“The role of family businesses and active family members in environmental performance”

There is a growing concern about environmental issues, particularly carbon emissions, in many countries. Indonesia, with its huge population, also suffers from excessive carbon emissions. This study aims to investigate the effect of family businesses on environmental performance, specifically carbon emission disclosure. This study also explores the role of the family supervisory board and management on the quality of carbon emission disclosure. The study employed 62 non-financial family-listed firms in 2017–2019 (186 observations). The analysis found a positive and significant relationship between family enterprises and the disclosure of carbon emissions, implying that family firms expose more information about their carbon emissions. It also revealed a significant positive association between the family supervisory board and carbon emission performance, suggesting that having family members on the supervisory board aligns with policies for reducing and maintaining accountability for carbon emissions. In summary, the findings suggest that family enterprises prefer to exercise their indirect control by holding a position on the supervisory board and owning a substantial percentage of the company’s stock corresponding to their socio-emotional wealth agenda. Additionally, there is a non-linear association between family firms and the disclosure of carbon emissions. Carbon emission performance decreases as family share ownership rises to 53.1% but increases when family equity exceeds this cut-off point. Finally, family shareholders in non-polluted firms report higher quality of carbon emission disclosure.


INTRODUCTION
Many nations are now concerned about environmental issues, particularly carbon emissions. Numerous research has been done on firms' carbon emission disclosures, particularly on the variables that may affect the level of carbon emission disclosure (Shen et al., 2020). The majority of research, however, including those by Gray et al. (1995), Iyer and Lulseged (2013), and Baalouch et al. (2019), have concentrated on the United Kingdom, the United States, and countries in Europe. Limited studies have investigated the type and scope of company disclosure in Asia, particularly Indonesia. According to Chau and Gray (2002), Asian businesses are less driven to share information openly than their counterparts in Anglo-American nations. Furthermore, Lam et al. (1994) argue that Asian corporate management and ownership structures significantly impact the information they disclose. Fan and Wong (2002) and Joni et al. (2020) claim that a family business group controls most Indonesian listed companies, holding many critical positions and a sizable number of the shares. Moreover, Andres (2008) and Prencipe and Bar-Yosef (2011) suggest that a company's

LITERATURE REVIEW AND HYPOTHESES
Research on family businesses contends that family management exploits earnings from minority shareholders to their benefit by taking advantage of their highly concentrated ownership (Chi et al., 2015). However, Anderson and Reeb (2003) show that concentrated ownership lessens the typical issues of managerial expropriation in listed family-owned enterprises because their wealth is thoroughly correlated with business welfare. Moreover, Chen et al. (2008) and Wang (2006) argue that the active participation of family members in the company's management will lower the risk of information asymmetry and agency conflicts between owners and managers, resulting in diminished incentives to manipulate earnings or withhold information. Additionally, Andres (2008) documents that families prioritize non-financial objectives and are concerned about company reputation and survival.
This study uses the socio-emotional wealth concept to explain the behavior of family firms, which argues that the primary goal of the family-controlled firm is to protect the family's socio-emotional wealth ( (1995) show that family management and supervisory boards are frequently more interested in social and ecological problems. Indonesia has a two-tier board structure, separating the governance duties (by the management team) and oversight (by the supervisory boards). The supervisory board has given the board of commissioners the formal name. Therefore, the terms "supervisory board" and "board of commissioners" are used interchangeably throughout the paper.
Based on a literature review, this study aims to explore how family businesses and family representation on supervisory boards (the board of commissioners) and executives (management) affect the quality of carbon emission disclosure of non-financial listed firms. As a result, the hypotheses of this investigation are:

