Corporate Social Responsibility Reporting: A Content Analysis in Family and Non-family Firms

Abstract

Family firms are ubiquitous and play a crucial role across all world economies, but how they differ in the disclosure of social and environmental actions from non-family firms has been largely overlooked in the literature. Advancing the discourse on corporate social responsibility reporting, we examine how family influence on a business organization affects CSR reporting. The arguments developed here draw on institutional theory, using a rich body of empirical evidence gathered through a content analysis of the CSR reports of 98 large- and medium-sized Italian firms. The grounded theory analysis informs and contextualizes several differences in the type and content of corporate social responsibility reports of family and non-family firms. Our findings show that in comparison to non-family firms, family firms disseminate a greater variety of CSR reports, are less compliant with CSR standards and place emphasis on different CSR topics. We, thus, contribute to the family business and corporate social responsibility reporting literatures in several ways, offering implications for practice and outlining promising avenues for future research.

Introduction

Corporate social responsibility (CSR) reporting is defined as the process of communicating the social and environmental actions of organizations to particular interest groups within society and to society at large (Campbell 2004; Gamerschlag et al. 2011; Gray et al. 1987, 2001). The literature on CSR reporting has greatly increased in the last few years alongside the development of CSR practices on the assumption that this implies that the company is accountable for its actions (Perrini 2005). Indeed, society generally attaches high importance to organizations that are committed to reporting their socially responsible principles and activities (Fisher et al. 2001). Companies spend a great deal of effort and money on disclosing information on their social and environmental initiatives as this allows them to generate moral capital (Gamerschlag et al. 2011). Moreover, there is a growing body of evidence of firms leveraging on CSR reporting to nurture their competitive advantage and overcome economic and financial downturns (Hooghiemstra 2000).

Alongside the increasing number of companies that report their social and environmental activities to the public, the diversity of the type of CSR reports issued is marked (Perrini 2005). In addition, social and environmental reports are increasingly accompanied by third-party assurance statements (Kolk and Perego 2008), often supplemented with additional material on corporate websites.

An increasing stream of the literature has investigated how specific firm characteristics such as size (e.g. Morhardt 2010) or contextual factors such as media exposure (e.g. Reverte 2009) affect firm CSR reporting behaviour. A more limited number of studies also focus on investigating how internal factors such as the appointment of a CEO (Campbell 2000) or the presence of a social reporting committee (Cowen et al. 1987) affect the level of CSR reporting. However, the studies on internal determinants of CSR reporting are few (Adams 2002; Prado-Lorenzo et al. 2009). Furthermore, no studies have investigated how a particular type of internal factor, namely, family influence on a business organization, affects CSR reporting (De Massis et al. 2012).

This scarcity is surprising given the ubiquity of family firms and their crucial role across all world economies (La Porta et al. 1999; Schulze and Gedajlovic 2010; Villalonga and Amit 2006). This is particularly true in Italy, where a joint survey led by the Italian Association of Family Enterprises (Aidaf) and Bank of Italy shows that family businesses represent 94 % of the nation’s companies and employ 98 % of the workforce in manufacturing companies (European Commission 2008). Some of the historical and most well-known family firms in the world are Italian, and Beretta (Ward and Lief 2005), Fiat (Tapies and Toninato 2005) and Alessi (Verganti 2009) are just some examples of famous family firms that actively disclose CSR reports. Anderson and Reeb (2003) also show that a third of S&P 500 firms are controlled by the founding families. Hence, CSR reporting within this important sector of the global economy warrants investigation, particularly since there are conceptual reasons to believe that the process of disclosing the social and environmental actions of an organization differs in family and non-family firms. Indeed, a vast body of empirical evidence points to the distinctive nature of family firms in terms of other business processes, for example, corporate governance (Randøy and Goel 2003), internationalization (Piva et al. 2013), innovation (De Massis et al. 2013a) and financing (Romano et al. 2001). The importance of both economic and non-economic goals in family firms depends on the presence of the family involved in the business and characterizes their behaviour (Kotlar and De Massis 2013). Indeed, family firms enhance their sustainability by preserving their socio-emotional wealth, meant as “the nonfinancial aspects of the affective endowments of family owners” (Berrone et al. 2012, p. 2). This suggests that there may also be important differences in the disclosure of social and environmental actions that distinguish family and non-family firms.

In this article, we attempt to fill this gap in the literature, advancing our understanding of how family influence on a business organization may affect the type of CSR report and content disclosed. Specifically, we aim to address the following research questions:

RQ1

How do family firms differ from non-family enterprises in the type of CSR reports issued?

RQ2

How do family firms differ from non-family enterprises in the topics communicated in similar types of CSR reports?

The arguments of this paper draw on institutional theory using a rich body of empirical evidence gathered through a content analysis of the CSR reports of a sample of 98 large- and medium-sized Italian firms. The grounded theory analysis informs and contextualizes several differences between family and non-family firms in their CSR reporting. Our findings show that family firms disseminate a greater variety of CSR reports, are less compliant with CSR standards and emphasize different topics in the content of their CSR reports when compared to their non-family counterparts.

The structure of the paper is as follows. The next section provides the theoretical background of the research. The third section illustrates the method employed in the empirical analysis, the fourth section reports and discusses the main findings of the empirical study, while the last section concludes and outlines avenues for future research.

Theoretical Background

Institutional Theory and CSR Reporting

This study draws on institutional theory to examine CSR reporting. Institutional theory, in turn, encompasses arguments from both the stakeholder and the legitimacy perspectives, which provide interesting insights to comprehensively present and discuss the topic under investigation. The concept underlying institutional theory is that firms are called on to respond to industry regulatory and competitive norms and values (DiMaggio and Powell 1983; Scott 1995, 2001). From an institutional perspective, the study scope is comparatively broad since CSR reporting is positioned within the responsibility systems in which business, government, legal and social institutions operate (Matten and Moon 2008). Firms are embedded in a wide array of political, economic and social institutions that affect their behaviours (e.g. Campbell 2007; Campbell et al. 1991) and are driven by the need to respond to environmental pressures. According to institutional theory, the sources of external pressure are multiple, i.e. regulative, normative and cognitive (Scott 1995). Regulative pressures are the rules, laws and political structures that govern an industry; normative pressures include the norms and values that organizations are expected to respect; cognitive pressures refer to the social interactions between business and community actors to develop a dominant strategic approach and avoid uncertainty (Dibrell 2010; Di Maggio and Powell 1983; Reay 2009). These institutional pressures explain organizations as business entities operating in coordination and with the support of other actors in their field, interacting with their stakeholders and striving for social acceptability and credibility, i.e. legitimacy, to survive in their context (Scott 2008). CSR reporting, thus, reflects the evolution of firm governance systems to respond to broader stakeholder concerns and institutional requisites, evidencing their key priorities and accountabilities (Young and Marais 2011). According to the stakeholder perspective, CSR reporting is viewed as part of the dialogue between the firm and its stakeholders, while CSR is considered a relatively successful means of negotiating relationships with stakeholders (Gray et al. 1995a). In addition, firms are motivated to undertake socially responsible initiatives and communicate these to achieve high visibility and respond to the continuous monitoring of their stakeholders and society (Chiu and Sharfman 2011). CSR reporting is seen as a governance practice to inform the firm’s stakeholders on ‘good’ practices (Clatworthy and Jones 2001; Godfrey et al. 2003; Merkl-Davies and Brennan 2007; Yuthas et al. 2004), which is closely linked to achieving and maintaining a good reputation (e.g. Bebbington et al. 2008; Fombrun et al. 2000; Young and Marais 2011). To build a solid reputation in the market and obtain stakeholder trust, companies must show their commitment to CSR by providing clear and verifiable data and information (Perrini 2005), thus, developing their image and identity, and achieving a competitive advantage (Hooghiemstra 2000). Therefore, business entities need to legitimize their actions (Hooghiemstra 2000) as predicted by arguments from the legitimacy perspective: an increase in social disclosure “represents a strategy to alter the public’s perception about the legitimacy of the organization” (Deegan et al. 2000). Legitimacy is defined as the perception that the actions of a firm are desirable in a socially accepted system of norms, values and beliefs (Suchman 1995). Hence, in line with Hooghiemstra (2000), CSR reporting aims to provide information that legitimizes the company’s behaviour by influencing stakeholder and society perceptions (e.g. Neu et al. 1998; O’Donovan 1997) in such a way that it is considered a ‘good corporate citizen’ and its actions justify its continued existence.

