Family Firms and Coupling among CSR Disclosures and Performance

: This paper aims to analyse the behaviours related to the decoupling of the disclosed information on Corporate Social Responsibility (CSR) and corporate sustainability, deepening these practices’ knowledge within family businesses. For this purpose, we deﬁned decoupling as a gap between social responsibility performance (internal actions) and disclosures (external actions). For a sample of 33,809 observations for the period 2011–2019, corresponding to 5029 companies, 19% being family ﬁrms, our empirical evidence supports that family ﬁrms present a less wide gap between performance and disclosure, conﬁrming the prevalence of socioemotional wealth dimensions in the decision-making of these companies. In ﬁrms without controlled shareholders, the quality of nonﬁnancial reporting could be understood as ambiguous, understanding that the most useful CSR information is found in the reports of family-owned companies.


Introduction
There is undoubted academic interest in knowing whether corporate decisions differ between companies depending on the presence or not of a family in the shareholding and key decision-making positions (i.e., Mariotti et al. 2020), especially in the field of corporate social responsibility due to the principles of socioemotional wealth that determine the decision-making processes in these companies (García-Sánchez et al. 2021b).
According to this paradigm, family businesses are interested in other aspects beyond financial ones to perpetuate the dynasty and the family's influence within their companies (Gómez-Mejía et al. 2007). These principles become a double-edged sword as an investment in sustainable projects will be oriented to satisfy the interests of these owners in the search for a more significant business reputation that guarantees their ultimate goal and may cause the economic interests of other nonfamily investors to be harmed (Morck and Yeung 2003).
This perspective of family firms suggests a more significant commitment to Corporate Social Responsibility (CSR) by favouring the transfer of the firm to future generations (Kim et al. 2016;Binz et al. 2017) and reinforcing the success and image of the company (Sharma and Sharma 2011). Although the literature is not unanimous, it can be argued that most of the time, but not always, they are socially and environmentally more responsible than other companies (i.e., Berrone et al. 2010;Cennamo et al. 2012;Cruz et al. 2014;Campopiano et al. 2014;Marques et al. 2014;Cuadrado-Ballesteros et al. 2015Martínez-Ferrero et al. 2016;Feliu and Botero 2016;Samara and Berbegal-Mirabent 2018;García-Sánchez et al. 2021a).
On the other hand, the academic literature has evidenced that CSR reporting content is ambiguous and presents impression management strategies to manipulate the opinion of stakeholders, limiting its usefulness in the decision-making process (García-Sánchez et al. 2020b). These practices are especially worrisome in less sustainable companies, as there is a decoupling between a firm's CSR disclosures and performance (Sauerwald and Su 2019). In addition, several firms adopt silent strategies around specific CSR practices in order to protect its interests from higher stakeholders' expectations. Therefore, there is an interest in determining whether or not family firms' practices follow other companies' information strategies, especially those relating to CSR decoupling, understanding like the gap that exists between firms' practices and the information disclosed in this regard.
In this sense, this paper aims to determine whether family firms show less decoupling in the disclosures they make concerning the CSR initiatives they develop. More specifically, the aim is to determine whether the family presence in shareholding and critical decisionmaking positions alleviates or aggravates CSR decoupling practices related to the voluntary omission of information on corporate behavioural responsibility or greenwashing practices.
From a theoretical point of view, we defend that these companies will be interested in demonstrating their commitment to CSR to avoid reputational losses (Koh et al. 2014) that would affect the family's success and image because of their close ties (Sharma and Sharma 2011;Marques et al. 2014). Therefore, one would expect the content of their nonfinancial reports-sustainability reports, integrated reports, nonfinancial information statements, etc.-to contain the necessary information that would enable the family to understand the company's reputation-collect essential information to allow stakeholders to assess the company's accountability.
To contrast the hypothesis research, we used an unbalanced panel data sample of 5029 international listed companies for the period 2011-2019, 19% of which are family-owned companies. According to the socioemotional wealth theory, the results obtained confirm our working hypothesis, showing that family firms pursue other additional and preferential purposes to maximise economic wealth and are interested in corporate reputation (mirroring the family reputation) favouring CSR coupling practices.
This paper contributes to previous literature on CSR decoupling and family business. First, we contribute to the literature on CSR decoupling (Winn and Angell 2000;Delmas and Burbano 2011;Walker and Wan 2012;Bowen 2014;Hawn and Ioannou 2016;Sauerwald and Su 2019;Tashman et al. 2019;Graafland and Smid 2019;García-Sánchez et al. 2020b), highlighting the role that one of the most relevant blockholders, family firms, can play in corporate decisions regarding the disclosure of more useful and reliable information to various stakeholders. Secondly, this study contributes to the family business literature by focusing on the decisions that these companies make about the relevance of the information they disclose and the adequacy to their CSR actions. Family researchers have currently focused on the analysis of family groups either in CSR practices (i.e., Berrone et al. 2010;Cruz et al. 2014;Cuadrado-Ballesteros et al. 2017;García-Sánchez et al. 2020c) or in corporate transparency (i.e., Cuadrado-Ballesteros et al. 2015;Gavana et al. 2017;Bansal et al. 2018). This paper is the first article that combines these two research lines.

