Role of Family Ownership in the Relationship between Corporate Social Responsibility and Firm Performance

This study examines the eﬀect of corporate social responsibility (CSR) on ﬁrm performance in Indonesia. Most Indonesian companies are family-owned; therefore, it is important to consider the family ownership’s role in the relationship between CSR and ﬁrm performance. The study sample consists of 285 Indonesian listed ﬁrms for the period 2015–2019. Our results show that CSR positively aﬀects performance. Companies that conduct more CSR activities perform better, indicating their importance. Further, the interaction between family ownership and CSR negatively aﬀects ﬁrm performance. Therefore, family ownership weakens the positive eﬀects of CSR. Family owners have signiﬁcant disincentives for the CSR’s positive eﬀect in improving ﬁrm performance.

One reason for the mixed findings is the controlling shareholders. Consistent with agency theory, managers' engagement in CSR reflects their alignment with shareholder interests. Given that family-owned firms most likely choose their family members as managers, family-owned companies will be more active in CSR to increase value. For example, nonfamily-owned companies, which have higher agency costs [15], invest less in CSR; hence, they have a weak performance. Based on this logic, the alignment hypothesis of agency theory broadly suggests that the presence of family business in the organization motivates the organization to exercise CSR appropriately and for value creation [16,17]. However, this important topic has not received direct attention in accounting and finance research and is thus poorly understood, a gap that this research aims to tackle.
is study thus aims to fill in the lacuna by examining the moderating role of family ownership in strengthening the CSR effect on firm performance. Indonesia offers an ideal research setting for examining this argument. First, like other emerging countries, especially in the East Asia region, family business still dominates the ownership structure in Indonesia [18]. Given that family business generally has stricter control and monitoring [19] and is oriented on value creation [20], any act or exercise of CSR from this type of business might be close to transactional or profit-sensible. Additionally, like other countries, even though CSR is regulated and required by law, its intensity and targeted stakeholders depend on the firms' deeds. e manager has the freedom to choose which type of social and or environmental activities, how many programs, or which stakeholders, leaving the full authority to the managers. Finally, the loosen and lenient CSR regulation combined with the issue of governance in Indonesia may give intriguing insights for future emerging countries' study.
Addressing the moderating role of family ownership on the positive effect of CSR on firm performance is our focus. It makes a number of contributions. First, we bridge the institutional and agency theories by addressing controlling shareholders of family business as a moderation to overcome agency issues arising from the CSR-performance relationship. However, rather than finding support for the excellent monitoring of family business in the CSR activities, we find that family business weakens the positive effect of CSR on firm performance. Specifically, our findings show that CSR does increase firm performance and family-owned businesses outperform nonfamily businesses in terms of financial performance.
Nevertheless, when CSR is handled by a family-owned business, it deteriorates financial performance. We explore it further in subsampling analysis by industry (mining vs. agriculture) or performance period (profitable period vs. losses period); the conclusion remains intact. ese results counter one of the most fundamental recommendations typically made by alignment hypothesis proponents. Instead, it suggests that family firms need to rethink, reconceptualize, or re-strategize their CSR activities to achieve better performance.
is empirical finding of the family business as a moderation variable may also give new insights for further research in this area, especially the emerging market studies. Additionally, this study establishes that family ownership may play a crucial role in determining a firm's performance in relation to CSR.
is paper is organized as follows: e next section reviews the literature and hypothesis development for the relationship between CSR and firm performance as well as family ownership's role in this relationship. e following section describes the study's methodology. e fourth section provides an analysis and discussion of the results, followed by the conclusion.