Along this stream of research, a relevant contribution for our purpose is provided by Matten and Moon (2008), who define corporate social responsibility as either ‘explicit’ or ‘implicit’. Explicit CSR generally consists of voluntary programs and strategies that combine social and business values. It is perceived as part of the firm’s social responsibility and, therefore, constitutes proactive decision-making. The initiatives in the explicit CSR category usually respond to the willingness of the firm to dialogue with its stakeholders, in line with a stakeholder perspective. Implicit CSR instead refers to the company’s role within institutions and its duty to address stakeholder issues and as such is considered a reflexive reaction to the corporation’s institutional environment (Porter and Kramer 2006). Implicit CSR, thus, relates to the need to respond to institutional pressures to which firms are exposed. Both these dimensions are responses to the different institutional pressures that organizations are subject to, and CSR reporting is a useful tool to manage the need to communicate their behaviour.

Prior empirical research on CSR reporting has focused on the documents that organizations disclose to communicate their efforts and commitment to CSR. Most studies concentrate on the analysis of codes of ethics and investigate the factors that affect CSR reporting and how these effects are achieved (Béthoux et al. 2007; Singh 2006; Singh et al. 2011; Wood 2000). This body of the literature provides evidence on how firm size, industry type, profitability, country and media exposure influence the level of CSR reporting (e.g. Adams et al. 1998; Chen and Bouvain 2009; Deegan et al. 2000; Gray et al. 1995b; Guthrie and Parker 1989; Hackston and Milne 1996; Jose and Lee 2007; Neu et al. 1998; Patten 1991, 1992; Reverte 2009; Tagesson et al. 2009). Worth noting is that the level of CSR disclosure increases with firm size (Morhardt 2010). The country the organization reports in and the firm’s country of origin are found to have a significant effect on the characteristics of CSR disclosure. Indeed, companies based in different countries have divergent perspectives on the importance of being publically perceived as socially responsible and the CSR issues that should be prioritized and disclosed (Maignan and Ralston 2002). Empirical studies on CSR reporting revolve around firm characteristics and the contextual factors affecting the level of CSR reporting. As affirmed by Adams (2002) and Prado-Lorenzo et al. (2009), the factors that potentially affect the level, quality and quantity of CSR reporting can be grouped into three categories: (i) corporate characteristics, e.g. size and industry; (ii) contextual factors such as country of origin, time, media or stakeholder pressure and (iii) internal factors, e.g. CEO appointment or a social reporting committee. Only a limited number of studies focus on the third category. For instance, Campbell (2000) identifies in his study that the level of CSR reporting depends on changes in the company chairperson; Cowen et al. (1987) show that a corporate social reporting committee is an antecedent of a high number of social reports. Prado-Lorenzo et al. (2009) find that the ownership structure in general matters: the presence of stockholders whose personal image and social reputation are strongly associated with the evolution of the firm usually fosters the development of CSR disclosure practices. To the contrary, investors with reduced equity shares have limited interest in CSR reporting. Based on these premises and the potential impact of internal factors on CSR reporting, it is rather surprising that no studies have yet been undertaken to understand the role of family influence as an internal factor that potentially affects firm decisions on the type of CSR report to issue and its content. The next subsection briefly reviews the literature on CSR in the family business context, highlighting the distinctive traits of family firms.

Overview of the literature on CSR in family business and distinctive traits of family firms

Scholars have only recently begun to study the issue of corporate social responsibility in the family business context (De Massis et al. 2012). Given the scarcity of studies on family business CSR reporting, we provide an overview of the findings of prior studies on CSR in family firms. This brief literature review is useful to understand the state-of-the-art of research on CSR in family firms and identify their distinctive traits that can potentially affect their CSR reporting behaviour.

A first category of studies involves empirical analyses on samples exclusively composed of family firms. Deniz and Suárez (2005), through the analysis of a sample of 112 Spanish family firms, show the heterogeneity of CSR behaviour in different types of family firms. The study identifies the system of values that characterize social interactions in family firms as important aspects affecting their CSR orientation; indeed, values often characterize relationships among family members and with external stakeholders, rendering formal information almost unnecessary (e.g. Steier 2001). Moreover, the authors identify the potential sources of heterogeneity of family firms, showing that part of the sample has a narrow spectrum of CSR behaviours due to lack of professionalism, while others have a broad spectrum of CSR behaviours when the focus is on reputation, long-term orientation and respect for tradition. An informal approach to manage the relationships with the stakeholders and a search for legitimization emerge as distinctive traits of family firms that are attentive to their image and reputation. Niehm et al. (2008), in a study on 221 small and mostly rural family firms, show that CSR orientation is beneficial to firm performance; they emphasize the role of the values and attitudes of the owning family on CSR behaviour, highlighting a relevant link with the surrounding community and, thus, a propensity to dialogue with the society’s constituents. A third contribution by Uhlaner et al. (2004) focuses on the relationships between the family firm and its stakeholders and, through in-depth interviews with 42 small and medium-sized Dutch family firms, establishes that family involvement in the business affects employee, customer and supplier relationships. Overall, family business research emphasizes that family firms are particularly inclined to be legitimized in their action, increasing their visibility and reputation with customers, suppliers and the entire community (Dunn 1996). In this regard, Ding and Wu (2013) find that younger firms among family firms are not particularly concerned about their reputation and consider the preservation of socio-emotional wealth (Gómez-Mejía et al. 2007) to a lesser extent. Instead, when family firms are more mature, consistent with a legitimacy perspective, they tend to behave more ethically to preserve their reputation and improve their image since the main aim of these family firm owners is intra-family succession (e.g. Chua et al. 1999; De Massis et al. 2008).

A second category of studies consists of comparative analyses on mixed samples of family and non-family firms. Adams et al. (1996) investigate the ethical behaviour of family versus non-family firms, but their results do not show any significant differences between the two subsamples. They provide contrasting arguments to predict different ethical behaviours of family versus non-family firms. On one hand, there is emphasis on the typical long-term orientation of family firms (e.g. Dyer 2003; Zellweger 2007), which is likely to result in more ethical behaviour (Long and Mathews 2011). This is also a consequence of the family-centred goals and values that characterize family firms, which further contribute to influencing their strategic decisions (Chrisman et al. 2012; Kotlar and De Massis 2013). On the other hand, the frequently unfair hiring strategies of family firms are acknowledged to be associated with less ethical behaviour (Donckels and Frohlich 1991). In this regard, family business research shows family firms as less progressive than non-family firms in terms of human resource issues such as employee involvement or adequate staffing (Aldrich and Langton 1998; Colombo et al. 2014; de Kok et al. 2006; Reid and Adams 2001). The authors, therefore, highlight how different family firms characteristics affect their business ethics and, thus, their behaviour towards stakeholders. A further study by Dyer and Whetten (2006) on a sample of S&P 500 firms demonstrates that family and non-family firms are not significantly different in relation to positive initiatives towards employees, society and the environment. Nevertheless, family firms are found to be more cautious in forestalling social concerns about potential damage to society constituents, thus, proving to be attentive to their reputation and to be legitimized in their local community. Finally, Berrone et al. (2010) perform a longitudinal comparative study on 194 U.S. firms required to report their pollution emissions. They rest on the preservation of socio-emotional wealth as the leading theoretical argument to explain why family controlled firms are more likely to respond to institutional pressures and are more compliant with environmental standards than their non-family counterparts. Highly related to the concept of socio-emotional wealth (Berrone et al. 2012; Gómez-Mejía et al. 2001), which is a non-economic utility and complementary to economic wealth (Gómez-Mejía et al. 2007), is the conviction that family firm behaviour is simultaneously driven by economic and non-economic goals (e.g. Kotlar et al. 2013).

