a long time ago and confirmed by Carroll in 1979 [6]. Environmental information is equivalent to the environmental category [29,66,68,86,97,131,138]. Information on the organization's governance is equivalent to the governance category [3,60,140]. The stakeholder profile is similar to the stakeholder category [3,58,59,66,68,75,139,141,142]. Finally, the ethical behavior of organizations is equivalent to the ethical category [6,8,10,11,59,68]. These results confirm that there is harmony between the expectations of stakeholders with the knowledge brought by the reviewed literature.
The composite indicator interaction and social investments could be considered an integral part of the social category [6,8,10,11,59,68] but it denotes the deeper expectation for specific information about how PHEOs are interacting with society and for which projects PHEOs are returning the expected social benefits, according to the social contract established between stakeholders and PHEOs [2][3][4][5].
The other composite indicators present specificities and details depth of the expectations of the PHEOs' priority stakeholders, especially with regard to the relationship between them, that is, the fulfillment of the existing social contract between the stakeholders and the PHEOs [3][4][5]. PHEO board representations composite indicators; work policies; students care; organizational history and behavior; and stakeholders satisfaction and support policies, express this specific and profound interest. Considering that priority stakeholders are the main legitimizers [48,51,54] of PHEOs, it is strategic that they meet their expectations. Legitimacy is the condition that arises when organizations operate in accordance with the value system of the society in which they operate, but the values differ between stakeholders [143]. In this sense, the composite indicators created in this research are a reference for knowing the expectations of priority stakeholders. They contribute both to the formulation of PHEOs disclosure policies, in the search for legitimacy, and to the stakeholders in the PHEOs social responsibility assessment process.
It is noteworthy that the composite indicators were not created with weighting, as well as the individual indicators. However, in Appendix A, there is an average importance degree, attributed by the stakeholders, for each indicator, which can serve as a basis for assigning weight to the indicators if the PHEOs and or the stakeholders deem it relevant to use.

Final Considerations
The advance of communication technologies allowed stakeholders to access information through social media, among other communication channels, which until recently was concentrated on media that had information hegemony. These are times when the manifestation of wills and what is accepted or not by society is no longer controlled by the classical media. In this context, there are organizations that may or may not be valued by the community in which they are inserted according to the image built by the disclosure of information about their responsibility towards their stakeholders. Stakeholders are increasingly interested in non-financial social responsibility information. They evaluate this information and include it in their decision-making about whether to legitimize organizations.
The study objective was to list composite indicators comprising information that enabled the assessment of PHEOs' social responsibility based on the interests of priority stakeholders. The field research with PHEO stakeholders in Brazil and the United Kingdom allows us to list eleven composite indicators representing social responsibility from a list of 81 indicators listed from the literature review and survey with priority stakeholders.
These composite indicators aggregate information regarding stakeholder profile, environmental information, social interaction and investments, PHEO governance information, economic and financial information, board representation, student care, work policies, organizational history and behavior, ethical behavior of PHEOs, and stakeholder satisfaction and support policies. Some of these composite indicators are close to the social responsibility categories reviewed in the literature. However, others show more specific and in-depth interests, especially regarding the stakeholders themselves. So, together, these 11 composite indicators guide managers in terms of disclosure policy and indicate the interests of stakeholders about the organization's expected social responsibility. In other words, these composite indicators of social responsibility, when disclosed, enable PHEOs to legitimize themselves in terms of what priority stakeholders understand as social responsibility, contributing to organizational sustainability.
Thus, assuming that social responsibility is a fundamental concept for management, which must involve the entire organization and value chain for the organization to legitimize itself with its stakeholders, this research presents the composite indicators valued by them. Such composite indicators should not be ignored by managers upon establishing disclosure policies whenever they are focused on seeking legitimacy in the social context of PHEOs. In this way, the composite indicators contribute to both stakeholders and PHEOs, as their disclosure reduces information asymmetry and facilitates the PHEOs legitimization by stakeholders.
This research differs from the reviewed ones in that it presents the stakeholders' perspective and the information regarding social responsibility they consider important, corroborating the other empirical studies on the subject. Given the differentiating aspect, but limited to some stakeholders, it is considered important to extend the research not only to samples of stakeholders from other countries but also to samples of stakeholders from organizations with other characteristics, like publicly traded universities.
Recognizing the importance of the theme studied and the respective limitations, we highlight here future research. It is worth noting that the development of analysis of these indicators, in communication channels of these associations, configure as a future study. Being no less important would be analyzing the legitimacy of these indicators with stakeholders of other organizations types.  Acknowledgments: We would also like to thank all the Philanthropic Higher Education Institutions in Brazil and the United Kingdom that allowed primary data collection to be carried out.

Conflicts of Interest:
The authors declare that they are professors at three educational institutions whose stakeholders were also respondents to the research. However, it is noteworthy that none of them responded to the forms and did not participate directly in the data collection when it took place at the university where they work.