1
RUG
MSc thesis IFM
Environmental performance, environmental
disclosure and institutional ownership: An
empirical analysis
In the light of the increased importance of environmental practices and accounting, and the enactment of Paris Agreement in 2015 to combat climate change, this study examines the relationship between environmental performance and environmental disclosure and the effect of institutional ownership. The amount of direct and indirect GHG emissions is used as a proxy for environmental performance, due to its comparability and fundamental influence of companies on the environment. A environmental disclosure score from ASSET4 is employed for a measure of environmental disclosure. For institutional ownership, I will categorize institutional ownership into potentially active and potentially passive in terms of monitoring power. Based on a longitudinal dataset of 356 firms listed in the MSCI Global Index for the period 2013-2019, I find empirical evidence that there is no correlation between the amount of GHG emissions and environmental disclosure. Moreover, active and passive institutional ownership does not appear to have an effect on either environmental disclosure and performance. Student number: S2558130
Name: Wesley Hop
Supervisor: prof. dr. L.J.R. Scholtens Study Programme: MSc IFM
2
1. Introduction
Ample literature has investigated the relationship between environmental performance and environmental disclosure, but found mixed results and in which studies used small samples (Giannarakis et al., 2017; Manning et al., 2018), focused on one country (Graafland and Smid, 2014; Luo and Tang, 2014; Yu and Lee, 2017) and solely used cross-sectional analysis (Clarkson, 2008; Cotter and Najah, 2012). Moreover, most studies have been conducted based on samples of firms from Anglo-Saxon countries (English speaking) (Clarkson, 2008; Giannarakis et al., 2017; Graves and Waddock, 1994; Walls et al., 2012). In this study, a longitudinal analysis of a seven year global sample of firms listed in the MSCI Global Index is done to investigate the relationship of corporate environmental performance and corporate environmental disclosure. The importance of sustainability guidelines and goals has increased in recent years and in 2015, the UN Climate Conference adopted the Paris Agreement (UNFCCC). This agreement outlined future global actions and goals to combat climate change, where the main goal is to keep the global average temperature below 2°C and to pursue efforts to limit temperature increase to 1.5°C above pre-industrial levels. Thus, I will use a time-period of 2013 to 2019, because findings of this recent time-period increase the understanding of motives of managers and firms and help to create guidelines for businesses to improve environmental disclosure and performance.
3 develop and use a categorization of institutional ownership to test its relationship with environmental performance and disclosure. As a final part of my study, a separation is made between firms originated in Anglo-Saxon and non Anglo-Saxon countries. These types differ in terms of laws, regulations, ownership structures and government involvement, which might affect the institutional ownership - environmental disclosure/performance relationship.
4 crucial for these stakeholders, because it enhances the rationality of their expectations for environmental responsibility (Giannarakis et al., 2017). Firms communicate their environmental practices through their environmental reports.
In this paper, ASSET4’s environmental disclosure score, imported from Eikon, is used to measure corporate environmental disclosure. This score consists of 51 data points in three categories: emission reduction, product innovation and resource reduction and is received from firms’ annual reports. Some papers (Clarkson et al., 2008; Graafland and Smid, 2014) used self-made indices to measure environmental disclosure, which can be valuable for studies focused on specific analysis (e.g. focus on one country or ability of implementation of corporate social responsibility (CSR)), but suffer from comparability issues. Data from ASSET4’s environmental score is available for a wide amount of large firms and industries, and covers a large time-period.
5 legitimacy theory is supporting the latter relationship, which argues that bad environmental performers can use disclosure to legitimize and justify their actions and conform with government environmental regulations. In line with this theory, Graafland and Smid (2014) and Cadez et al. (2018) stated that environmental disclosure can possibly be used as ‘window dressing’ or ‘greenwashing’.