Overall, this brief overview of past research on CSR in family firms highlights the need for a study on CSR reporting in family versus non-family firms. To our best knowledge, no study on this topic has been conducted to date. What is more, as the stream of the literature on CSR in family business has highlighted, institutional pressures, the search for legitimization and the willingness to establish long-standing relationships with stakeholders drive the socially responsible behaviour of family firms and may also represent relevant interpretations for potential differences in their CSR reporting activity. Moreover, CSR reporting is “the source of the major advances in our understanding of CSR” (Gray et al. 1995a, p. 48), because firms institutionalize CSR within the organization through reporting (Reynolds and Yuthas 2008; Schultz and Wehmeier 2010). Thus, a study on CSR reporting in family firms would also allow to contribute to the family business literature by extending, complementing and enriching our current knowledge on CSR in family firms. In the next section we present the method used in this study.

Method

Our aim is to build theory in the area of CSR reporting, demonstrating and understanding the differences between family and non-family firms in the disclosure of information on their corporate social responsibility actions. We use a grounded theory approach (Glaser and Strauss 1967) to identify the factors that affect the type and content of CSR reports disseminated by firms. Grounded theory uses an inductive approach to generate theory for unexplored research topics (Charmaz 2006; Hodkinson 2008; Locke 2001) and is based on an iterative process referred to as ‘double back steps’ where collected data are coded and recoded following a comparative analysis (Glaser 1978).

More specifically, we use a content-analytic procedure to collect relevant information. Content analysis is a method of codifying written text into various groups or categories based on selected criteria (Krippendorff 1969, 1980; Unerman 2000). To achieve this, we applied the so-called ‘third party approach’ where the analysis is carried out by someone who is neither the provider nor the receiver of the report (Gamerschlag et al. 2011). Content analysis is a highly utilized means of obtaining otherwise unavailable information (e.g. Kabanoff et al. 1995) as it tends to avoid recall biases typical of interviews (Barr et al. 1992) and because it generally affords greater reliability and replicability (Potter and Levine-Donnerstein 1999).

Sampling criteria and context

We first focus on firms with a strong commitment to CSR principles and actions. Having selected companies that consider CSR a critical determinant of their competitive advantage, the differences in the disclosure of information on CSR actions due to the heterogeneity of the strategic relevance assigned to CSR cannot be misinterpreted. We, thus, started by drawing up a list of firms included in well-known CSR networks. With the help of the Corporate Responsibility Programme Manager of SCS Consulting, a key consultancy in Italy in the CSR field, five main data sources of potentially interesting firms were identified: (i) Accountability Rating¹; (ii) CSR Manager Network²; (iii) Sodalitas Foundation³; (iv) ImprontaEtica⁴ and (v) National Report of Good practices on CSR.⁵

Second, we followed recommendations to sample broadly in order to generate novel and theoretically grounded insights (Glaser and Strauss 1967) and focused our study on large- and medium-sized enterprises. This choice was dictated by the evidence that large- and medium-sized companies have considerably more stakeholders demanding information than small- and micro-sized firms, and are also acknowledged as having a greater impact on society (Young and Marais 2011). By excluding small- and micro-sized firms, we also reduced the risk of unobserved heterogeneity due to marked differences in the size of family and non-family firms. Indeed, larger firms are acknowledged to have a number of traits that foster CSR communication and reporting (Baumann-Pauly et al. 2013).

Finally, we clustered the firms to gain insight into the differences in CSR reporting between family and non-family firms. Consistent with several previous studies (e.g. De Massis et al. 2013b; Westhead et al. 2001), we adopted two definitional criteria to identify family firms: self-identification as a family firm and family involvement in ownership and management. We examined the corporate websites to establish whether the firms self-identified themselves as a family business. Thereafter, we assessed the ownership structure and management composition on AIDA, the Bureau van Dijk database, which includes financial, demographic and commercial information on Italian firms. We identified familial relations among shareholders from their family name(s) (e.g. De Massis et al. 2013c, d, in press; Gómez-Mejía et al. 2001) and operationally considered a company as a family firm when at least two members with the same family name owned at least 10 % of equity shares. In addition, we verified whether at least one member of the same family, as identified by the family name, had a role in the firm’s top management team. This criterion accounts for the firm’s ability to behave as a family firm. The adoption of the aforementioned criteria enabled considering both family willingness and family ability to influence firm behaviour, which are acknowledged as two necessary conditions for family firms to behave differently from non-family firms (De Massis et al. 2014b).

Data

CSR reporting can take many forms but most commonly refers to the dissemination of information reported in either annual reports and accounts packages (including both voluntary and mandatory information) or stand-alone reports, which are usually, but not always, voluntary (Bebbington et al. 2008). To classify a document as a CSR report, we referred to the Global Reporting Initiative (2000; 2006) guidelines which define a corporate social responsibility report as an organizational document that provides information on the economic, environmental, social and governance actions of a firm. We, therefore, considered the documents through which organizations communicate their principles of conduct, their relationships with stakeholders and their actions towards the surrounding society and environment. We collected information and documents on CSR directly from their corporate websites (Esrock and Leichty 1998; Maignan and Ralston 2002; Morhardt 2010), which is consistent with the fact that corporations increasingly turn to the Internet to communicate their principles and to disseminate information to a much wider public (Snider et al. 2003).

To define the set of documents to be considered in the content analysis, we first performed a pilot study on four Italian firms listed in the 2011 Global Fortune 500 ranking that fulfilled the following criteria: headquarters in Italy, large size, inclusion in at least one of the five aforementioned CSR networks and belonging to four different industries. This preliminary research phase led us to identifying the following types of CSR reports: (i) Social (or Sustainability) Report, defined as a report of economic, environmental and social actions; (ii) Code of Ethics, a written set of guidelines issued by an organization to its workers and management to help them conduct their actions; (iii) Code of Conduct, the document that defines the platform of the acceptable set of behaviours within the company whose main objective is to promote a higher standard of practices within the organization; (iv) Stakeholder Map, a document that formalizes stakeholder engagement in the organization and graphically shows the groups involved in public dialogue with the company; and (v) Environmental Report, which provides information on the impact of an organization’s business activities on the environment. Only one firm in the pilot sample issued a Stakeholder Map or Environmental Report, whereas all four firms disseminated CSR-related topics within the other three types of CSR reports. To ensure comparability in the content analysis, Sustainability Reports, Codes of Ethics, Codes of Conduct⁶ as well as the sections of corporate websites dedicated to disclosing CSR actions were considered. In addition, we found that some of the firms in the pilot study had established corporate foundations to undertake socially responsible and philanthropic activities. We, thus, decided to also account for the presence of a corporate foundation as a further proxy of the firm’s commitment to disclosing its socially responsible initiatives. Corporate foundations not only play an increasingly important role in supporting firms in their social mission but also have been gaining increasing attention from academic research (Lungeanu and Ward 2012).