Literature Review and
Hypothesis Development e tenet of stakeholder theory argues that a company has many parties as stakeholders [21], such as customers, employees, regulators, and others. Each party has its agenda and interests. erefore, managers must manage the relationship between stakeholders and companies to positively affect companies and society [22]. is concept is further explored, establishing the institutional theory by DiMaggio and Powell [23]. It is well tested in CSR literature by proposing that CSR's benefits are higher than its costs (i.e., [22,24]).
Empirically, this area is dominated by showing a positive effect of CSR on firm performance [6][7][8]25]. For instance, Cho et al. [7] found a positive relationship between CSR and financial performance on the Korean Stock Exchange. A study by Jang et al. [26] also found a positive effect of CSR on listed Korean firms' financial performance (ROA and return on equity). is result aligns with the study by Javeed and Lefen [8] who found that CSR in Pakistan positively affects firm performance.
is positive association is also found in Indian firms by Laskar and Maji [27]. Meanwhile, Sardana et al. [28] take marketing performance as the dependent variable but result in the same conclusion of the positive effect of CSR and performance. Beck et al. [29] conducted a cross-country study on the relationship between CSR and performance, confirming the expectation that CSR and firm performance have a positive relationship. is cross-country study is also well documented in emerging markets [30], Latin America [31] and Europe [32]. Further, Chen et al. [33] analyzed the green initiatives' effect on firm performance using 500 companies from 34 countries, but the results varied by country. e same conclusion is also documented by Walker et al. [34] who found that the effects of CSR differed between coordinated economies and liberal market economies.
Conversely, several research findings reveal a negative association between CSR and firm performance. Oh et al. [35], for instance, conducted a survey on the effects of CSR on performance, revealing a negative relationship between CSR and firm performance. A company that treats CSR as a charitable activity rather than a strategic activity perceives it as expensive. Such companies might not strategically engage in CSR to benefit stakeholders and society. us, CSR might not improve a firm's competitive advantage.
is was confirmed by Crisóstomo et al. [9], Setiawan et al. [36], Garg and Gupta [11], and Lahouel et al. [12], who found a negative relationship between CSR and firm performance. Intriguingly, empirical findings in this area also document the trivial CSR's effect on firm performance. For example, McWilliams and Siegel [37], who investigated in the US context, revealed that CSR had no significant effect on firm performance. is result was confirmed by other findings, such as those by Aras et al. [13]; Sekhon and Kathuria [14], Oeyono et al. [38], and Sekhon and Kathuria [14]. For the Indonesian context, Oeyono et al. [38] showed that CSR had no significant effect on EBITDA and EPS, concluding that CSR does not improve firm performance.
e results of previous studies on the effect of CSR on firm performance are therefore inconclusive. However, institutional theory predicts that firms attempt to maximize stakeholders' interests through CSR because their benefit is higher than CSR's cost. CSR adds value to the company, thereby improving its competitive advantage. CSR positively affects firm performance [6][7][8][26][27][28]. us, CSR is expected to have a positive effect on firm performance. H 1 : Higher CSR might lead to strong firm performance.

Role of Family Ownership.
ere are two competing theories regarding family ownership's effect on CSR: the alignment effect and entrenchment effect [39]. e alignment effect predicts that family ownership positively affects CSR implementation, aligning their interests with other stakeholders. For example, a previous study showed that family firms engage more in CSR [40] to increase CSR performance [41]. Family firms are also more concerned about environmental investments [42]. Meanwhile, the entrenchment effect argues that family ownership puts more effort into profitability. Family owners perceive CSR as obligations rather than strategic decisions. e cost of CSR is higher than the benefits earned from them; therefore, family ownership is expected to affect CSR negatively. Labelle et al. [43] provide evidence that family firms have lower CSR performance than nonfamily firms.
It is consistent with the institutional theory argument. For example, Chen et al. [33] and Walker et al. [34] argue that CSR's effects on firm performance should consider the institutional context, which might be an important factor affecting the relationship between CSR and performance. One such aspect is the ownership structure [16,17]. In the context of Korean firms, large shareholders have significant effects on CSR. Large shareholders, such as chaebols, have the ability to drive CSR [16]. Further, Zaid et al. [17] provide evidence that government ownership, institutional ownership, and foreign ownership positively affect CSR. eir results showed the importance of the ownership structure. Indonesia has unique characteristics, as most Indonesian firms are family-owned [44].
Most Indonesian firms are owned by families. Family owners significantly affect firms' decisions regarding CSR [42,43]. erefore, it is important to consider family ownership in the relationship between CSR and firm performance. Previous studies have considered family ownership as a moderating variable on CSR and earning management [45], CSR and entrenchment strategy [46], CSR and firm performance [47], and CSR and information asymmetry [48]. erefore, the second hypothesis is as follows: H 2 : e financial performance of family-owned firms outperforms nonfamily firms. H 3 : Family ownership strengthens the positive relationship between CSR and firm performance.  [42] and Wang et al. [25]. is study used the Global Reporting Initiative (GRI) G4 as a proxy for CSR. Family ownership is defined as the percentage of shares owned by the family [44,45].