Along with environmental performance and environmental disclosure, institutional ownership has also shifted toward social issues that are relevant for stakeholders (climate change, labor rights and corruption) (Walls et al., 2012). Fragmented and inconclusive evidence has made theory building difficult. For example, some studies find a positive relationship (Cotter and Najah, 2012; Ho and Tower, 2011; Manning et al., 2018), while others show insignificant results (Brown et al., 2006; Calza et al., 2016; Graves and Waddock, 1994; Lo, 2014). Among others, Ho and Tower (2011) state that the positive relationship between institutional investors and environmental disclosure is explained by the agency theory, because they can push firms to disclose more information. However, according to Walls et al. (2012), the agency theory falls short in explaining this. Further study of this relationship is needed to gain an enhanced understanding of the subject.
6 categorization can lead to invalid results and implications. This paper will focus on the ability of institutions to use their monitoring and voting power to influence managerial decision-making. Institutions differ across their legal, regulatory, competitive environments and investment strategies (Bennett et al., 2003). These institutions also differ in terms of their monitoring costs. In this study, and also following Almazan et al. (2005), I will label these institutions as potentially active or potentially passive (1), based on the difference in monitoring costs and institution characteristics. Active institutions are more likely to have more skilled employees, are less strictly regulated on their investments and have a lower potential for business relations with the company opposed to passive institutions. The ability for active institutions to monitor and exercise power over company management, allows them to have a significant impact on decision-making and sustainability practices. Active institutions are able to earn higher amounts of profit from monitoring and exercising power due to the significant amounts of capital that are at stake (Crespi and Renneboog, 2010). Institutions that fall under this category are investment institutions, investment advisors and pension funds. These institutions gain profits from portfolios that include stocks from multiple firms, which is their primary source of income. These institutions do not have personal relations with management (Almazan et al., 2005). On the contrary, passive institutions are primarily bank trust funds, research firms and insurance companies. These institutions are generally expected to be passive in their monitoring and voting power (Crespi and Renneboog, 2010). Besides this, they are more dependent on other
7 sources of income rather than depending on stock portfolios and possibly want to sustain a healthy relationship with a firm, because the firm might provide financial services for them.
Accordingly, I can provide evidence for the two main research questions in this paper:
What is the relationship between corporate environmental disclosure and corporate environmental performance?
What is the effect of active and passive institutional ownership on both corporate environmental disclosure and corporate environmental performance?
Contributions of this study to the literature are as follows. This paper is one of the few papers that investigates environmental disclosure and performance based on a longitudinal, global dataset, in which I use GHG emission and a 51-point environmental disclosure score as proxies for environmental performance and disclosure respectively, because of high comparability. With the creation of the two groups active and passive institutional ownership I can highlight possible effects and differences in monitoring power.
8 environmental performance can highlight variances in monitoring power and influence on decision-making, which leads to implications for the involvement of institutional ownership in sustainability practices.
With this thesis, I contribute to all three aspects in the field of international financial management. The international aspect is covered in this study by using a global sample and making a distinction between firms from Anglo-Saxon countries and from non-Anglo-Saxon countries. For finance, findings will lead to an increased understanding of the relationship between environmental disclosure and performance for stakeholders, which is valuable for their investing strategies and choices. This holds for institutional stakeholders in particular, because the effect of their monitoring power on environmental disclosure and performance is included in this study. In the field of management, findings will gain insight for managers on the effect of institutional pressures on sustainability engagement.
The remainder of this paper is structured as follows: the next section provides the theoretical framework and hypotheses development related to environmental disclosure, institutional ownership and environmental performance. Then, section 3 presents the data and methodology. Section 4 covers the results and discussion. And finally, section 5 contains a conclusion, implications and opportunities for future research.
