Three theories were used in this study to give varying perspectives on the relationship between board independence and dividend decisions in family firms. The agency theory is the most used theory in family firm research and promotes the appointment of outsider directors on a board, as it enhances the monitoring of managerial actions. The agency theory highlights three different types of conflicts, namely principal–agent, principal–principal and principal–creditor conflicts. Although the most cited type of agency conflict is principal–agent conflict, this is less prevalent in family firms (Al-Najjar & Kilincarslan 2016; Dinh & Calabrò 2019).

Family owners often possess more information than other owners, limiting the information asymmetry and managerial exploitation between agents and principals (Bendickson et al. 2016; Habib et al. 2019). Furthermore, family members often occupy managerial positions and may act as stewards for the firm, given their socioemotional involvement in the business (Bhattacharyya, Elston & Rondi 2014). The principal–principal conflict is largely relied on as a theoretical construct in family firm-related research (Al-Najjar & Kilincarslan 2016; Anderson & Reeb 2003; Dinh & Calabrò 2019). The type of agency conflict refers to the expropriation of minority shareholders’ wealth by controlling shareholders (Al-Najjar & Kilincarslan 2016). Using this theory as an explanatory framework is especially applicable in developing economies where investor protection is weak (Dinh & Calabrò 2019). In the context of the current study, family owners are regarded as controlling shareholders.

Djebali and Belanès (2015) aptly note that because family shareholders often have managerial ties, this may result in them prioritising their interests at the expense of minority shareholders. Family shareholders may engage in the tunnelling of assets and profits, which refers to the unfair transfer pricing between entities owned by these shareholders (Djebali & Belanès 2015). Transfer pricing refers to the practice of pricing transactions between and within firms that share common ownership or control (Kumar et al. 2021). Further, family owners may act in their own interests by lessening firm risk to preserve firm continuation at the expense of other shareholders (Bhattacharyya et al. 2014).

To address shortcomings of the agency theory to explain managerial decision-making in family firms (Molly & Michiels 2021), Gómez-Mejía et al. (2007) developed the SEW theory. Socioemotional wealth is an essential feature of family firms and is often used to differentiate them from non-family firms (Cuevas-Rodríguez et al. 2023; Hasenzagl et al. 2018). In short, decision-makers in family firms are motivated as much, or even more, by achieving non-financial (socioemotional) outcomes or gains while minimising non-financial (socioemotional) losses. When their stock of socioemotional endowment (non-financial aspects) is threatened, controlling family members may make decisions that are not necessarily financially logical but rather made with the intent to preserve their non-financial (socioemotional) wealth (Berrone, Cruz & Gomez-Mejia 2012). McCracken (2020) is of the opinion that family firms are characterised by their informal governance mechanisms and managers who are motivated by the values and non-economic objectives of the families in business, such as reputation, family image, family cohesion and power protection. Consequently, family firms attempt to preserve SEW to the best of their ability even if it comes at a financial cost to the firm (Naldi et al. 2013).

Berrone et al. (2012) added to Gómez-Mejía et al.’s (2007) original SEW research by developing a framework that describes the dimensions of SEW. This framework is known as the ‘FIBER’ framework and has been widely used by family firm scholars (Cuevas-Rodríguez et al. 2023). The five FIBER dimensions include family control and influence, identification of family members, binding social ties, emotional attachment of family members and the renewal of family bonds to the firm through dynastic succession. These dimensions represent aspects that family members regard as most important and are often prioritised.

According to the SEW theory, decisions concerning control and influence, stakeholder relationships, business venturing and corporate governance are affected by the intended preservation of SEW (Hasenzagl et al. 2018). The impact of SEW on corporate governance processes is particularly relevant in this study. For example, two of the FIBER dimensions, namely family influence and control and dynastic succession, enable family members to remain in control of the family firm by appointing family members on the board or in executive positions (Berrone et al. 2012; Naldi et al. 2013; Rudyanto 2023). Although the appointment of a family member as chief executive officer (CEO) and/or chairperson of the board may aid family firms in preserving their SEW endowments, it comes at a cost to board independence (Berrone et al. 2012; Naldi et al. 2013).