Data Analysis

Before starting the content analysis, the two authors independently conducted an in-depth analysis of the CSR reports issued by each of the four pilot study firms in order to develop the coding scheme.⁷ We used thematic content analysis to categorize the content of CSR reports, which is a technique whereby categories for analysis are identified in a predetermined coding scheme (e.g. Phillips 1994; Short and Palmer 2003). Content analysis can be broken down by word, clause, sentence, paragraph or page to assess the meaning of a document (Weber 1985). Using each CSR report or dedicated website section as a unit of analysis, the two authors analysed these clause by clause to capture multiple codes embedded within a single sentence. We identified keywords and topics (henceforth codes) related to CSR to enable creating an objective categorization for the content of each report. Since we followed a multiple coder procedure (Barr et al. 1992), we assessed its reliability through the evaluation of the interrater agreement, which was 90 % on completion of our pilot study. The authors then independently analysed the different reports issued by the firms that fulfilled the sampling criteria defined in the previous section. Each clause was placed into one of the categories as previously identified according to the topic; for example, the clause “Honesty, sincerity in relationships and the wisdom of farmers were the principles that inspired the establishment and growth of what is now a modern, concrete and courageous entrepreneurial reality: our Box Marche” (cf. Box Marche’s Social Report, p. 11) refers to the code ‘Honesty, fairness, integrity, respect’ which is included in the ‘Values and general interests’ topic of our categorization. However, once the coding process began, any passages where a coder had doubts (which only occurred in less than 10 % of instances) were discussed between the two authors, until agreement was reached on the appropriate code (disagreement resolution in four cases required additional coding from a research assistant who read the report and assigned the coding). In sum, the content-analytic procedure was performed by adapting the seminal content analysis methodology of Neuendorf (2002) following four stages: (i) we inductively generated first-order codes from reading the pilot study cases, integrating them with insights from existing theoretical perspectives and empirical studies, adding codes when new facts or factors emerged in the reports (Tsui-Auch 2004); (ii) when codes were named, and categories were constructed, we reviewed the data for the other companies in the sample to see which, if any, fitted each category. At times, the data did not fit well into a category, which led to either abandoning or revising the category (Pratt et al. 2006). The final list of codes served as organizing devices to recode the documentary data; (iii) based on the final interpretive results, we conducted within-case analysis and cross-case analysis on the entire sample (Eisenhardt 1989): the within-case analysis enabled familiarization with each case as a stand-alone entity prior to any generalization, while the cross-case analysis enabled comparing and contrasting cases; (iv) the fine-tuning of codes and categories during the analysis continued until saturation; we abandoned or modified tentative hypotheses and iterated between data and theoretical perspectives, until a stage of theoretical saturation was reached (Eisenhardt 1989; Glaser and Strauss 1967; Pratt et al. 2006). Although guidelines to determine non-probabilistic sample sizes are virtually nonexistent (Guest et al. 2006), we ended when no new information or themes were observed in the data, which resulted in a sample of 24 family firms and 74 non-family enterprises. This sample is representative of the Italian population, where family firms account for around 25 % of large firms according to the statistics provided by AIdAF (Italian Family Firms Association) and by the Italian Ministry of Labour and Social Policies in its national document on good CSR practices.⁸

The final sample includes 51 out of 98 companies listed on the Italian Stock Exchange (seven are family firms), 75 firms are large sized (14 are family firms), and according to the first-digit US SIC code, 33 companies (21 are family firms) operate in the manufacturing industry. Within the non-manufacturing industry, firms operate in the construction, finance, services, transport and utilities industries. Detailed information on the distribution of family and non-family firms is reported in Table 1.

Table 1 Distribution of the no. of sample firms (N = 98) by firm category and CSR document issued

The findings of the content analysis were then subjected to statistical inference. In order to answer our research questions, the content categories were used as the object of the statistical analysis, and we performed Chi squared tests between the groups of family and non-family firms; we also controlled for size, listing and industry in order to verify whether these variables affect the findings of the study. The Chi squared test allows assessing whether paired observations on two groups of firms are independent of each other, e.g. if family and non-family firms differ in reporting frequency. The choice of this statistical procedure was dictated by the type of data collected that consist of the observations of occurrences of codes within the CSR reports (Abraham and Moitra 2001; Llena et al. 2007).

In the following section, we illustrate and discuss the results of our analysis and show that family firms differ from non-family firms in the type and content of CSR reports issued.

Results and Discussion

In this section, we show the findings of our analysis of CSR reporting in Italian firms. First, we offer insights into the types of CSR reports disseminated, providing figures on the diffusion of different types of CSR reports, the establishment of foundations dedicated to corporate social responsibility as well as compliance with CSR standards. We specifically focus on the differences that emerged between family and non-family firms. Thereafter, we offer insights into the variety of topics that emerged from the content analysis of the CSR reports and provide statistical inference to identify significant differences in the reported topics between family and non-family firms, with additional controls for size, listing and industry.

Types of CSR Reports

Diffusion of CSR Reports and Establishment of Foundations

Table 1 reports the main figures relating to the diffusion of the types of CSR reports and establishment of foundations among the sampled firms.

All 98 companies issued a sustainability report, whereas 96 firms (98 %) published a code of ethics and 87 firms (89 %) provided a dedicated CSR section on their websites. We also considered the presence of other CSR documents, i.e. environmental reports, sustainability policies and CSR indicator reports. We found that 13 (13 %) of the 98 firms in the sample produced an environmental report; 19 (19 %) disclosed their sustainability policies with dedicated documents and 4 (5 %) purposefully created ad hoc reports to communicate their CSR performance. Several firms disseminated additional types of CSR reports such as (i) work-life balance principles (SACE), (ii) charter of equal opportunities (BPM), (iii) quality culture report (De Cecco), (iv) other reports based on corporate case studies on social issues (Saipem), (v) environmental policies characterizing the firm’s actions (Box Marche), (vi) documents dedicated to human care (Loccioni) and (vii) charter of services provided (Came Cancelli Automatici).

With regard to the distinction between family and non-family firms, the former were found to place more emphasis on creating a section dedicated to disclosing CSR information on their corporate websites. Indeed, as shown in Table 1, 23 out of 24 family firms (96 %) and 64 out of 74 non-family firms (86 %) have a CSR website section. Family firms are also more likely to establish foundations (50 % of family firms vs. 42 % of non-family firms have established a foundation). Finally, beyond the code of ethics and sustainability report, family firms are more inclined to publish supplementary CSR documents such as environmental reports as it is clear when considering that 15 out of 24 family firms (63 %) provided additional CSR reports, which is higher than the proportion of ‘other CSR documents’ issued by their non-family counterparts (37 %, reported by 27 out of 74 non-family enterprises).

This higher propensity of family firms to disclose CSR information through a dedicated section on their websites, to communicate the establishment of foundations and to disseminate a wider range of CSR reports can be interpreted in light of the greater importance that family firms attach to the actions that affect their reputation and that foster dialogue with their external stakeholders. These types of reporting activities refer to ‘explicit’ CSR, i.e. the set of CSR programs and strategies that firms undertake on a voluntary basis to enhance their external reputation of engaging in social responsibility. Since the company name often includes the name of family members, and their wealth is tied to their firms, family members are more likely to invest in CSR initiatives that can be easily disseminated and used to build a good reputation in the eyes of their external stakeholders (Dyer and Whetten 2006; Whetten and Mackey 2005; Wiklund 2006). This is also consistent with studies showing that family firms tend to promote their visibility and family reputation within their social environment (Dunn 1996) since they are well known within their communities (Block and Wagner 2010).

Strong pressure is felt by family firms from local communities that, even informally, prompt them to generate and disclose documents to dialogue with and meet the expectations of stakeholders (Uhlaner et al. 2004); family firms are, thus, legitimized to operate in their market. This ‘explicit’ CSR reporting is considered an activity that affects society’s perception of the firm (Deegan et al. 2000; Hooghiemstra 2000). Hence, we propose:

P1

Family firms are more likely than non-family firms to proactively and voluntarily provide a wider range of stand-alone CSR reports, to disclose CSR information through a dedicated website section and to communicate the establishment of foundations since these ‘explicit’ CSR reporting initiatives enable them to meet the informal expectations of proximate external stakeholders, reflecting the typically higher attention they pay to promoting their visibility and family reputation, and, thus, increase their legitimacy in society.

Finally, we can consider that family firms are usually characterized by a longer tenure of chairpersons (Jorissen et al. 2005); therefore, a more continuous engagement in communicating the CSR actions is expected since changes in chairpersonship may change engagement in CSR reporting (Campbell 2000).