2. Theoretical framework and hypotheses
2.1 Environmental disclosure and environmental performance
11 Besides these positive and negative relationships, studies also found no association between environmental disclosure and environmental performance (Fekrat et al., 1996; Freedman and Jaggi, 2011). One reason that can explain the mixed results in this research area is the use of different proxies for environmental disclosure and performance. But even while using the same proxies in some cases, there are still mixed results. For example, by using GHG emissions as a proxy for environmental performance and the CDP for environmental disclosure, Freedman and Jaggi (2005) investigated electric utilities in the US and found that firms emitting the highest carbon dioxide emissions also present the most extensive disclosures, which is similar to the results of Luo and Tang (2014). However, Freedman and Jaggi (2011) extended their own research of 2005, by also including the EU, Japan and Canada and revealed that there is no significant relationship between disclosures and carbon emissions. Also within studies mixed results were found (Hassan and Romilly, 2017, Manning et al. 2018). Hassan and Romilly (2017) found a highly positive and significant relationship for the full sample by using GHG emissions as a proxy for environmental performance, but they also found a negative significant relationship when only looking at developed countries. A model with a sample of solely developing countries remained significantly positive. For environmental disclosure they used a environmental disclosure score from Bloomberg, which measures the most commonly disclosed fields of environmental information on a 1-100 scale. Manning et al. (2018), using a sample of Dutch firms during the years 2012-2016, found evidence for both the signaling and the legitimacy theory.
12 1A: environmental disclosure is positively related to environmental performance (lower GHG emissions), based on the voluntary disclosure theory
1B: environmental disclosure is negatively related to environmental performance(higher GHG emissions), based on the legitimacy theory.
2.2 Institutional ownership, environmental disclosure and environmental performance 2.2.1 Institutional ownership
13 strategies, which will lead to improved environmental performance. Focused on environmental disclosure, Ho and Tower (2011) used institutional ownership as an explanatory variable and concluded that, based on an eleven-year period of 1996-2006, institutional shareholders influence managerial decision making and increase voluntary disclosure for Malaysian firms.
Cotter and Najah (2012) support this by finding a positive association of institutional investors and climate change disclosure. Studies also find institutional ownership to be an important determinant of corporate environmental performance (Niedertscheider et al., 2018; Wang et al., 2019). Walls et al. 2012 only found evidence of institutional ownership when environmental performance was low, which can possibly be explained by the different types of investment institutions that exist.
14
2.2.2 Active vs. Passive institutional ownership
15 effect of active stakeholder engagement on sustainability reporting standards and sustainability reporting quality, indicating that active shareholders, over time, are able to push firms into more sustainable business practices. Passive institutions are not expected to have such effects. They are reluctant to influence decision-making, since it could negatively influence the relationship between managers and these institutions. Banks and trust funds, insurance companies and research firms are good examples within this category. Besides income from investing, they also gain income from other sources of income, such as financial services and information. Compared to active institutions, passive institutions can be seen as institutions that have a two-way interest with firms. Based on these findings and characteristics of institions I develop the following hypotheses:
H2: Active institutional ownership has a positive and significant effect on environmental performance
H3: Passive institutional ownership has no significant effect on environmental performance
H4: Active institutional ownership has a positive and significant effect on environmental disclosure
H5: Passive institutional ownership has no significant effect on environmental disclosure
2.2.3 Anglo-Saxon and non Anglo-Saxon countries
16 differences of institutional ownership between Anglo-Saxon countries and non Anglo-Saxon countries. The origin of the Anglo-Saxon model can be dated back to the views and ideas of Adam Smith. Countries that use this model have fewer laws and regulations compared to non Anglo-Saxon countries. They use the common law system, which operates with lay judges, broader legal principles and oral arguments. On the other hand, civil law systems, used in non AS countries, rely on written statutes and professional judges. In these countries, ‘ownership structures and governance mechanisms present well-established peculiarities, because there is often less separation between ownership and control, and there are frequently dominant blockholders among the owners of the firm’ (Calza et al., 2016, pp 376). They conclude that the role and influence of institutional investors in these countries is not as relevant as in Anglo-Saxon countries and that institutional investors in non Anglo-Anglo-Saxon countries are traditional and focus on short-term activities. Another paper that focused on the influence of different institutional environments on sustainability policies stated that Anglo-Saxon countries score higher on most CSR, including environmental, dimensions (Jackson and Apostolakou, 2009). Therefore, I expect that active institutional investors in Anglo-Saxon countries have a higher effect on environmental performance compared to investors from non Anglo-Saxon countries and create the following hypotheses:
H6: Active institutional ownership has a bigger effect on environmental performance in Anglo-Saxon countries compared to non Anglo-Saxon countries.