As stated earlier, dividends are also regarded as effective corporate governance mechanisms. A large number of theories have been developed to guide managers in making optimal dividend distribution decisions. One of these deemed relevant for this study was the catering theory of dividends. Scholars also found that the demand for dividends differs between shareholder groups and over time (Baker & Wurgler 2004; Shefrin 2018). The varying demand for dividends will reflect in the share price of dividend and non-dividend paying firms accordingly (Baker & Wurgler 2004).

The considered period of the study is 2006–2022. This timeline covers economic upturns and downturns, including the global financial crisis of 2008 and the coronavirus disease 2019 (COVID-19) pandemic. The research period started before the 2008 crisis to ensure that we investigate the possible impact of the pandemic (if any). Given the study’s period (2006–2022), market sentiment would have varied accordingly. Thus, it is important to consider the catering theory of dividends, in conjunction with the agency theory because family members, as managers and majority shareholders, may change their dividend decisions according to current market conditions and personal needs (Sekerci 2020). Rudyanto (2023) proposes that the COVID-19 pandemic placed considerable pressure on family firms, impacting the emotions of family managers negatively. Thus, during crises, family members in managerial positions may opt to retain earnings in the firm instead of distributing earnings as dividends to ensure they have sufficient cash reserves for emergencies, which may be detrimental to the sustainability of the firm. This decision may come at the expense of minority shareholders who may have preferred an extra source of income (in the form of cash dividends).

Dividends serve as a crucial source of income for many shareholders (Baker, Mendel & Wurgler 2016). Consistent with the agency theory, dividends are argued to reduce agency conflicts by constricting the FCF available to managers. By paying dividends, managers have less FCF that may be used for personal objectives or to pursue unprofitable projects (Djebali & Belanès 2015). There are valid arguments for a family owner’s preference for higher dividends or lower dividends. Family owners may prefer higher dividends for reasons related to their personal income needs (Isakov & Weisskopf 2015). In addition, family firms may opt to pay higher dividends to build their reputation. In doing so, they simultaneously reflect good treatment towards minority shareholders (Isakov & Weisskopf 2015).

For a different perspective, family firms may prefer lower dividends and a reinvestment of excess FCF. This preference is aligned with the aim to maximise long-term revenues and guarantee survival for future generations (Djebali & Belanès 2015). It is evident that for both preferences, the SEW theory may be used as a guide to understand dividend preferences in family firms. Two of the FIBER dimensions, namely identification of family members and strengthening of family bonds to the firm through dynastic succession, could be used as explanations for varying dividend preferences.

When reviewing over 30 research articles that investigate dividend distributions in family firms, several proxies were used to measure dividend payments and analyse the main findings of past studies. It is evident from reviewing these studies that there are mixed results regarding the nature and regularity of dividend distributions in family firms. On balance, research indicates that family firms tend to pay lower dividends than non-family firms (e.g., Briano-Turrent et al. 2020; Djebali & Belanès 2015; Duygun et al. 2018). The over-retention of earnings may thus be a problem in family firms and could be caused by dominant shareholders who are intending to extract for private benefits. For example, major shareholders may engage in the tunnelling of profits and assets and empire building (Setia-Atmaja, Tanewski & Skully 2009). Alternatively, the over-retention of earnings may be explained by firm longevity, as discussed in the preceding paragraph (Djebali & Belanès 2015). Through receiving lower dividends, minority shareholders in family firms are considered severely expropriated (Djebali & Belanès 2015). This problem is also called as the principal–principal agency conflict.

One way of addressing the over-retention of earnings in family firms, which appears to be a problem, is the existence of strong internal governance controls. In this study, emphasis was placed on board independence. In the following section, an overview will be presented of studies that investigated the relationship between dividend distributions in family firms and board independence in these firms.

Board independence has been considered the most important corporate governance mechanism to reduce agency problems (Siebels & Knyphausen-Aufseß 2012). According to King IV, the most recent and widely used guideline for corporate governance in South Africa, boards should be composed of 50% or more non-executive directors, of which most should be independent (Institute of Directors in South Africa [IoDSA] 2016). To maintain the firm’s interest, it is crucial to appoint INEDs on the board. It is argued that these directors, who are also called outside or external directors, are vital for a board to be an effective monitoring mechanism (Duru, Iyengar & Zampelli 2016; Muniandy & Hillier 2015). Although definitions differ, an independent director is one who exercises objective and unconfined judgement (IoDSA 2016). Having a higher percentage of independent directors than affiliated directors on a board has been shown to be more effective when it comes to monitoring managers and protecting shareholders’ interests (Muniandy & Hillier 2015). Furthermore, despite the fact that studies have mixed results on the topic, the majority of researchers state that there is a positive relationship between board independence and firm performance (Johennesse & Budidarma 2022; Liu et al. 2015). In a South African study, Muniandy and Hillier (2015) found that there is a positive association between firm performance and independent directorship.