It is, however, interesting to investigate how distributional differences along specific dimensions such as industry, firm size and listing status may affect our results. This can be done by controlling for these dimensions when comparing family and non-family firms. For example, focusing only on the manufacturing industry, family firms are found to disclose code of ethics and additional CSR documents in a higher proportion than non-family firms (respectively, 100 % of family firms versus 92 % of non-family firms as regards code of ethics, and 24 % of family firms versus 6 % of non-family firms as regards other CSR documents), whereas there are no marked differences between family and non-family firms as regards the publication of a sustainability report and a website section dedicated to CSR. Similar findings can be found if we focus on the two subsamples of large-sized firms and listed companies. On a different level of analysis, it is also interesting to show, for each of the dimensions introduced as control variables in the study, the results emerging from comparisons within the family and non-family firm subsamples. The number of listed non-family companies disclosing CSR information through various types of reports is higher than that of non-listed non-family firms. This can be explained by considering that listed companies are usually under the lens of multiple stakeholders, especially dominant shareholders (Mroczkowski and Tanewski 2007) and financial analysts (Boubaker and Labégorre 2008). Conversely, in case of family firms, the previous prediction is reversed with a high number of non-listed companies providing additional CSR reports and establishing a foundation. This can be explained by considering that the ownership and governance structures of non-listed family firms are generally composed of a higher number of family members than those of listed family firms (Boubaker and Labégorre 2008). Accordingly, a non-listed family firm tends to be more willing and able to establish its own foundation, which usually takes the name of the founder and is directly managed by a family board (Lungeanu and Ward 2012). Likewise, large-sized firms in both the family and non-family subsamples are more likely to publish additional reports than medium-sized firms since larger firms manage a more extensive range of activities and have more resources to invest in communication (Baumann-Pauly et al. 2013; Young and Marais 2011). Finally, with regard to industry clusters, firms in the manufacturing industry are more inclined to disclose information through a dedicated CSR section of their website and to establish foundations with respect to firms in non-manufacturing industries since firms that operate in the manufacturing industry are usually subject to greater industrial and environmental pressure especially in terms of their pollution activities (Chen and Bouvain 2009).

Compliance with CSR Standards

From the analysis of firm compliance with international CSR standards emerged that 26 (27 %) out of 98 firms followed International Labour Standards as defined by the International Labour Organization (ILO), 13 (13 %) declared their adherence to the World Business Council for Sustainable Development (WBCSD), and to the World Resources Institute (WRI), 38 (39 %) followed Global Compact principles, while 39 (40 %) took part in the Carbon Disclosure Project. We found that 61 (62 %) out of 98 firms declared following the Global Reporting Initiative (GRI) guidelines.

With regard to the comparison between family and non-family firms, the former show a lower propensity towards compliance with CSR standards than the latter. As the figures in Table 2 suggest, only three out of 24 (12 %) family firms followed International Labour Standards, while 23 out of 74 (31 %) non-family firms complied with this type of standard. Similarly, 21 % of family firms vs. 45 % of non-family firms followed Global Compact principles. A lower percentage of family firms (8 %) declared their adherence to WBCSD and WRI than their non-family counterparts (15 %). In terms of GRI guidelines, family firms (38 %) were found to follow these less than non-family firms (70 %). Finally, only 17 % of family firms stated participating in the Carbon Disclosure Project, while 47 % of non-family firms declared this within their reports.

Table 2 Distribution of the no. of sample firms compliant with CSR standards by firm category and CSR standard

Thus, regardless of the higher number of types of CSR reports issued by firms with family influence (Proposition 1), family firms are found to be less compliant with CSR standards. Although the dissemination of a wide range of CSR reports is relevant to family firms to dialogue with their stakeholders and to meet their specific expectations, compliance with CSR standards does not seem to be as important as it is for non-family businesses. This finding can be understood when considering the meaning of compliance with CSR standards. Compliance with CSR standards entails conforming to external norms and rules; reporting this kind of information, therefore, merely implies demonstrating corporate efforts towards that which constitutes ‘implicit’ CSR, i.e. the firm role within institutions, consisting of all requirements that the companies must passively satisfy (Matten and Moon 2008). Family firms are acknowledged to be less dependent on their institutional context (Dunn 1996; Jaggi et al. 2009) and are less subject to regulatory and normative pressures (Dyer and Whetten 2006), which may explain why on average they devote less attention than non-family firms to complying with CSR standards. Moreover, compliance with CSR standards requires formalizing business practices and a high level of bureaucracy. However, family firms are acknowledged as business organizations whose agents tend to be exempt from bureaucratic constraints and strictly formalized business practices that limit managerial authority and inhibit ownership priorities and firm autonomy (Carney 2005; Massis et al. 2013). In view of the foregoing, we propose:

P2

Family firms are less compliant with CSR standards than non-family firms since these ‘implicit’ CSR reporting initiatives, which entail passively satisfying requirements to be considered compliant with institutional norms and rules, contrast with their more autonomous nature and lower dependency on the institutional context.

Also in this case, we controlled for specific dimensions such as industry, firm size and listing status in order to investigate how distributional differences along these dimensions may affect our results. Analyses on the subsamples identified by isolating manufacturing, non-manufacturing, listed, non-listed, large, and medium-sized firms lead to results that are consistent with those found in the main analysis. Listed firms show higher levels of compliance with CSR standards in both family and non-family firms, and the difference with respect to non-listed firms is particularly evident for ILO, Global Compact principles and Carbon Disclosure Project. This can be understood by considering that publicly traded companies are more eager to emphasize their CSR commitment and have more duties towards their shareholders (Singhvi and Desai 1971). Firm size is positively related to the propensity to comply with international standards, which can be understood by considering that medium-sized enterprises compared to large-sized firms are less subject to pressure from stakeholders and the media in terms of aligning with universally recognized standards (Young and Marais 2011).

Content of CSR Reports

The contents of CSR reports are shown and discussed in this section. In particular, we show the topics that emerged from our content analysis drawing on some exemplar quotes extracted from the analysed reports. We then discuss each CSR topic in detail in light of the statistical tests performed to assess the significance of the differences identified between family and non-family firms, and to control for the size, industry and listing variables. Table 3 provides some illustrative examples of the CSR topics disclosed by the sample firms as emerged from our content analysis.

Table 3 Evidence of the content analysis by CSR topic

The method adopted provides a wealth of information as well as details on the CSR topics including the aim of firms to manage the trade-off between serving their own needs, those of their public of interest and obligations to humankind. This is consistent with Snider et al.’s (2003) results of the analysis of the CSR sections of the websites of large corporations.

Table 4 provides a synoptic view of the occurrences of the content of CSR reports and the results of the statistical tests, enabling a straightforward comparison and analysis.

Table 4 Synoptic representation of the content of CSR reports and the results of statistical tests

A discussion on the content of CSR reports for each single topic identified in the content analysis is offered hereafter.

Values and General Interests

The topic ‘Values and general interests’ refers to what is considered good, important, useful and desirable; this topic includes concepts such as honesty, fairness, integrity, respect, gender equality, support for cultural and sporting activities.

The results of the Chi squared test show a significant difference between family and non-family firms. Non-family firms were found to more frequently communicate issues related to this topic. This can be explained by considering that family firms typically state and share their values informally, predominantly within the firm, as they are generally acknowledged to be protective of their value systems (Dunn 1996; Miller et al. 2008). In terms of attitudes towards their institutional context, family firms are usually less inclined to publicize the cultural elements that characterize their behaviour (Le Breton-Miller and Miller 2006). This is consistent with the tendency of family firms to preserve their socio-emotional wealth (Berrone et al. 2012; Gómez-Mejía et al. 2007, 2011), which is more salient in the presence of highly emphatic family members (Goel et al. 2012). Moreover, the family council is often the privileged locus to communicate the values within the firm (Carlock and Ward 2001; Davis and Harveston 1998), and this may explain why family firms are less likely to make this type of communication public. Values are considered part of the family legacy that is internally developed and transmitted.

Listed non-family firms disclosed a significantly higher amount of information than non-listed non-family firms in relation to values and general interests, whereas Chi squared tests for listed versus non-listed firms resulted in a non-significant difference between listed and non-listed family firms. This evidence further highlights the significant role of family involvement in the level of disclosure of information related to family culture. The controls for size showed that, in both family and non-family firms, large-sized firms are more concerned about communicating CSR values and interests than medium-sized companies. With regard to industries, neither family nor non-family firms are significantly affected by the industry variable in the level of information disclosed on this topic.

Shareholders

The codes here are shareholder value creation, consideration of their points of view, attention to their interests, equal treatment of different shareholder categories, and clearness and honesty in communications.

Family and non-family firm CSR reports were found to significantly differ with respect to this topic; specifically, non-family firms disclosed more information on this topic. This can be explained by considering the high relevance of shareholder claims in non-family firms; in particular, non-family businesses aim at maximizing return on equity for shareholders as their primary goal. Conversely, family firm owners are generally concerned with the pursuit of non-economic goals (Chrisman et al. 2012; Kotlar and De Massis 2013), and shareholders may, thus, be less relevant than in the case of non-family firms. Put differently, the existence of family-centred non-economic goals as a consequence of the interplay between the family and business systems (Zellweger et al. 2012), renders shareholders less relevant to family firms when compared to non-family firms, and this is reflected in a lower emphasis on disclosing topics related to shareholders in their CSR reports.

