Factors influencing levels of CSR disclosure by forestry companies in China

: With the international community’s increasing concern for social and environmental problems, the fulﬁlment and disclosure of corporate social responsibility (CSR) has been advocated and promoted across the world. Forestry companies, which are particularly sensitive to environmental and social issues, are increasingly developing and improving their levels of CSR disclosure. However, information on emerging country contexts is still lacking. To ﬁll this gap, this study focuses on Chinese forestry companies’ CSR disclosure and introduces new disclosure indices through content analysis of annual reports by listed companies between 2011 and 2015. It then builds a correlation analysis of the factors inﬂuencing these companies’ disclosure indices in order to gain a better understanding of the current situation for CSR implementation by forestry companies in emerging economies like China. Although context-speciﬁc, our ﬁndings can provide a reference for researchers and policy makers, and promote sustainable development via improved CSR disclosure by forestry companies, especially in developing regions.


Introduction
In recent years, with the international community's increasing concern for social and environmental problems, a wide range of stakeholders and issues apart from meeting core shareholder interests, has become increasingly influential in determining the long-term viability of businesses [1,2].Consequently, the improvement and disclosure of corporate social responsibility (CSR) has been advocated and promoted across the world [3,4].International institutions, such as the Global Reporting Initiative (GRI) and ISO26000, have encouraged enterprises to undertake their due share of economic, environmental, and social responsibilities (GRI, in particular, is among the most common guidelines for CSR disclosure globally [5,6].In 1999, the United Nations Environment Program (UNEP) promoted the GRI initiative and, in 2002, officially recognized it.In 2002In , 2006In , and 2013, the GRI released the revised versions of Sustainability Reporting Guidelines, respectively, G2, G3, and G4).This has exerted a positive influence on CSR disclosure on a global scale [7].CSR disclosure is particularly salient in the forest sector, which faces the formidable challenge of reconciling growing fibre demand, while guaranteeing the sustainable management of forest ecosystem services, including water resources, climate control, and cultural values [1,[8][9][10][11][12].CSR in the forest sector has been driven largely by legitimacy pressures [13][14][15][16][17][18].In the context of emerging economies like China,

Literature Review and Hypothesis Development
Stakeholder management theory [42,43] states that managers organize diverse information depending on multiple stakeholders.Traditionally, publicly owned companies have mainly been accountable to their shareholders, and their principal aim has been to pursue value maximization.Stakeholder management theory, however, proposes that the long-term viability of companies also needs to include the concerns of multiple stakeholders who are influential or important for the company, including customers, employees, suppliers, local communities, governments, and environmental groups [44,45].Indeed, with the development of standardized reporting guidelines, such as Global Reporting Initiative, CSR disclosure in recent decades has reflected this notion by becoming more inclusive towards all stakeholders' needs [46].All that said, this may not be the case for China.Previous studies have shown that Chinese companies tend to be more concerned with whether or not they should release CSR reports and, if so, how much information should be contained within the reports [47].
This study aims at exploring the factors influencing the CSR disclosure by Chinese forestry enterprises in light of four research hypotheses enumerated below.Hypotheses are developed in accordance with current literature on the topic, and are enumerated below.
The larger a company is, the more likely it is to draw attention from government regulatory institutions, environmental protection organizations, the media, and other social groups [7].Larger companies, due to their higher visibility, are more likely to disclose information on their CSR to demonstrate that they are socially responsible enterprises.Research on factors influencing CSR disclosure indeed show that there is a positive correlation between the firm size and CSR disclosure levels [25,26,[29][30][31][48][49][50][51][52].For example, Li et al. [23] and Han and Hansen [26] found that larger, international forest companies are more likely to implement a broader range of CSR activities than smaller firms.We, therefore, hypothesize that: Hypothesis 1 (H1).Corporate social responsibility disclosure is positively associated with firm size.
Management theories postulate that CSR performance is positively correlated to corporate financial performance [5].However, there is little consensus on the correlation between CSR disclosure and profitability.Some empirical studies indicate that enterprise profit is closely correlated to levels of CSR disclosure [30,53,54], while other research (e.g., McWilliams and Siegel [55]) has shown that there is, in fact, a negative correlation or no clear correlation between CSR disclosure and profitability of a company.We, therefore, hypothesize that: Hypothesis 2 (H2).Corporate social responsibility disclosure is positively associated with profitability.
Previous studies in other countries and regions have shown that levels of CSR disclosure are correlated to equity concentration; companies that are publicly owned are more likely to have a CSR program in place and to release information related to CSR activities.[52,56,57].We, therefore, hypothesize that: Hypothesis 3 (H3).Corporate social responsibility disclosure is positively associated with equity concentration.
Finally, previous research suggests that CSR activities and disclosure are dependent on the availability of financial resources within a company [53,56].A company's financial position relates to the degree of financial leverage available to them [50].We, therefore, hypothesize that: Hypothesis 4 (H4).Corporate social responsibility disclosure is positively associated with financial leverage.