Although, in theory, there are a vast number of benefits that arise from having independent directors on a board, firms may make false claims on the independent state of their boards or fill positions of independent members on a board for the sake of tick-boxing. A firm can claim that INEDs had been on the board for more than 9 years. Given their exceptionally long tenure on the board, their independence can be questioned. Another concern regarding board independence is the varying definitions and perceptions of independence.

These perceptions differ from one another in organisations, regulatory bodies and investors. For example, firms may vary in their opinions of what the tenure threshold is to no longer consider an NED as independent. Director tenure presents a conundrum to firms (Huang & Hilary 2018). Lengthy tenures are likely to impose on the independence of directors, as independent directors are more likely to become affiliated with other directors and the firm as tenures lengthen. However, directors with longer tenures gain more experience and enhanced decision-making abilities (Huang & Hilary 2018). Therefore, this trade-off is up to the subjective discretion of shareholders and the nomination committees.

The majority of literature, however, indicates a lack of independence on family firm boards (Kilincarslan 2021b; Shaw, He & Cordeiro 2021). In line with the family influence and control dimension of SEW, several scholars who have conducted research in the context of Malaysia, Turkey and India suggest that family firms tend to avoid appointing INEDs to maintain family control and facilitate their influence on management (Jong & Ho 2019; Kilincarslan 2021b; Shaw et al. 2021; Siebels & Knyphausen-Aufseß 2012). A widely used cross-country research article by La Porta, Lopez-de-Silanes and Shleifer (1999) states that, in family firms whereby the family holds a minimum of 10% in shares, at least 69% of them have family members on the boards. Family firms prefer to appoint affiliated and family members as directors even if it restricts a board’s ability to monitor management (Jones, Makri & Gomez-Mejia 2009). In a similar manner, Viviers (2022) questioned the effectiveness of INED categorisations of family firms listed on the JSE.

Another South African study concurred with the above-mentioned findings and stated that, on average, boards of non-family firms are more independent than boards of family firms in South Africa (Mashele 2021). In essence, a key way for family owners to exercise control is through board representation (Kilincarslan 2021b). The lack of board independence in family firms may provide the advantage of reducing agency costs between principals and agents, as family executives closely monitor managers on the behalf of shareholders (Jong & Ho 2019). However, there is also a disadvantage to minority shareholders, as INEDs play an essential role in alleviating conflicted interests between major and minority shareholders (Kilincarslan 2021b). Thus, overall, the lack of independence may reduce the board’s effectiveness as a monitoring mechanism (Kilincarslan 2021b).

There are, however, studies that oppose the notion that family firms lack board independence. In their systematic review of 106 studies conducted, Habib et al. (2019) suggest that family firms tend to appoint INEDs to signal good governance to shareholders. Similarly, Jong and Ho (2019) found that Malaysian family firms appoint more INEDs than family directors on their boards. The findings of a study conducted on Turkish family firms also suggested that boards have evolved to become more independent and that family decision-making is being more contested by INEDs (Kilincarslan 2021b). Although most literature supports the notion that family firms typically lack board independence, recent articles suggest the opposite. These mixed findings once again advocate the need for further research on this topic.

Research investigating the relationship between dividend payments and board independence in family firms is scant. The majority of studies were conducted in a developed economy context (specifically Europe and the United States of America and the Asian context) (Al-Najjar & Kilincarslan 2016; Djebali & Belanès 2015; Duygun, et al. 2018). As far as can be established, to date, there are no existing South African studies that investigated this topic. Upon reviewing these studies, it was found that, on balance, family firms tend to pay lower dividends (i.e. retain more earnings) than non-family firms. The over-retention of earnings may signal the presence of the principal–principal conflict, whereby majority (family) shareholders are expropriating the wealth of minority shareholders (Djebali & Belanès 2015; Setia-Atmaja et al. 2009).