Data and Methodology
e control variables consisted of five variables: leverage, firm size, auditor firm, growth, and firm age. Leverage is measured using the debt-to-equity ratio.
is study used total assets as a proxy for firm size. Further, this study used a dummy variable for the audit firm: 1 if the audit firm was a Big-4 firm and 0 otherwise. Growth was measured using the market-to-book ratio, and firm age is the number of years since the firm's establishment. As pointed out by Dang et al. [49], the firm size is an important variable in finance. us, we included firm size in this study. Our study follows the method used by Li et al. [50], Ikram et al. [51], and Dunbar et al. [52] that use total assets as the measurement of firm size.
e following equation was used to test the hypothesis: where Perf � firm performance, measured using Tobin's q; CSR � corporate social responsibility using the GRI index. Fam � family ownership, the percentage of firm shares owned by the family. Lev � Leverage, debt-to-equity ratio.
Size � size of the firm, ln total assets. Aud � audit firm, dummy variable � 1 if the audit firm is Big-4 and 0 otherwise. Growth, market-to-book ratio. Age � firm's age, the number of years since the firm's establishment. Tables 1 and 2 provide the descriptive statistics used here. Table 1 shows that Tobin's q value ranges from 0.0575 to 4.5888, with a median (mean) value of 0.6839 (0.8002). is shows that, on average, Tobin's q is below 1. Further, Table 1 shows that CSR disclosure ranges from 24.27% to 49.26%, whereas the mean CSR value is 33.09%. us, Indonesian firms disclose 33.09% of the information required by the GRI index. For family ownership, Table 1 shows that the median (mean) is 59.62% (55.89%) shares owned by the family. erefore, family owners have a higher opportunity for firm control. e leverage of the research sample ranges from 2.36% to 225.78%, and the mean leverage value was 53.37%.

Descriptive Statistics.
On average, the composition of debt and equity in the sample is almost balanced. e mean firm age is 28.12 years. From Table 2, we see that most mining and agricultural firms in Indonesia are audited by non-Big 4 audit firms. Table 3 provides information on CSR's effects on firm performance with family ownership as a moderating variable. First, it shows that CSR has positive effects on firm performance (P < 0.01), supporting our first hypothesis (H1: a higher CSR leads to stronger financial performance). Firms enjoy more benefits from CSR costs. CSR's positive effects on firm performance are robust, as presented in Columns 1 and 2 of Table 1. It aligns with institutional theory, whereas managers need to manage the relationship between companies and stakeholders to obtain the best interests of both parties [22]. Managers should Complexity 3 expend more effort to achieve the best interests of stakeholders and companies through CSR involvement. Our findings confirm previous studies that demonstrated the positive effect of CSR engagement on performance [8,[26][27][28]. Table 3 also shows the positive effect of family ownership on firm performance, supporting our second hypothesis (H2: the financial performance of family-owned firms outperforms the non-family-firms). It shows that the financial difference between family-owned and nonfamilyowned was 1.4%, supporting the alignment hypothesis of agency theory. It implies that family ownership aligns its interests with those of other shareholders; therefore, family owners use their discretion to improve performance and retain their reputation. Our findings confirm previous studies that family owners positively affect performance [15,53].