Data Collection
In order to conduct a systematic qualitative and quantitative analysis, the study sample comprises 42 publicly listed forestry companies, including, among others, well-known companies like Jilin Forest Industry Group (Changchun, China), Dare Global Co., Ltd.(Danyang, China), Fenglin Group Co., Ltd.(Nanning, China), Dehua Tubaobao Co., Ltd.(Huzhou, China).The companies in this analysis are among the first in China to release CSR disclosure reports, either voluntarily or under the auspices of the CFIA and CNFPIA.Due to a lack of consistency, the separate CSR reports were not used in this analysis; instead, the source of data was their annual reports.The companies were selected from those listed by the Shenzhen Stock Exchange [58], Shanghai Stock Exchange [59], and CNINF [60] (a website with information disclosure for listed companies in China).Thus, the data came from 209 annual reports published by these 42 companies during the period 2011-2015 (The 2011 annual report is missing for one company, Sleemon Co., Ltd.).Appendix A shows the reports for Chinese forestry companies included in this study.

Analysis
This paper adopts a qualitative content analysis and quantitative correlation analysis to explain factors influencing CSR disclosure levels by Chinese forestry enterprises.The analysis was conducted as follows.First, based on the literature reviewed, a working framework was created (Table 1) to assist in the categorization of CSR disclosure information.This framework categorized the disclosed CSR information into different themes, each subdivided into several dimensions in order to further analyze CSR disclosure information.
Second, the CSR reports themselves were analyzed using a content analysis.Content analysis is performed by thoroughly reading through texts and developing suitable codes that are consolidated into categories describing the text.During this process, relevant literature is used to support the interpretation of the data and the development of the codes [61][62][63][64].Consequently, the data was coded first based on the framework (Table 1), and then with a refined category system formed by using a data-driven (abductive) approach [65].The entire process was performed iteratively.In other words, the texts of the CSR reports were analyzed with NVivo 11.0 software, and when the category system was created, a process of manually going back and forth between the text and the framework was implemented.As a measure of reliability of content analyses, previous studies [61,63] generally refer to three characteristics: stability, reproducibility, and accuracy.To ensure reliability, we based our content analysis on two CSR disclosure guidelines released by Chinese Academy of Social Sciences and Chinese National Forestry Products Industry Association (CASS 3.0 and CNFPIA 2.0); these can be considered stable and currently valid frameworks, and the codes for this study are reproducible.
Third, the quality of CSR information disclosed in enterprises' reports was quantified by assigning scores to each specific item.This method (i.e., the application of disclosure index) has been widely used in CSR disclosure studies [23,26].The scores were assigned according to the following criteria: 0 points for non-disclosure, 1 point for qualitative disclosure, and 2 points for quantitative disclosure.
Fourth, a Pearson correlation analysis using SPSS 24.0 software was run to investigate the influence of companies' firm size, ROE (return on equity, a measure of profitability), equity concentration, and leverage on CSR disclosure level.Correlation analysis was used over regression because the intent of this analysis was merely to determine association.
Table 1 shows the categorization of CSR disclosure information based on CASS 3.0 and CNFPIA 2.0, released by Chinese Academy of Social Sciences and Chinese National Forestry Products Industry Association.This framework represents currently accepted standards for Chinese forestry companies disclosing information and, consequently, serves as the foundation for this study.Table 2 shows the variables considered in the correlation analysis, including firm size, return on equity (ROE), leverage, and equity concentration.Firm size was computed using the natural logarithm of total assets of the companies in order to reduce the variance.ROE is the reported return on equity of the sample companies, while leverage refers to debt ratio of the companies and equity concentration is the proportion of the firms' ownership controlled by common shareholders.