In non-family firms, we found that listed firms were more concerned about reporting on this topic than non-listed firms, and the same result emerged from the Chi squared test between large- and medium-sized firms, in favour of the former. Conversely, in family firms, listing and size are not significant variables in explaining the disclosure of topics related to shareholders. Finally, in both family and non-family firms, companies operating in manufacturing and non-manufacturing industries did not significantly differ in the disclosure of these codes.

Employees

Good working conditions, the involvement of employees in business strategies, sharing business culture, safety, training and skills development, non-discrimination and symmetrical treatment of human resources are just some of the codes considered essential in the relationship with employees.

Family firms were found to disseminate less information on these types of issues than their non-family counterparts. Family firms build their relationships with employees based on normative commitment rather than on financial performance motivations (Stavrou et al. 2007). Employees are considered as valuable assets rooted in the family resource endowment, so that family firms are aware of their importance and, therefore, do not sacrifice them for short-term returns (Donckels and Frohlich 1991; Miller and Le Breton-Miller 2007). In addition, employees are often considered as part of the family (Uhlaner et al. 2004; Ward 1988) rather than as external parties to be involved in communication exchanges. This is consistent with the family business embeddedness perspective (Khanin et al. 2012), which foresees a high degree of alignment between the values and objectives of employees and those of the family firm (Tagiuri and Davis 1996). Put differently, family firms tend to develop highly intertwined relationships with their employees to the point of overshadowing any issues on employee-related topics. To the contrary, non-family firms are incentivized to develop full sections in their CSR reports on these topics. Employees are considered important stakeholders whom non-family firms are inclined to protect and maintain constant dialogue with to communicate the objectivity and fairness of critical aspects such as the working place, health and safety, training and careers, to provide some examples.

Listed and non-listed family firms do not significantly differ in the disclosure of this topic, while for non-family firms listing on the capital, market plays a significant role in predicting the tendency to disclose information on issues related to the management of employees. Large-sized firms are more attentive to employees since these firms are generally under the lens of institutions and the media (Young and Marais 2011). Industry does not appear to affect the statistical results that emerged from the study on this topic.

Customers

The codes related to the ‘Customers’ topic include aspects such as satisfying customer expectations, creating customer loyalty, customer involvement in corporate social responsibility activities and objectives, ethical advertising and fair prices. Here the Chi squared test revealed a significant difference between family and non-family firms, as shown in Table 4. Consistent with prior literature (Gray et al. 1987), customers are one of the most important stakeholder categories and the non-family firms in our sample tended to be more inclined to disclosing information on customer-related CSR issues. This evidence can be explained by considering the differences between family and non-family firms in their logic to approaching customers. Contrary to non-family firms, which have a propensity to dialogue with their customers in order to meet institutional pressures and gain legitimacy in the market, family firms tend to build personal and informal relationships with their customers (Brokaw 1992; Gómez-Mejía et al. 2001; Lyman 1991; Uhlaner et al. 2004). Positive customer perceptions of family ownership and relationship-based business interactions with customers create stakeholder efficiencies (Aronoff and Ward 1995; Dyer and Mortensen 2005). Family firm customers are, therefore, less exposed to formal communication flows. This provides some rationale for why family firms may have less interest in communicating issues in relation to their customer relationships, thus, explaining the scarcity of information on this topic in the reports analysed.

In non-family firms, we found that listed and large firms were more concerned about reporting on this topic than their non-listed and medium-sized counterparts. Conversely, in family firms, listing and size do not affect the disclosure of customer related topics. Finally, in both family and non-family firms, companies operating in manufacturing and non-manufacturing industries did not significantly differ in the disclosure of these codes.

In view of the foregoing, we propose the following:

P3

The CSR reports disseminated by family firms are less focused on topics related to ‘Values and general issues’, ‘Shareholders’, ‘Employees’, and ‘Customers’ since family firms tend to informally convey their values within the firm and protect themselves against information leaks on their value system, are driven by both economic and non-economic goals in assessing their relationships with shareholders, consider employees as part of the family system rather than as external parties to be involved in communication exchanges, and build informal relationships with customers that are, therefore, less subject to formal communication flows.

Environmental and Green Issues

This topic relates to concerns for environmental conservation and improving the state of the natural environment including issues such as responsible use of energy and material resources, reduction of pollution emissions, green and sustainable research, design and innovation.

The results of the Chi squared test show a significant difference between family and non-family firms. Family firms were found to disclose more information on green issues than non-family firms as the environment is a key stakeholder in this type of business organization with strong family influence on pursuing proactive environmental strategies (Sharma and Sharma 2011). This is consistent with existing evidence showing that family firms react more to environmental pressures since the concern for environmental issues offers rewards in terms of non-economic goals through which they increase their socio-emotional wealth (Berrone et al. 2010). Moreover, family firms tend to pay particular attention to maintaining firm reputation since the family name is strongly associated with the firm name, and to avoiding potentially negative outcomes arising from the loss of reputation in relation to the business and the family (Dunn 1996; Godfrey 2005). The higher emphasis given by family firms to communicating environmental issues can also be interpreted by considering that family firms are more likely to institute environmentally friendly policies than non-family firms since natural environment policies are more positively associated with innovation and greater financial performance (Craig and Dibrell 2006). This finding is also in line with our previous results on the types of CSR reports issued whereby family firms disclose a higher number of additional documents than non-family firms especially with regard to the environment.

As to the control variables, listed and large-sized non-family firms disclose environmental and green issues more than, respectively, non-listed and medium-sized counterparts. This can be explained by considering that listed and large firms are monitored to a greater extent by public opinion which is very sensitive to environmental issues (Reverte 2009). Moreover, firms in the manufacturing industry significantly differ in the amount of environmental codes disclosed with respect to those operating in non-manufacturing industries. The manufacturing industry has the greatest impact on the environment, and the firms in this industry are, therefore, more active in communicating how they deal with environmental concerns. In family firms, no significant difference emerged between listed and non-listed firms, large- and medium-sized firms, companies in manufacturing and non-manufacturing industries. Finally, a more fine-grained comparison between family and non-family firms highlights that, within the subsample of listed firms, family firms on average disclose less information on environmental issues than their non-family counterparts. This difference can be explained by considering that the family firms in this subsample are characterized by dilution of family ownership, and this weakens the effect of family influence on the business and hence deteriorates the drivers that would lead family firms to disclose more than non-family ones on this topic.

Philanthropy

Within this CSR topic, the codes include respect for local communities, their rights and specific customs, engaging in projects for the quality of local community life, engaging in projects for the social and economical development of the local community, donations, charity and sponsorships to support the social initiatives of local communities, attention to changes in the surrounding civic environment and efforts to make the local community aware of firm commitment to CSR.

The Chi squared test shows that family firms disclosed significantly more information than non-family firms on this topic. This result can be interpreted by considering that family firms are characterized by family altruism (Karra et al. 2006; Schulze et al. 2003), which leads them to engage more in philanthropic activities targeted at their local communities in order to be recognized as good corporate citizens (Dyer and Whetten 2006; Uhlaner et al. 2004). This evidence may also be explained by considering the higher long-term orientation of family firms (Zellweger 2007; Zellweger et al. 2012b), which induces them to create stable and long-lasting relationships with the communities in which they operate and to maintain a positive image (Miller and Le Breton-Miller 2005). This is consistent with the existing literature depicting family firms as more embedded in social relationships with the local community than their non-family counterparts (Aldrich and Cliff 2003; Le Breton-Miller and Miller 2009; Steier et al. 2009; Veliyath and Ramaswamy 2000), so that they are very attentive to their neighbours’ expectations.