Theoretical Framework and Research Hypothesis
The current business environment is characterised by a significant increase in companies' demands about the process of accountability to various stakeholders and society regarding the use of resources and the environmental and social impact of their activities (García-Sánchez et al. 2020a).
However, the underlying business motives for CSR disclosure are characterised by a high degree of ambiguity due to the consideration of possible use of CSR as a mechanism to manipulate external opinions regarding the company's behaviour or to manage relations with a specific group of stakeholders (i.e., García-Sánchez et al. 2020c).
Thus, several studies confirm the disclosed information's subjectivity, which hinders its comparability and limits its usefulness. This situation is aggravated in companies with less sustainable behaviour. In this regard, the literature has begun to refer to them as CSR gap, a term that identifies the existence of decoupling between what the company does and what it says it does (Sauerwald and Su 2019), either because companies need to legitimise themselves before their stakeholders or because they avoid creating high expectations among their stakeholders, which would lead them to face a possible risk of not being able to satisfy them, and could be accused of being hypocritical (García-Sánchez et al. 2020b).
More specifically, the literature uses the term CSR decoupling to refer to the "degree of misalignment between a firm's CSR reporting and CSR performance" (Tashman et al. 2019, p. 158), contemplating the possible existence of a complete divergence that identifies that the information disclosed is associated with "purely ceremonial CSR" (Graafland and Smid 2019, p. 231). Practices that would refer to symbolic corporate responsibility (Delmas and Burbano 2011;Walker and Wan 2012;Bowen 2014).
However, in general, the term CSR decoupling usually refers to the fact that the firm has chosen to overstate its CSR performance in its disclosure (Winn and Angell 2000;Delmas and Burbano 2011;Hawn and Ioannou 2016;Tashman et al. 2019). Conversely, Silent Firms are also used to identify those practices in which companies with good CSR performance do not communicate about all of these responsible projects and actions (Delmas and Burbano 2011).
Previous literature is inconsistent in determining whether or not the family firm's commitment to CSR is superior to that shown by other companies (Samara and Berbegal-Mirabent 2018). On the one hand, scholars argue that family firms are more concerned about social and environmental issues because their socioemotional wealth increases with the reputation they acquire with stakeholders (Cennamo et al. 2012). Sustainable growth and a better image favour the family's socioemotional interests (Gómez-Mejía et al. 2007Cruz et al. 2014). On the other hand, family problems of selfishness and family conflict, and nepotism, among others, would lead to the prevalence of socioemotional wealth being opposed to the demands of various interest groups, which could lead to decisions detrimental to their sustainability strategy (Zellweger et al. 2012).
In general, the literature argues that family firms' interests are associated with socioemotional wealth and, therefore, they perceive more benefits than risks from CSR (i.e., Berrone et al. 2010;Cruz et al. 2014;Cuadrado-Ballesteros et al. 2015García-Sánchez et al. 2021a). For that reason, their commitment to CSR is usually higher, which would justify the idea that they will be less prone to CSR decoupling practices.
Additionally, this decoupling would be oriented towards the search for legitimacy derived from the symbolic adoption of CSR actions, practices that could be identified by lobbying or other pressure groups, which could lead to sanctions and reputational losses jeopardise the preservation of the socioemotional endowment. Therefore, it seems appropriate to think that CSR decoupling would be a decision instead of the socioemotional criteria that families use when making their decisions. Wiseman and Gómez-Mejía (1998) and Gómez-Mejía et al. (2001, 2007 argue that preserving the socioemotional endowment is fundamental for the family and shapes the formulation of problems, becoming the primary reference point to guide management choices. On the other hand, the quest for a better reputation will encourage a match between what they say and what they do. A more significant commitment to CSR and greater alignment in their disclosure policies would favour the five dimensions of socioemotional wealth (FIBER) proposed by Berrone et al. (2012): influence and control of the family; identification of family members with the company; generation of social rootedness; emotional attachment of family members; and renewal of family ties through succession. Therefore, we propose the following research hypothesis: Hypothesis 1 (H1). Family firms' CSR reports present a lower decoupling with CSR performance than nonfamily firms.