CSR Disclosure Levels
A descriptive analysis of CSR disclosure indices (CSRDI, i.e., the qualitative summative rating of information disclosed in the reports for different items) was conducted for the 42 Chinese forestry companies in this sample based on their annual reports (Table 2).The average total score for disclosure is 37.07, with the highest and lowest scores being 60 and 23, respectively, indicating a considerable gap between different disclosure strategies.Among the assessed dimensions, information on environment ranks highest, with a CSR disclosure index of 8.18.The shareholders (8.00) and employees (7.37) dimensions rank second and third, respectively, followed by government (5.38),customers (3.50), and community (2.58).The bottom of the list is rounded out by suppliers (2.06).
In terms of variance, the shareholders dimension has the lowest standard deviation (0.00), indicating a uniformity of information disclosed among reports.This, coupled with the mean the shareholders dimension suggest that the companies have maintained stable high levels of information disclosure on this topic.The environment dimension has the highest standard deviation (4.40), reflecting a greater variation of the information disclosed.The minimum CSDRI value for the environment dimension is 0, revealing that some listed forestry companies have not discussed environmentally-related issues at all in their annual reports.The standard deviations of other dimensions range from 1.01 to 1.84, suggesting little disparity among the companies with regards to information disclosure levels on customers, employees, suppliers, the community and the government dimensions.Importantly, all companies equally seem to provide very little or no information disclosure on customers, suppliers, and the community dimensions.
Figure 1 reflects the development of the CSDRI on the seven dimensions outlined in Table 3 over a period of five years (2011)(2012)(2013)(2014)(2015).The data show a steadily rising tendency in the levels of information disclosed by companies, from an average total of 35.25 in 2011 to 38.34 CSDRI in 2015.For the specific dimensions, disclosure of shareholder information has been stable and comparatively high, while environmental responsibility information grew, especially in 2012, when it surpassed shareholder levels and remained at the highest level.Disclosure information on customers, communities, and suppliers are comparatively lower than the other dimensions for each of the years under study.

Influence of Company Characteristics on CSR Disclosure Levels
Simple correlation analysis was run to determine the influence of firm's background characteristics on the CSR disclosure level [33] for the year 2015.Table 4 provides summary statistics computed for the variables in this analysis.In order to conduct the correlation analysis, the log of CSRDI was taken to transform nonnormality of data.Table 5 shows the resulting coefficients of correlation between dependent and independent variables.The correlation coefficients indicate that CSR disclosure has positive and significant relationships with firm size and equity concentration.While the relationships with ROE and leverage are negative, they are statistically insignificant.This reveals that firm size and equity concentration are positively correlated with the disclosure levels among the sample companies, and that ROE and leverage had no significant impacts on disclosure levels.