In family firms, non-listed organizations were found to disclose more information on their philanthropic actions than listed firms. This evidence can be explained by considering that the equity share owned by the owning family in non-listed firms is generally higher (Anderson and Reeb 2003) so that attention to reputation and image within the local community is intensified in order to be legitimized in their activities. Listed and large-sized non-family firms disclose more information on philanthropy than, respectively, their non-listed and medium-sized counterparts, while companies in non-manufacturing industries are slightly more inclined to disclose information on this topic. A more fine-grained analysis within individual subsamples of firms provides additional insights for the overall results that we obtained. First, considering only firms working in the manufacturing industry, we find a higher rate of family firms disclosing information on philanthropy than non-family firms. The difference is, however, marginal, and this means that in the manufacturing industry, the propensity of non-family firms to report on philanthropic issues is more aligned to the propensity of family firms. This can be explained, for example, by considering that the production facilities of manufacturing firms are usually perceived negatively by the surrounding community, and engaging in philanthropic activities and communicating these actions is a commonly used expedient to mitigate this negative perception and strengthen their image in front of society, regardless of the family status. Second, a focus on the subsample of listed firms shows that family firms on average disclose less information on philanthropy than non-family firms. As already mentioned, listed firms are subject to the detailed scrutiny of consumers, associations, the equity market and the environment. When we consider, within the group of listed firms, family businesses, we cannot ignore that being listed weakens some of the determinants of their higher propensity to report on philanthropy. Specifically, family firms are spurred to pursue short-term payoffs to the detriment of their long-term orientation, and family altruism is more difficult to be exercised due to the fact that the family dominant coalition has a lower managerial discretion, because it is more constrained to respond to the demands of institutional actors, and family ownership is less concentrated than in the case of non-listed firms. Thus, the overall higher propensity of family firms to disclose more on philanthropy than non-family firms does not hold when we focus on the subsample of listed firms.

In view of the foregoing, we formally state:

P4

The CSR reports disseminated by family firms are more focused on topics related to ‘Environmental and green issues’ and ‘Philanthropy’ since family firms are particularly concerned about environmental pressures to protect their socio-emotional wealth and are more likely to establish environmentally friendly policies given that they have a long-term orientation and a higher degree of altruism, and are more embedded in social relationships with the local context.

Processes and Products/Services

This topic is related to production efficiency, quality guarantees and improvements, waste reduction, recovering and recycling materials, product safety, sustainability-oriented planning and sustainable supply chains.

Family and non-family firms were not found to significantly differ in the disclosure of CSR issues on processes, products or services, even if the former publish a relatively higher proportion of issues associated with this CSR topic. The lack of significant differences may be due to the recognized standardization in corporate practice in terms of communicating issues related to production and service supply (Mueller et al. 2009), which must respond to well-defined practices regardless of whether there is a family influencing firm behaviour.

In non-family firms, we found that listed companies placed more emphasis on disclosing this CSR topic than non-listed firms as well as a strong significant difference between large- and medium-sized companies in the degree of completeness of information on this CSR topic, which increased with firm size. This is reasonable if we consider that the complexity of firm processes and the supply chain increases with firm size. The Chi squared test resulted in a significant predominance of codes related to this topic in firms in the manufacturing industry; this is again reasonable when we consider the traditionally greater attention paid by manufacturing firms to the efficiency and quality of their supply chain.

General Stakeholder Management Issues

This topic includes all issues related to satisfying stakeholder claims, stakeholder dialogue, stakeholder involvement in decision-making and stakeholder management reliability.

The Chi squared test showed that family and non-family firms do not significantly differ with regard to communicating general issues on stakeholder management. In analysing the types of CSR reports, we found that family firms provided a wider variety of CSR reports to meet their stakeholder expectations. When focusing on the content of these reports, all sampled firms discussed the general issues on stakeholder management in a very similar way. This is probably due to the fact that stakeholder management refers to the high level and very general principles that each firm engaged in socially responsible actions, regardless of family influence, must declare. In non-family firms, we found that listed and large-sized firms are more concerned about reporting on this topic than their non-listed and medium-sized counterparts. To the contrary, in family firms, listing and size are not significant variables in explaining the disclosure of topics related to general stakeholder management issues. Finally, in both family and non-family firms, companies operating in manufacturing and non-manufacturing industries did not significantly differ in the disclosure of these codes.

Suppliers

The codes related to this topic deal with the assessment of supplier engagement in CSR, whether suppliers reduce waste in provisions, open communications, loyalty, fair contracts and equal opportunities in the relationship between the firm and its suppliers.

Family and non-family firms were not found to significantly differ in the Chi squared test results. What emerged from the content analysis is that, regardless of family influence on the firm, high emphasis is placed on communicating principles such as respect and achievement of the requirements that suppliers must fulfil. Indeed, all firms engaged in corporate social responsibility disclosed a large amount of information on supplier respect of CSR principles and their engagement therein.

In family firms, the Chi squared tests showed no significant difference between listed and non-listed companies, whereas medium-sized firms were found to disclose more information on suppliers than large firms. This evidence can be explained by considering that medium-sized family firms are more likely to deal with local suppliers who through their proximity become long-term partners and are engaged in a constant dialogue with the firm (e.g. Uhlaner et al. 2004). Conversely, suppliers of larger family firms are not necessarily located in the same community, are more dispersed and, therefore, tend to have less constant dialogue with the firm. In terms of industry, family firms in the manufacturing industry declared more information related to suppliers, while no significant differences between organizations in manufacturing and non-manufacturing industries were found in non-family firms. Listing and size, instead, were found to significantly predict the extent of disclosure of this topic in non-family firms, suggesting that listed and large firms declare more information than, respectively, non-listed and medium-sized firms.

Hence, we state:

P5

The CSR reports disseminated by family and non-family firms do not significantly differ in topics related to ‘Processes and products/services’, ‘General issues on stakeholder management’, and ‘Suppliers’ since the issues associated with these topics are expected and required by the social and institutional context, and all firms, regardless of family influence, are incentivized to pay attention to these in their CSR reports.

Conclusions, Implications and Limitations

Considering the ubiquity of family firms and their crucial role across all world economies as well as the importance of disclosing social and environmental actions for the competitive advantage of any firm, this article has shed new light on the important yet overlooked topic of CSR reporting in family versus non-family firms. Drawing on institutional theory and on a rich body of empirical evidence gathered through a content analysis of the CSR reports of 98 large- and medium-sized Italian firms, this study has evinced several differences between family and non-family firms in CSR report type and content disclosed. First, family firms were found to disseminate a wider range of CSR reports than non-family firms since these ‘explicit’ CSR reporting initiatives enable them to meet the informal expectations of proximate external stakeholders (Wiklund 2006). This finding provides further insights on the higher likelihood of family firms to disclose information on ‘explicit’ CSR with respect to their non-family counterparts and adds to prior non-significant results found by Dyer and Whetten (2006). Explicit CSR reporting initiatives reflect the typically higher attention paid by family firms to enhancing their visibility and family reputation (Block and Wagner 2010) and increasing their legitimacy in society (Hooghiemstra 2000). Family firms can, thus, enhance the image of both the family and the business and are legitimized in their actions, which enable them to pursue their long-term sustainability goals (Chrisman et al. 2003; Long and Mathews 2011). Family firms are characterized by unique traits due to the interplay of two institutionalized entities, the family and the business, extending the set of organizational goals and generating a number of changes in the nature and scope of the relationship between the organization and its stakeholders (Mitchell et al. 2011). Second, family firms were found to be less compliant with CSR standards than non-family firms. This has been interpreted by considering that these ‘implicit’ CSR reporting initiatives, which entail passively satisfying requirements in order to be labelled as compliant with institutional norms and rules (Matten and Moon 2008), contrast with the more autonomous nature of family firms and their lower dependence on the institutional context (Dunn 1996; Jaggi et al. 2009). Third, as concerns the content of CSR reports, our findings show that family and non-family firms place emphasis on different CSR topics. On one hand, although family firms are commonly acknowledged to devote particular attention in their everyday operations and business life to topics such as ‘Values and general interests’, ‘Shareholders’, ‘Employees’, and ‘Customers’, (e.g. et al. 2004), they report less information on these than their non-family counterparts. This evidence has been explained by considering that family firms tend to informally convey their values within the firm and to protect themselves against information leaks on their value system (Dunn 1996). They are driven by both economic and non-economic goals in assessing their relationships with shareholders (Chrisman et al. 2012; Kotlar and De Massis 2013), consider employees as part of the family system (Ward 1988) rather than as external parties to be involved in communication exchanges, build informal relationships with customers (Gómez-Mejía et al. 2001) and are, therefore, less subject to formal communication flows. On the other hand, family firms appear to disclose more information than non-family firms on topics related to ‘Environmental and green issues’ and ‘Philanthropy’. This evidence has been interpreted by considering that family firms are particularly concerned about environmental pressures in order to protect their socio-emotional wealth (Sharma and Sharma 2011) and are more likely to establish environmentally friendly policies (Craig and Dibrell 2006). This suggests that family influence has a role in linking institutional pressures and firm response, which is driven by a set of ‘socially worthy’ non-economic preferences (Berrone et al. 2010) towards a continuous search for legitimacy in its institutional context. Moreover, family firms are characterized by a long-term orientation (Miller and Le Breton-Miller 2007; Zellweger 2007), a higher degree of altruism (Schulze et al. 2003; Karra et al. 2006), and are more embedded in social relationships with the local context (Miller and Le Breton-Miller 2006), providing a rationale for the greater emphasis on ‘Philanthropy’.