METHODOLOGY
The study uses family-owned non-financial companies publicly listed on the Indonesia Stock Exchange (IDX) from 2017 to 2019. To ensure data uniformity, the paper focuses on non-financial companies because they dominate the Indonesian economy (Craig & Diga, 1998). The family-owned business was initially highlighted in an article published in the July edition of GlobeAsia Magazine (GlobeAsia, 2019). Then, the investigation tracked down each family business group's websites and discovered that 91 companies consistently released their annual reports from 2017 to 2019. Nevertheless, 29 companies are financial institutions or do not offer sufficient information to measure the study's variables. Therefore, the useable sample of this study consists of 62 entities or 186 observations (Table 1). This study uses the carbon emission reporting level of the sampled firms as a dependent variable and utilizes a carbon emissions checklist created by Choi et al. (2013). The dependent variable is measured using the unweighted disclosure index technique, which assigns equal weight to each disclosure item. Compared to a weighted index approach, this method is less subjective and judgmental (Cooke, 1993). The study uses family ownership and family members' involvement as predictors for carbon emission disclosure. The proxy for a family firm is the proportion of a firm's outstanding shares held by family members; the minimum is 20% (Arosa et al., 2010). The paper uses two family representations and controls in a firm, considering the family member's participation in the supervisory board (board of commissioners) and executive (management). The current study considers other factors that might affect carbon emission disclosure. It includes corporate governance attributes (board size, board independence, board meeting frequency, members of audit committee, and board gender) in the analysis. , where CED = A disclosure index for carbon emission. A company scored one if it discloses information per the checklist's requirements; otherwise scored zero. FO = the proportion of a firm's outstanding shares held by family members; the minimum is 20%. FACOM = total number of family members holding supervisory (board of commissioner) positions. FADIR = the total number of family members in the executive (management) position. BSIZE = the number of board of commissioner members. BIND = the percentage of independent members on the board of commissioners. BMEET = the frequency of annual board of commissioner meetings. AUCOM = the total number of audit committee members. FEMALE = has a value of one if at least one supervisory board member is female, zero otherwise. ROA = the ratio of net income to total assets. AGE = the number of years, expressed as a natural logarithm, since the corporation was founded. it = a company i in year t. YEAR and INDUSTRY FIXED EFFECT = the industries and year-specific fixed effects.