Sample
To demonstrate the level of coupling of CSR reporting by family-owned companies, the largest listed companies worldwide were selected as the target population. A larger group of stakeholders than most observes these companies due to the impact their economic activity has on society and the environment. Subsequently, family firms were identified according to the criteria of ownership, management and control used in previous studies (i.e., Cascino et al. 2010;Singla et al. 2014;García-Sánchez et al. 2021b). Thus, companies whose majority shareholder is a family or family group owning at least 20% of the voting rights are considered family firms. Additionally, at least one family member must be part of the management team or the board of directors.
The information necessary for the analysis was obtained from Thomson Reuters. The final sample consists of an unbalanced panel of 33,809 observations for the period 2011-2019, corresponding to 5029 companies. Of the companies in the sample, 19% are family-owned companies. Table 1 shows the information on the sample distribution by sector and geographical area, identifying biases due to the variability inherent to international samples, which requires the inclusion of control variables at the country, sector and institutional environment levels.

Analysis Model
Equation (1) represents the empirical model designed to determine the interest that family firms have in CSR disclosure coupled with their commitment to the environment and society.
The CSRgap variable is a numerical variable that takes values between -1 and 1 to identify the decoupling level between companies' CSR performance and the disclosure of information associated with this strategy. It is constructed based on the proposal of Hawn and Ioannou (2016). More concretely, we determined the CSRgap variable as an absolute gap between the score of the 21 internal actions and 24 external that these authors used.
The FamilyFirm variable is a variable that takes a value of 1 to identify whether the company is a family firm, taking a value of 0 otherwise. The consideration of family firm has been made taking into account the most relevant characteristics of family firms are their presence in the ownership, controlling a percentage of voting rights over 20% (Campopiano and Massis 2015), as well as the presence of the founders and/or their descendants in management positions and/or on the board of directors (Chen et al. 2008;Cuadrado-Ballesteros et al. 2015).
To avoid biased results, a set of control variables have been included in the model by García-Sánchez et al. (2020b): company size measured by the logarithm of assets (Fsize); the age of the company represented by the number of years that have elapsed since its creation (Fage); growth opportunities identified by the average growth in sales over five years (GrowthOpp); the economic return on assets (ROA); the level of financial indebtedness (Flev); the internationalisation of the company represented by the percentage of investments in assets in other countries (Finter); working capital (Fwc); the number of analysts covering the company (CovAna). The adequacy of internal corporate governance mechanisms is also monitored using the GoverScore variable extracted from Thomson Reuters. Institutional pressures are identified with the country-and sector-level composite indicators, NCSRPI and ICSRPI, from Amor-Esteban et al. (2018a, 2018b, 2019).
The analysis techniques correspond to panel data regressions to control for unobservable heterogeneity (η), considering a one-period lag in the independent and control variables to control endogeneity problems. The variables Year, Country and Industry are included to control the variation across time, country and industry. According to the Hausman test, we used fixed-effects and a robust approach to avoid multicollinearity problems.

Basic Descriptive
The descriptive statistics for the different variables proposed for the analysis are reflected in Table 2. In this respect, it can be observed that the mean value of the CSRgap variable is 0.011, indicating that companies disclose less information about their commitment to CSR than the actions they have implemented to protect the environment and contribute to sustainable development.  Table 3 shows the bivariate correlations between the variables selected to estimate the empirical model. An analysis of the correlation coefficients shows the absence of multicollinearity problems.

Basic Model Results
The third column of Table 4 shows the results for the Equation (1) proposed to test the working hypotheses. In addition to the dependent variable CSRgap, Equation (1) was estimated, considering CSRPerf and CSRDisclo as dependent variables, to show whether family firms show more significant commitments to CSR and corporate transparency than the rest of the companies, and these results are reflected in the first two columns of Table 4. Table 3. Bivariate correlations (*** p < 0.01, ** p < 0.05, * p < 0.1).  In this respect, it can be observed that the variable FamilyFirm is statistically nonsignificant in the explanatory models of CSR practices, performance (Equ. CSRPerf: coeff. = −0.000517; p-value = 0.980) and transparency (Equ.CSRDiscl: coeff. = −0.0170; p-value = 0.426). These results would indicate that family businesses' commitment to sustainable development and corporate transparency is identical to that of other companies. However, this variable has a negative, econometrically significant impact on the CSRgap variable (coeff. = −0.0000384; p-value = 0.042). These results would be in line with our working hypothesis, confirming that family firms show lower decoupling levels between what they do and what they say in terms of CSR.
Together, although these results would not be in line with those obtained by Berrone et al. (2010) or García-Sánchez et al. (2020b), among others, which show that family businesses are more oriented towards social and environmental responsibility, we would confirm that family firms are just as sustainable as other companies. However, they disclose more relevant and useful information for stakeholders' decision-making. Specifically, the CSR information disclosed is more closely aligned with the sustainability strategies of these companies. It should be understood as a corporate decision subject to the preferences of a group of owners who see sustainability as a way of guaranteeing the family legacy (Feldman et al. 2016).
On the other hand, we confirm that the most profitable companies, with more excellent analyst coverage and located in countries more sensitive to CSR, are those with the highest CSR coupling. Conversely, larger companies operating in sectors more committed to CSR and better governance indices are set to silence part of their sustainability strategy. Table 5 shows different robust analyses relating to the CSRgap variable's complementary definitions to confirm the results. In this regard, the first column presents an alternative measure of decoupling following García-Sánchez et al. (2020b), named CSRgap2. The second column uses a dummy variable, DSilent, which identifies those companies whose nonfinancial reports present incomplete information that does not allow us to know the r CSR unit initiatives. Table 5. Robust models (*** p < 0.01, ** p < 0.05, * p < 0.1).