Regression Results.
For the main findings, we reveal that the interaction between CSR and family-owned firms has a nontrivial effect on performance, consistent with the findings of Kumala and Siregar [45], Martínez-Ferrero et al. [46], and the authors of [54]. However, the interaction has negative signs Perf � firm performance, measured using tobin's q; CSR, corporate social responsibility, using the GRI index; Fam � family ownership, the percentage of firm shares owned by the family; Lev � leverage, debt-to-equity ratio; size � size of the firm, ln total assets; growth, market-to-book ratio; and age � firm's age, the number of years since the firm's establishment.  , * * , and * * * refer to the 10%, 5%, and 1% significance levels, respectively. e figures are beta coefficients except for the figures inside the parentheses, which are p-values. Perf � firm performance, measured using tobin's q; CSR, corporate social responsibility, using the GRI index; Fam � family ownership, the percentage of firm shares owned by the family; Lev � leverage, debt-to-equity ratio; Size � size of the firm, ln total assets; ise; growth, market-to-book ratio; Age � firm's age, the number of years since the establishment of the firm; and Aud � audit firm, dummy variable � 1 if the audit firm is big-4 and 0 if otherwise.
(β � −3.0461), indicating the negative moderating effect. is result implies that family-owned firms weaken the positive effect of CSR on firm performance.
ere are three explanations for this conclusion. First, family owners might not support CSR as a strategic decision to enhance firm value but treat CSR as a stakeholder liability. CSR is required by law, and the family-owned firms view this as an additional cost. Instead of using this CSR for reducing contingency cost [55] or strengthening market competition [56], family-owned firms treat CSR as additional nonoperating costing resulting in deteriorated earnings.
Second, the human capital and nonmarket cost to access the CSR activities. Most family businesses are run by family members; hence, they might not be exposed to effective CSR activities due to human capital [57][58][59] or additional nonmarket costs [60,61]. Additionally, Gavana et al. [62] argue that CSR activities are a nonmarket strategy to manage their earning reporting. ey found that family owners use CSR to mask downward earnings management.
Meanwhile, for the control variables results, Table 3 displays that leverage positively affects firm performance. Firms with higher Leverage are more likely to achieve higher performance. Further, firm age and growth positively affect firm performance. Older firm age and higher-growth firms improve firm performance. On the other hand, audit firms have a negative effect on firm performance, and firm size negatively affects firm performance.

Robustness Check: Subsampling.
Our sample consists of two important industries with different profitability characteristics. us, it is possible that the significant baseline results from the pooled sample might be the net effect of varying relationships across industries offsetting each other. Moreover, the intensity of CSR may impact firms differently because of their profitability. To tackle this issue, we reestimate the model under different groups, which is called subsampling.
e study sample was divided into mining and agriculture industries, and the sample was divided into profit and loss subsamples. Columns 1 and 2 of Table 4 show that subsample mining and agriculture results are qualitatively the same. CSR positively affects firm performance, while family ownership weakens CSR's positive performance effect. Furthermore, the results were different when the sample was divided into profit and loss subsamples. In Column 3 of Table 2, the result was qualitatively the same as the main test.
us, in profitable firms, CSR positively affects firm performance, and family ownership moderates this relationship. However, the subsample of unprofitable firms showed a different result. During loss periods, CSR still has a positive effect on performance; however, family ownership has no significant effect.
Further, the interaction between CSR and family performance during the loss period has no significant effect on firm performance. e results of the additional tests show that CSR positively affects firm performance. Further, family ownership moderates the effect, except when the firm faces a loss; then, family ownership has no significant effect on performance.