Minority shareholder wealth expropriation may be mitigated through the implementation of stricter corporate governance mechanisms in family firms (Benjamin & Zain 2015; Djebali & Belanès 2015). This study focused specifically on the influence of board independence on dividend distributions. Research on the aforementioned relationship is limited. Existing studies yielded mixed results on the topic; however, the majority of the studies indicated a negative relationship between board independence and dividend distributions (Al-Najjar & Hussainey 2009; Benjamin & Zain 2015; Sener & Selcuk 2019). This common result suggests that when the percentage of independent directors increases, dividend payouts decrease. This finding contradicts previously mentioned studies that state the principal–principal conflict may be reduced through enhanced corporate governance implementation (Benjamin & Zain 2015; Djebali & Belanès 2015) but rather suggests that dividend distributions and board independence are substitute monitoring mechanisms.

Table 1 summarises the findings of a handful of studies investigating this relationship and demonstrates that the relationship between board independence, as measured by the percentage of INEDs serving on the family firm’s board, and dividend payments is complex and still requires further investigation and clarification. Studies on this topic were conducted in a mixture of both developed and developing countries. Furthermore, the majority of the studies conducted cover a short research period (of approximately 6–8 years) and fall short in covering more recent years such as 2018–2022. Including these years in the current research may be prudent as it includes the impact of the COVID-19 pandemic. Regarding chair independence, Sener and Selcuk (2019) found that the existence of a family chairperson reduces dividends. This finding seemingly contradicts the above-summarised studies that report a negative relationship between dividend distributions and board independence.

All studies on this topic mentioned in Table 1 used the agency theory as their theoretical framework. Limited research has been conducted using the SEW and catering theory of dividends. Furthermore, existing literature predominantly focuses on the comparison between family firms and non-family firms and lacks an awareness or mention of business heterogeneity, which has led to an oversimplification in results (Molly & Michiels 2021). Heterogeneity is a vital factor to consider in the study of family firms, as they differ in terms of managerial decision-making and behaviour. In light of the preceding discussion, the following null hypotheses were formulated:

The population comprised all South African JSE-listed family firms over the research period (2006–2022). Even though the most common approach to define family firms is based on shareholder ownership, this approach could not be used in the South African context because of a lack of data availability. Family firms were thus defined based on family influence at the board level. The dataset constructed by Viviers (2022) of listed South African family firms over the period 2011–2019 was used as the sampling frame, and the dataset was interrogated, extended and refined to include data for the years evaluated in the current study. It was created through an extensive review of online sources (such as credible newspapers, financial magazines, company websites and integrated reports). Keywords used in the process included, but were not limited to, ‘family firm’, ‘founding member’, ‘family business’, ‘family-owned’, ‘family-controlled’, ‘founding family’ and ‘family business empire’. Viviers (2022) had a sample of 31 family firms. To ensure validity, the authors used the same methodology as Viviers (2022) to identify three additional family firms (the sample of this study thus consisted of 34 family firms).

A database containing the names of directors of JSE-listed firms as 2010 was furthermore reviewed. This study used specific criteria to identify family firms for the sample. A firm was classified as a family firm if the surname of a director corresponds with the name of the firm, or the surname of the director corresponds with the founding family of the firm, or a director was not a founder but was associated with the family firm for more than 30 years, or the firm was a well-known family firm in South Africa. Family firms that were suspended or that were delisted during the research period (2006–2022) were included to address potential survivorship bias.

Data on the 34 identified family firms’ industry association and year of establishment were collected from their integrated reports and websites, as well as other credible online sources. The dependent variable in this study, dividend payments, was measured by using three proxies, namely, dividend payout ratio (DPR), dividend per share (DPS) and the propensity to pay dividends. Dividend payout ratio refers to the percentage of earnings distributed to shareholders in a given year and is calculated as follows (Equation 1): DPR=Total ordinary dividendsIncome before extraordinary items-Minority and preference dividends×100[Eqn 1]

Dividend payout ratio data were collected from the Bloomberg database. Missing values were replaced with zeros based on the assumption that no dividend was paid. Propensity to pay was based on DPR. If DPR was zero, a value of zero was allocated. If DPR was negative (which occurs when the denominator in Equation 1 was negative) or greater than zero, a value of one was assigned to indicate that a dividend was paid. DPS data were also sourced from Bloomberg. This ratio, expressed in South African Rand, was calculated as follows (Equation 2): DPS=Total ordinary dividendsNumber of outstanding ordinary shares at financial year end[Eqn 2]

The independent variable, board independence, called for the collection of demographic data for each of the 719 directors who served on a family firm board over the research period. The following data were hand collected from the family firms’ integrated reports: the director’s surname and initials, membership of the founding family (1;0), status (executive; non-executive; INED) and date appointed to the board. The surnames of women directors who married or divorced during the research period were carefully checked to avoid double counting. Each director’s tenure was determined by comparing the year under consideration with the year in which the director was appointed to the board.