Equation (1)
Equation ( Total effect of the variable FamilyFirm on CSRgap2 is negative and econometrically significant (coeff. = −0.00112; p-value = 0.000). For the dependent variable DSilent, the effect is similar (coeff. = −0.0211; p-value = 0.000). Therefore, the results confirm those obtained in the initial model, allowing us to validate the interest that family businesses have in disclosing information linked to their sustainability performance.
Additionally, in the third column, we estimated Equation 1 with the initial CSRgap measure without any control variables that were nonsignificant from the statistical point of view. The omission of these variables is oriented to evidence that the models are not over-specified. The results obtained show the same effects as those observed in the previous analysis. The effect of FamilyFirm on CSRgap2 is negative and econometrically significant (coeff. = −0.0000289; p-value = 0.032).

Conclusions
The ultimate purpose of corporate information is its use in decision-making processes, sometimes by third parties outside the company. For the information to be useful and relevant, it must faithfully represent the reality of the companies, referring to its correctness and reliability, which should not be understood as the demand for mathematical accuracy or absolute veracity, but simply a reasonable approximation to this extreme veracity, and by correctness, not only formal correctness but the adequacy or correspondence of the data and figures with reality and business performance.
In the search for symbolic legitimacy or as a prevention mechanism, previous literature has highlighted a dissociation between the information disclosed on CSR and corporate sustainability, focusing on analysing the role that financial agents and corporate governance systems can play in this regard.
In this paper, given the peculiarities of family businesses, companies in which the family and the company interact, decisions were not taken according to the economic logic of the postulates of agency theory regarding profit maximisation. Instead, decisions were made according to an accumulated affective legacy or socioemotional wealth aimed at retaining ownership and control over the company and passing it on to future generations.
This decision-making framework leads these companies to make decisions to protect socioemotional wealth, finding in CSR a sustainable growth strategy that favours the image and reputation of the company and the family. In this sense, we argue that these companies will be less prone to greenwashing practices due to the risks that it may entail for their image and legacy, opting for a more significant coupling between what they do and what they say. We confirmed these results empirically for an international sample of listed companies.
These results have different implications for academia, practitioners and regulators. From a theoretical point of view, we extend the socioemotional wealth perspective, providing evidence that family firms show a greater CSR coupling preference than other companies. More concretely, using a sample with the largest multinational companies allowed us to contrast the differences between companies subject to the same pressures from the stock market and other common actors at a global level but differ in the ownership composition. In this sense, our results confirm that the different socioemotional wealth dimensions lead family businesses to behave more ethically when informing their stakeholders about their commitments to the environment and society. Although, unlike previous studies, it was not observed that they are more sustainable or transparent than the rest of the companies, only that the information they disclose is more closely related to their actions and, in consequence, more useful in decision-making.
From a practical point of view, the main conclusions can be drawn for the different stakeholders, who use the information reported by companies to make decisions and determine their relationships with them. In this sense, family businesses have turned out to be the cluster of companies that disclose the most useful information for decision making by reporting more accurately on their internal actions.
In this line, our evidence should be taken into account by governments, public institutions and policymakers in regulatory processes regarding corporate transparency so that nonfinancial reports are of higher quality by containing relevant information for decision making.
It also necessary to remark that these research results have significant consequences for society, to the extent that family firms help reduce the information gap between what companies talk and make in terms of CSR. This coupling discourse favours the knowledge of the real business commitment to the environment and society, determining with greater accuracy the contribution of companies to sustainable growth and the solution of the most pressing problems worldwide. In this sense, the use of information on CSR provided by family businesses can be used by different interest groups in their decision-making processes, favouring relationships with those organisations with higher contributions to social well-being.
Finally, it should be noted that this work is subject to several limitations that should be considered in future studies, especially those related to the possible confrontations that may exist within family businesses and the divergences that may exist between family businesses in terms of the degree of affectivity of the emotional legacy that may differ between generations or, even, the capacity to intervene in decision-making bodies. Additionally, the use of a sample of the largest family firms worldwide suggests the presence of a size and internationalisation bias that could be corrected through the analysis of family firms with an activity restricted to a specific geographic zone.