Robustness Check: Moderation Plot.
We further examine family-owned firms' moderating effect using a moderation plot. We follow Dawson [63] to portray the moderation effect; the results are shown in Figure 1. 79 * , * * , and * * * refer to the 10%, 5%, and 1% significance levels, respectively. e figures are beta coefficients except for the figures inside the parentheses, which are pvalues. Perf � firm performance, measured using tobin's q; CSR � corporate social responsibility, using the GRI index; Fam � family ownership, the percentage of firm shares owned by the family; Lev � leverage, debt-to-equity ratio; size � size of the firm, ln total assets; ise; growth, market-to-book ratio; age � firm's age, the number of years since the establishment of the firm; and Aud � audit firm, dummy variable � 1 if the audit firm is big-4 and 0 if otherwise. is study has three key findings. First, the performance of family and nonfamily firms is significantly different, with a minimum level of CSR. It implies that if family firms engage in minimum CSR, their performance is much lower than nonfamily firms. Additionally, when both groups pursue CSR, nonfamily firm performance increases steadily. However, the performance of family firms drops drastically when CSR increases. Lastly, the performance gap between family-owned and nonfamily-owned firms is massive, with a maximum level of CSR. is figure confirms our findings that family-owned firms weaken the positive effect of CSR on firm performance.

Endogeneity.
We further examine the potential endogeneity in this study. We follow the suggestion of Li [64] to test the endogeneity using the lagged dependent variable in this study, which is firm performance t−1.
e result of the endogeneity check using the lagged dependent variable shows that the main effect of CSR and the interaction between CSR and family ownership remain the same as in the main test. CSR has a positive effect on firm performance; therefore, a higher CSR improves firm performance. Furthermore, the interaction between family ownership and CSR has a negative effect on firm performance. erefore, family ownership weakens the positive effects on firm performance (Table 5).

Conclusions
is study provides empirical evidence about the importance of CSR engagement in achieving firm performance. e higher a firm's engagement with CSR, the greater the positive effect and performance. CSR positively affects firm performance. Companies must focus greater attention and effort on CSR activities. Firms that increase CSR enhance stakeholder value because of their commitment to increase sustainable practices. Rather than being a liability, CSR makes a good strategic sense for a company and its stakeholders. Second, this study shows that family ownership has a significant effect on the relationship between CSR and firm performance. Family ownership weakens CSR's positive effects on firm performance. CSR's positive performance effect is reduced in family-dominated firms.
is study focuses on CSR's impact on performance in the Indonesian context. It would be interesting to widen the study sample to investigate CSR's effects on performance in other countries, such as ASEAN members, Asian countries, and emerging markets, with the role of family ownership as the moderating variable. e study on the effect of CSR to the firm performance with family ownership as moderating  227 * , * * , and * * * refer to the 10%, 5%, and 1% significance levels, respectively. e figures are beta coefficients except for the figures inside the parentheses, which are p-values. Perf � firm performance, measured using tobin's q; CSR � corporate social responsibility, using the GRI index; Fam � family ownership, the percentage of firm shares owned by the family; Lev � leverage, debt-to-equity ratio; size � size of the firm, ln total assets; growth, market-to-book ratio; and age � firm's age, the number of years since the firm's establishment. 6 Complexity variables will provide an interesting discussion with crosscountry studies, such as ASEAN or Asian countries where families are still the dominant owners of the firm. We will continue this in a future study. Previous studies show the importance of CEO characteristics such as the power to engage in CSR [65], compensation [51] and CEO age [66]. us, the relationship between CSR and performance may be affected by CEO characteristics. Powerful CEOs tend to invest less in CSR. Also, another interesting aspect is the effect of CEO succession [67] on CSR decisions [68]. erefore, investigating the effect of family CEOs on CSR and how they affect firm value could prove fruitful. Another study by Li et al. [50] argues for the importance of corporate visibility in CSR performance. Companies with higher visibility perform better in terms of CSR ratings. Company visibility results in greater stakeholder attention and pressure to engage in CSR.
us, future research should investigate how company visibility affects family firms' CSR activities and firm value in the emerging market context. We leave this for future studies.

Data Availability
e data for the current study were taken from a public source at Indonesia Stock Exchange.

Conflicts of Interest
e authors declare that they have no conflicts of interest.

Authors' Contributions
Conceptualization was conducted by DS, AA, and RKB. Data collection was performed by DS, HPR, and MWW. Methodology was developed by DS and RKB. Formal analysis was carried out by DS, AA, RKB, HPR, and MWW. Original draft was written by DS, AA, and RKB. final draft was written by DS, AA, RKB, and MWW.