Influence of Company Characteristics on CSR Disclosure Levels
Simple correlation analysis was run to determine the influence of firm's background characteristics on the CSR disclosure level [33] for the year 2015.Table 4 provides summary statistics computed for the variables in this analysis.In order to conduct the correlation analysis, the log of CSRDI was taken to transform non-normality of data.Table 5 shows the resulting coefficients of correlation between dependent and independent variables.The correlation coefficients indicate that CSR disclosure has positive and significant relationships with firm size and equity concentration.While the relationships with ROE and leverage are negative, they are statistically insignificant.This reveals that firm size and equity concentration are positively correlated with the disclosure levels among the sample companies, and that ROE and leverage had no significant impacts on disclosure levels.

Discussion and Conclusions
This study analyses the disclosure development of CSR reports published by the leading listed Chinese forestry companies from 2011 to 2015.The focus of this research was on company characteristics so as to further probe the influencing factors of CSR disclosure by forestry enterprises in an emerging economy context.The results show that the CSR disclosure by Chinese forestry enterprises is still in a nascent stage, dealing primarily with economic, social, and environmental dimensions, but in varying degrees of depth.During the period of study, however, the scope and quality of CSR disclosures has expanded and improved.
Based on a descriptive analysis of summative CSR disclosure indices, the companies in this study seem to have maintained a relatively stable and comparatively high level of stakeholder information disclosure.Disclosure on environmental issues and employees is also relatively commonplace.There are, however, significant gaps in the provision of environmental information among companies, with some enterprises not disclosing any information at all.Finally, disclosure of information related to suppliers, customers, and the community is generally sparse.
These findings are in line with other international studies on CSR in the forest sector, which show that disclosure is particularly focused on environmental dimension and, to a certain degree, other economic and social issues [1,2,26] (see also a review by Ranängen and Zobel [66]).Previous studies [1,2] have pointed out that Asian forestry companies are typically more concerned with environmental performance, in particular the control of air emissions, energy efficiency, and recycling programs, compared to other parts of the world: health programs prominently featured in Africa and European companies considering CSR activities in a broader sense from both environmental and social perspectives.This study, however, shows that there is still space for improvement for Chinese forestry enterprises in terms of CSR disclosure, since they lag behind their counterparts in developed economies.That said, it should be noted that companies in emerging economies have tended to adopt disclosure frameworks appropriate to their actual conditions, which may explain global variation in the types and frequency of information that is released for the public [67].This study also indicates, in line with previous studies [1,2], that Chinese forestry companies are increasing the amount of information that they put out on economic and social dimensions.Although the information disclosed may not be particularly significant at this juncture, it seems that CSR reporting in Chinese forestry companies is indeed moving towards a broader and more inclusive consideration of sustainability.That said, it is vital to keep in mind that this study reflects the characteristics, practices and reporting of only 42 Chinese forestry companies out of thousands.While the forestry companies in this study are among the largest and well-known forestry companies in China, signifying that they are considered business leaders, potentially, our sample could be biased towards early adopters of CSR practices and companies that are more susceptible to public criticism [2,10].In other words, this analysis does not provide conclusive evidence that CSR disclosure in the Chinese forestry sector has been widely adopted.However, the adoption of CSR disclosure by these companies may catalyze others in the sector to follow suit.These findings are interesting in comparison to Meng et al. [7], which found that environmentallyaware enterprises in China, such as those analyzed in this study, tend to produce more thorough CSR information.Two possible reasons for this trend may be as follows.First, the policy of the Chinese government has emphasized environmental disclosure information, with the new environmental laws (2015) adding an entire chapter on this topic, including explanations and requirements for disclosure and public participation.Second, in 2015, the Chinese Forestry Industry Federation (CFIF) and the Chinese National Forestry Products Industry Association (CNFPIA) included detailed rules for the forestry industry on how to engage in corporate social responsibility [34].