The results of our study have strong implications for academics, practitioners and policy makers. First, our findings could benefit both family business and CSR scholars. Family business scholars have only recently begun to investigate the socially responsible behaviour of family firms (De Massis et al. 2012). This study provides evidence on the CSR actions conveyed in social reports that firms are called on to disclose; as such, our findings advance our understanding of the strategic aspects of social responsibility that are prioritized by family firms, and pave the way for further studies on the important issue of prioritizing stakeholder claims in family firms (Kotlar and De Massis 2013). CSR scholars could benefit from theoretically and empirically considering how family influence on a business organization could affect CSR reporting. As such, our study extends the dialogue on CSR reporting by pointing to the degree and type of family involvement as a new category of internal factors affecting the level, quality and quantity of CSR reporting. For instance, the perspectives thus far adopted to predict the CSR reporting dynamics of the relationship between the firm and its stakeholders do not consider whether willingness and ability to disseminate CSR actions differ between family and non-family firms. Our analysis identifies several theoretical reasons suggesting that the ‘family’ variable would be a relevant moderator of the relationship between a firm’s willingness or ability to embrace CSR reporting, the types of CSR reports issued and the topics therein. The involvement of family stakeholders in the organization is a unique feature of family firms (Zellweger and Nason 2008), and our study shows that this factor is likely to play an important role in explaining their distinctive CSR reporting processes. However, our findings show that family firms are on average more willing to disclose CSR reports but less compliant with CSR standards. This contrasting evidence highlights the complexity and heterogeneity of family firms in dealing with the disclosure of social issues. Thus, our findings are consistent with prior studies providing arguments for the Janus-face of family firm behaviour and suggesting that family firms are extremely heterogeneous as regards CSR (Adams et al. 1996; Deniz and Suárez 2005; Dyer and Whetten 2006) and raise important questions, e.g. concerning whether family firms are better or worse corporate citizens than non-family firms. Understanding the relationship between CSR reporting and the firm’s propensity to act responsibly as a ‘corporate citizen’, and how family involvement moderates this relationship, is an area ripe for future research. Moreover, our study further adds to the debate on accounting ethics (Gunz and McCutcheon 1998), providing useful indications on topics and issues that academics should include in their accounting courses. Second, the findings of our study would also be useful to family firm managers and consultants in the area of corporate social responsibility reporting, who are encouraged to not assume that the good practices proposed by corporate social responsibility, corporate accounting and reporting handbooks are universally applicable. They should carefully consider how family influence on the organization could affect the effectiveness of these good practices and how these should be revised to best suit the firm’s distinctive characteristics. Finally, this research could be a suitable background policy document for policy makers. Corporate social responsibility actions are the focus of increasing attention in the design of public policies and family firms, given their ubiquity (Astrachan and Shanker 2003; Anderson and Reeb 2003), are critical in the development of economies across the world (Villalonga and Amit 2006; La Porta et al. 1999). In this regard, the results of our research are especially useful as they provide suggestions on how to build a system of supporting initiatives for CSR reporting that fits the idiosyncratic characteristics of family firms. For example, the findings of our study could provide support for policy makers in decisions on the type of CSR reports to be included in mandatory disclosure. Policy makers are also invited to develop adequate incentive systems to foster CSR reporting as a competitive lever in the current business context, with the twofold benefit for family firms to achieve, through CSR reporting, both a superior competitive advantage, and a legitimization of their status as good corporate citizens in their community (Hooghiemstra 2000), reinforcing in this way their image and reputation.

Like all studies, our work has some limitations, which, however, provide opportunities for future research. First, our final sample includes large- and medium-sized enterprises, while a study on small- and micro-sized enterprises would be useful to understand how the typical features associated with smaller firm size (e.g. differences in resource availability or in institutional constraints) could affect our findings. Second, we only collected information disclosed through corporate websites although CSR information is generally also disseminated through other communication channels such as newspapers and other media; these alternative channels could be considered in future research to establish whether our predictions hold. Third, our analysis is based on data gathered from secondary sources, which may have a low level of accuracy, especially with regard to the family involvement measures. A survey would be useful to obtain more precise and complete information on each company. Ideally, such studies would be large-scale in order to provide further statistical tests of our results. Fourth, this study is cross-sectional and does not permit a causation analysis; longitudinal studies would ensure capturing the cause-effect and temporal relationships between the type of CSR report and content disclosed and their antecedents. Moreover, firm behaviour may change as the firm ages and goes through different generations of family control (De Massis et al. 2014a). Likewise, changes in chairpersonship may vary engagement in CSR reporting (Campbell 2000). However, we did not collect information on the generation in control or on the occurrence of generational transitions in the family firms in our sample. We hope that future scholars will conduct longitudinal studies aiming to investigate the roles played by different stages of the family firm life cycle and by the succession event on CSR reporting. Fifth, our sample is drawn from one country (Italy), and we encourage scholars to add evidence on CSR reporting in family versus non-family firms from other countries to ensure the relationships found are not linked to Italian institutional or cultural variables. For instance, CSR reporting may be specifically bound to cultural contingencies. Important differences across natural cultures, for example, social collectivism versus individualism (Earley 1989) and perceptions of social distance (Akerlof 1997), may inform institutional perspectives on how the effect of family influence on CSR reporting could vary in firms from different regions. Moreover, each country is characterized by its own set of regulatory norms and constraints that the firms have to be aware of and accomplish in order to avoid sanctions (e.g. regulatory policies to ensure fair practices, workplace safety, respect of the environment, product quality). We, therefore, encourage others to conduct cross-country comparative studies in order to shed new light on the effect of national regulations on the type and content of CSR reports disclosed.

In addition to studies that investigate the implications of our work for the family business and CSR literature or that overcome our study’s limitations by extending the scope of inquiry in terms of sampling frames, variables or research methods, further research directions emanate from our findings. Better social performance is associated with an increase in the level and quality of CSR reporting (Ullman 1985), while family involvement in ownership and governance has been found to affect the firm’s social performance (Bingham et al. 2011). Exploring how the type of CSR report and content disclosed affect social and economic performance and how this relationship is moderated by family influence constitutes a promising avenue for future research. Specifically, it would be interesting to establish whether the changes in the type of CSR report and content disclosed by family versus non-family firms reflect and explain their social and economic performance trends. Future academic work in this direction would complement the contributions of our study by offering new insights into the entire chain of causality between family influence, CSR reporting and firm social and economic performance. Moreover, our study is focused on CSR reporting, which refers to the process of disclosing the social and environmental actions of organizations to particular interest groups. Although such disclosure is certainly related to CSR actions, our research does not consider the effects of family influence on the firm CSR actions. Investigating the relationship between the way firms disclose their CSR actions and the socially responsible actions they undertake constitutes another avenue for future research.

It is our hope that this study will inspire future work in this new field at the intersection of CSR reporting and family business, with a strong impact on both academic theory and management practice. We have offered some initial insights into a very complex topic and would, therefore, strongly encourage others to continue this line of inquiry.