Two proxies for the board independence were employed. The first, %INEDs as reported, was computed as the number of directors who were categorised as INEDs by their firms divided by board size at the financial year end. As in the foundational dataset, alternate directors were also excluded in this study. The second proxy, %INEDs re-categorised, was deemed important to obtain a more reliable measure of board independence. INEDs who were family members and those who had tenures of more than 9 years were re-categorised as non-executives. This decision was based on recommendations contained in King IV and international best practices such as the UK Corporate Governance Code (Cassim 2022). The percentage of re-categorised INEDs was also calculated relative to board size at the financial year-end.

In line with previous researchers (Benjamin et al. 2016; Djebali & Belanès 2015; Setia-Atmaja et al. 2009), a number of control variables were included, namely board and company size (measured in terms of market capitalisation), FCF, return on equity, operating profit margin and market-to-book ratio. With the exception of board size, all data for the control variables were collected from the Bloomberg database.

Several cross-checks were performed to ensure that the data captured in the Bloomberg database were accurate. As outliers were detected in the market capitalisation and FCF data, Log10 transformations were performed. Trends in the data over time were investigated using mixed model analysis of variance (ANOVA) tests along with Fisher’s least significant difference (LSD) tests. The primary hypothesis was examined using three panel regressions (one for each dividend proxy). Multicollinearity was evaluated by calculating variance inflation factors (VIFs), which were all found to be smaller and acceptable (< 2.5). All panel regressions were adjusted for heteroscedasticity.

This study made a theoretical contribution to the field of board independence and dividend distributions, in general, and among listed South African family firms, in particular. Specifically, the relationship between board independence and dividend distributions in JSE-listed family firms has not yet been investigated in South Africa. The study’s results gave insights on the trends present for both dividend distributions and board independence over the research period. Furthermore, family involvement on family firm boards was examined. This study produced new evidence that dividend distributions and board independence can be used as substitute monitoring mechanisms in family firms. The study ultimately demonstrated the uniqueness of family firms and confirmed their heterogeneity.

Given the extensive focus on the agency cost theory in existing literature, this study brought a new perspective from behavioural corporate finance by considering the SEW theory and the catering theory of dividends. Furthermore, this study used several measures for dividends (particularly the inclusion of DPS), as it provided a new perspective on perceived losses and gains of dividend income according to behavioural finance theories. Lastly, the literature review and empirical analyses exhibit many insights that may be converted to recommendations for several stakeholders.

This study provides practical contributions to shareholders, board committees, policymakers and commerce educators. All shareholders of family firms will benefit from becoming more educated on the unique characteristics present in family firms (heterogeneity), including their predilection of maintaining family control and, therefore, potentially lower board independence than non-family firms. However, given their own investment in such firms, family managers and directors have personal vested interests that may contribute to the overall success of the firms. Consequently, they may be regarded more as stewards rather than agents. It is thus critical for more education on the topic. It is further advised that family members who are appointed to the boards are trained extensively on how to be effective directors. Successful boards are committed to the improvement and development of their firms. Similarly, external directors are encouraged to gain a deeper understanding on the family values and history of the family firm. The additional training and educating of both types of directors may enhance the effectiveness of board decisions in such a way that all stakeholders are considered.

Majority shareholders, which are often family members in the case of family firms, have substantially larger voting power at annual general meetings and other shareholder meetings than minority shareholders. Consequently, their votes have a large impact on which individuals are elected as board members and who the chairperson of the board will be. Considering the SEW theory, it is advised that majority shareholders aim for a delicate balance in board composition of family involvement and independent directors and that strong consideration should be given to electing an independent and outside director as the chairperson of the board.