Our study also reveals that the two major firm characteristics influencing levels of CSR disclosure by Chinese forestry enterprises in 2015 are firm size and equity concentration, both of which are positively correlated with the disclosure levels among the analyzed companies.The other variables under investigation, such as ROE (measured by return on equity) and leverage (measured by debt ratio), had no discernable impacts on disclosure levels.Therefore, our hypotheses H1 and H4 are confirmed, while and H2 and H3 are rejected.This is in contrast to other studies that have been conducted in other industry contexts on CSR activities of Chinese companies [68].That said, the results of this study are in line with prior studies regarding the role of firm size on influencing CSR disclosure [23,26,66,68].
Results also show no significant association between firm profitability and CSR disclosure.This correlation has often been proposed in theoretical and empirical research on CSR, but there is no consensus on whether such a correlation is characterized as positive or negative.That said, some management theories postulate that CSR disclosure is positively correlated to corporate profitability [3,20,36] because better performing companies are more likely to engage with CSR, and because firms engaging with CSR are more likely to gain indirect strategic or financial benefits [47,67].Other studies [15,18,53] at the international level also show that there is a positive correlation between profit and CSR activities.This phenomenon was not borne out in this context of the Chinese forestry sector.A possible explanation of this could be that positive/negative correlations occur between CSR and, respectively, long/short-term profitability [67].CSR-compliant firms may incur immediate costs, but enjoy reduced costs in the longer term.Less proactive firms, instead, may avoid initial costs, but pay a higher price over time for neglecting their CSR.Following this logic, current investments in improving CSR in the forest sector in China could bring about financial benefits for compliant firms, but they have yet to materialize.
As a shortcoming, our sample included only listed Chinese forestry companies with their annual reports being used as the data source.Consequently, the results cannot be generalized to a vast number of existing small-scale forestry enterprise in China, or even the Chinese forestry sector as a whole.Furthermore, the reporting provided by different companies in this study is very heterogeneous in style and content.Due to the small number of listed forestry companies which have released annual reports, as well as separate CSR disclosure reports, this investigation can be thought of as a case study of such companies rather than a systematic analysis.It is, in fact, possible that some CSR practices are not reflected or reported in annual reports or that others are over-reported [69].All that said, the use of annual reports as a data source is a valid means of systematizing data collection and the framework for the information disclosed can be considered reliable (e.g., Li et al. [23], Lähtinen et al. [24]).In the future, as CSR disclosure reports in China become increasingly available, future research could attempt an even more systematic analysis, with additional information being captured through other types of documents or media, such as company websites, to create a predictive model to explore causality.A further focus for research could be to gain a better understanding of CSR practices by small and medium-sized forestry enterprises, both in China and in other emerging economies.Since the forest companies in this study are comprised of state-owned companies and privately-owned companies, it would also be interesting to shed further light on the impacts of ownership structure [70][71][72], especially in China.The role of political interventions on CSR disclosure, especially relevant in emerging economies like China [67], could also be investigated.The disclosure index of each dimension is the mean of the sample companies of each year from 2011-2015.

Table 1 .
Framework for the categorization of CSR disclosure information.quantity, kind and risk to human and environment of toxic or exhaust emission CNFPIA 2.0, 2013 Research, development, application and sale of the environment production and devices CASS 3.0, 2013; CNFPIA 2.0, 2015 Energy resources conservation CASS 3.0, 2013; CNFPIA 2.0, 2015 Reduce pollution and decrease drain CASS 3.0, 2013; CNFPIA 2.0, 2015 Ecology restoration CNFPIA 2.0, 2015 Volunteer working for environment protection CNFPIA 2.0, 2015 Government Enterprise management abided by rule CASS 3.0, Tax payment CASS 3.0, Employment security policy CASS 3.0, Employment amount over the report periods CASS 3.0, CASS: Chinese Academy of Social Sciences; CNFPIA: Chinese National Forestry Products Industry Association.

Table 2 .
Descriptions of variables.

Table 3 .
Descriptive statistics of corporate social responsibility disclosure indices (CSRDI).

Table 3 .
Descriptive statistics of corporate social responsibility disclosure indices (CSRDI).