RESULTS AND DISCUSSION
Descriptive statistics and early indications of associations between the key variables are presented in Tables 2 to 4. Table 2 exhibits a firm's disclosure percentage for each carbon emission item by each sample year. Table 2 reports the most frequently reported item: 'Identifying the board committee in charge of climate change initiatives (ACC1)', at 100% annually. Following that is 'CER1-Plans or strategy details to lower emissions' (87.10%, 87.10%, and 88.71%). The companies did not disclose the 'CER4-Future emission costs are taken into capital expenditure planning'. Table 2 also illustrates the upward trend in carbon disclosure from 32.44% in 2017 to 33.96% in 2018 and to 35.66% in 2019. These numbers show a rise in business spending on environmental initiatives. Table 3 shows the proportion of items disclosed by topic and industry classification. Companies that deal with agriculture (Industry Group 1) disclose the most (61.73%) information about carbon emissions. Companies in the basic industries and chemicals (Industry Group 3) reveal 43.52% of carbon emissions, while those in the trade, services, and investment firms (Industry Group 9) disclose emissions at the lowest rate (22.49%). On average, the accountability of carbon emissions (ACC) classification subject is the highest frequently (88.71%) disclosed by companies in Industry Groups 3 and 7, respectively. The second and third highest disclosures are CCR = Climate change: risks and opportunities (47.58%) and ECA = Energy consumption accounting (42.83%) themes. Interestingly, the sample firms' disclosures for the CEA = Carbon emissions accounting theme are the lowest. With a mean of 34.02%, the carbon emission disclosure score ranges from 22.49% (Trade, services, and investment = Industry Group 9) to 61.73% (Agriculture = Industry Group 1). Table 4 summarizes descriptive statistics and correlations between variables in the analysis. According to Table 4, family members own an average of 62.43% of the sample company's shares. The average number of family members on the supervisory board (board of commissioner) and executive (management) positions is one, respectively. The average number of board commissioner members is five. The percentage of independent  Additionally, the findings show a weak correlation between the independent variables. Therefore, the variables have no severe multicollinearity problems (Cooper & Schindler, 2003). However, the study assesses variance inflation factor (VIF) values to confirm whether multicollinearity issues existed in the regression model (Table 5).    The results also support the argument that the family board of commis-sioner members frequently has a greater interest in ecological matters (Ibrahim & Angelidis, 1995). Models III and IV show that the FADIR coefficient is positive but statistically insignificant, suggesting that family members in managerial or leadership positions do not influence carbon emission performance. Therefore, H 3 is rejected.
This study finds that BSIZE and FEMALE help explain carbon emission performance. The regression coefficients for BSIZE are all positively and significantly (p < 0.01 and p < 0.05) related to CED. The results suggest that supervisory boards with more significant members are more effective at setting CED agendas and promoting the communication of CED information to meet social needs  general, women are more aware of others, sensitive to social needs, and stakeholder-oriented, as well as concerned about ethical issues than men (Tate & Yang, 2015). The remaining five control variables (BIND, BMEET, AUCOM, ROA, and AGE) do not significantly affect CED.
This study conducts several additional analyses to strengthen the reliability of the main results. Firstly, the study examines the non-linear effects of family ownership on carbon emissions. Anderson and Reeb (2003) state that family-run businesses can positively or negatively affect firm performance. Companiy performance might improve with greater family ownership. However, the relationship between the two variables can be negative if family ownership is low. The association between family-controlled businesses and carbon emission practices is predicted to be non-linear by these two hypotheses.
Similarly, Boone et al. (2007) argue that the size of the supervisory board represents a trade-off between the firm-specific benefit and cost of monitoring. Therefore, many empirical studies have tried to find the optimal size of a company's board of commissioners. According to Lipton and Lorch (1992), a board should have at most ten members to function optimally and to be less susceptible to manipulation by the assigned commissioner. In comparison, Jensen (1993) suggests that a board should have at most eight commissioners. Table 6 displays the test results for the impact of non-linearities of family ownership, board of commissioners' size, and carbon emission disclosure.
The coefficients of FO and its square reported by Models I and III are negative and positive (p < 0.10 and p < 0.05), demonstrating a non-linear association between family entities and the disclosure of carbon emissions. Family ownership can lead to different behaviors. The increase in family ownership to 53.1% (the inflection point is not reported) supports the expropriation hypothesis. Beyond this threshold, however, the emission quality improved, supporting the monitoring hypothesis. In other words, carbon emission performance decreases as family share ownership rises to 53.1% but increases when family equity exceeds this cutoff point. The positive and negative coefficients for BSIZE and BSIZE-Square (see Models II and III) indicate that the supervisory board size and carbon emission information are not linearly related. However, these coefficients are statistically insignificant; thus, the results do not confirm a non-linear relationship between the board of commissioners' size and carbon emission performance. In the second test, this study excludes firms from a large industry cluster (Property, real estate, and building construction, which comprise 29.03% of the sample size) to ensure that no industry sector dominates the main findings (see Model I of Table  7). The study also analyzes whether alternative tests for disclosing carbon emissions are robust to a sample of companies engaged in environmentally sensitive sectors (see Table 7, Models II and III). Three sectors (Infrastructure, utilities and transportation, Basic and chemicals, and Mining) are designated as polluting under Indonesian Law No. 32/2009. The polluting firms are thus a binary variable with a value of one if included in one of these three industry classifications and zero otherwise.
Model I of Table 7 shows that the coefficients of the FO and FACOM are statistically significant (p < 0.05 and p < 0.01) and in the same direction as presented in Table 5. Also, the coefficient on FADIR is positive but statistically insignificant. Therefore, these findings support the main results of the re-gression analysis presented in Table 5. In conclusion, the main results summarized in Table 5 do not drive by an individual industry. Model III of Table  7 indicates a significant influence of family-owned and family supervisory boards on carbon emissions, only robust in non-polluted firms' classification. These findings support Baalouch et al. (2019), who suggest that family-controlled firms and supervisory boards do not primarily use environmental reporting to justify their actions and strengthen their standing among various stakeholder groups.
These findings are significant because they demonstrate how crucial it is to understand the relationship between socio-emotional wealth precepts and families to comprehend the carbon disclosure of Indonesian corporations. The evidence that family entities disclose more information on carbon emissions suggests that families' socio-emotional wealth agenda aligns with the Indonesian government's carbon emission reduction and accountability strategies.

CONCLUSION
This study explores the impact of family businesses and active family members (in the supervisory boards or management) on carbon emission disclosure in Indonesia during 2017-2019. The result shows a significant positive link between family ownership and the disclosure of carbon emissions. The findings align with socio-emotional wealth, which claims that family-owned enterprises frequently have