Most corporate governance codes encourage structures and practices that are predominantly tailored for largely dispersed-owned public firms (Balderrama 2015). Family firms, in particular, however, have distinctive goals and governance practices compared to non-family firms. The family’s emotional attachment to the firm, concern for dynastic succession and retaining control are merely a few reasons as to why family firms may differ from non-family firms in their corporate governance practices such as board independence (Hasenzagl et al. 2018). Consequently, a ‘one-size-fits-all’ approach to policymaking may not be as beneficial for different types of organisations as it seems.

South African policymakers such as the IoDSA and the King Committee would also benefit from further educating themselves on the cultural differences present in family firms. For example, policymakers who are aware of the unique characteristics in family firms will have a clearer understanding on the preference for family involvement at the board level. While family involvement may hinder independence, it could ultimately benefit all shareholders and stakeholders. The presence of family stewards in these firms may be a call for a more nuanced approach to governance regulations. Further, policymakers are encouraged to offer training and education on the topic of family firm governance to investors and directors.

In addition, shareholder activists will benefit from gaining an understanding that public disclosures do not display the full extent of family governance in these firms. Although family firm boards are becoming increasingly more independent, the empirical findings in this study, however, indicated a large disparity between the percentage of INEDs as reported according to integrated reports and the percentage of INEDs when reclassified according to tenure and family affiliation over the research period. This finding may be a cause for concern for shareholders. It is thus recommended that nomination committees are more comprehensive in their reporting practices.

Proper compliance with corporate governance standards is essential to increase investor confidence and credibility in publicly listed firms (Kilincarslan 2021a). Nomination committees are encouraged to report clearly on the rationale behind the classifications of INEDs and the value brought by non-independent board members such as family members. In short, it is advised that there is a higher level of transparency and comprehensiveness in the integrated reports of JSE-listed family firms.

Tertiary educators in commerce faculties are at the forefront of developing and enlightening students who will become future business leaders and shareholders. Given their impact, it is vital that tertiary educators enlighten students on corporate governance (beyond financial implications) for different types of firms, including publicly listed family firms. Cultural competence may be a positive contributor to the long-term success of these firms and ultimately the broader economy in which they operate (Tipu 2018).

The foremost limitation of this study was regarding the sample size. In South Africa, there is a lack of accurate data on the shareholder names and ownership structures of JSE-listed firms. Accordingly, family firms were identified according to the family presence on the board of directors. Although the researcher attempted to be as thorough as possible, there is a possibility that there are other family firms that may have been identified through their ownership structure. The restricted sample size may also have had an influence on the number of significant relationships that were identified.

The second constraint was a lack of information regarding director relationships with the families of family firms. To be more specific, it was difficult to discern whether a director is a family member or directly affiliated with the family based purely on their surnames. Consequently, there was a margin of error regarding the percentage of INEDs on the board and the level of family involvement in these firms. The last limitation relates to the exclusion of alternate directors in the dataset. These directors were excluded, as their extent of influence on strategic decision-making was not clear and may have differed from firm to firm. The barring of alternate directors may have had a minor effect on determining the true level of family involvement in family firm boards. Thus, a recommendation for future research would be to complement the current sampling and data collection methods with qualitative research such as surveys whereby researchers can gain a better understanding on the nature of the relationships present on these boards and whether these firms consider themselves as family firms. Qualitative research such as surveys and interviews may also help future scholars to gain insights on what motivates family firms more – financial or non-financial goals.

As previously mentioned, research regarding South African-listed family firms is limited. Thus, any contribution to this body of knowledge is encouraged. The researcher believes that it would be interesting to investigate the relationship between dividend distributions and other corporate governance mechanisms such as executive remuneration and internal auditing in South African-listed family firms. Further, because of time restraints, the researcher was not able to compare the results of family firms to a matched sample of non-family firms. Valuable insights may be gained from using a matched sample, as the extent of differences between listed family firms and their non-family firm counterparts in the South African context would be investigated.

The last recommendation for future scholars is to investigate the current topic in a cross-country study. Differing legal protection for shareholders in various countries might have implications for family firms’ payout policies and corporate governance compliance. Further, there may be discrepancies identified in the results based on a country’s level of economic development. As data availability on ownership structures may vary among countries, the researcher recommends future scholars to use the same approach as the current study to identify family firms – that is, through family influence.