17
3. Methodology and research design
3.1 Model
18 of serial correlation and heteroskedasticity, so both will be controlled for by clustering standard errors. The two equations are constructed as follows:
1
2
3.2 Variables
Industry and year fixed effects are left out of the equations for simplicity. The first equation presents environmental performance as a dependent variable, used to test hypotheses H1-H3. The second equation presents environmental disclosure as a dependent variable and is used to investigate institutional pressures and environmental disclosure. With this equation I can test hypotheses H4 and H5. H6 and H7 are tested by using both equations for the two subsamples. The models show similarities with models of Hassan and Romilly (2017), Walls et al. (2012) and Almazan et al. (2005), from which I use elements I consider relevant and valuable for this study.
19 et al., 2018). Luo and Tang (2014) and Yu and Lee (2017) suggest the combination of using 2 indicators of GHG emissions, where scope 1 measures direct GHG emissions and scope 2 indirect emissions. GHG scope 1 and scope 2 can be found on Thomson Reuters Eikon and will be added together as total GHG emissions in tonnes for the model. The following gases are included in GHG in Eikon’s database: carbon dioxide (CO2), methane (CH4), nitrous dioxide (N2O), hydrofluorocarbons (HFCS), perfluorinated compound (PFCS), sulfur hexafluoride (SF6) and nitrogen trifluoride (NF3).
20 environmental disclosure (Bernardi and Stark, 2016; Hassan and Romilly, 2017; Nollet et al., 2016)
For the measurement of active and passive institutional investors, I retrieved data from the Eikon database. This database provided institutional ownership data as a percentage of total shares outstanding and also divides institutional ownership into multiple types, which is important for this study. I retrieved the eight largest types that can be defined as (potentially) active or passive, which are banks and trust funds, hedge funds, insurance companies, investment advisors, investment advisors/hedge funds, pension funds, research firms and sovereign wealth funds. This distinction of institutions is made in their legal, regulatory, competitive environments and investment strategies (Bennett et al., 2003). Other types will not be included in this study, because they cannot be defined as either active/passive or are too small to have a statistical meaning in the model. Accordingly, I divide these 8 types in two groups, active institutions and passive institutions, based on definitions of Eikon and Almazan et al. (2005).
21 Hence, this category is placed in the group of active institutions. Research firms and sovereign wealth funds are added to passive institutions, based on definitions and categorization made in Eikon.
It is expected that both active and passive institutional ownership will not have an immediate effect on environmental disclosure and performance the same year, because institutional pressures are not implemented in disclosure and performance instantly. Instead, I expect that institutional pressures will have an effect on disclosure and performance the next year, so active and passive institutional ownership will be lagged by one year. This approach is also taken by Almazan et al. (2005).
For control variables, I use variables that are used in previous literature using CO2/GHG emissions and environmental disclosure (e.g. Calza et al., 2014; Hassan and Romilly, 2017; Giannarakis et al., 2017; Walls et al., 2012; Wang et al., 2019). Variables related to firm characteristics are leverage (LEV), firm size (SIZE), return on assets (ROA), capital expenditure (CAPEX) and sales growth (SALESG). Leverage ratio (LEV) is included as an indicator of the firm’s financial status. A firm with a high leverage ratio is more prone to financial constraints and risks. ROA is a proxy for the firm performance. CAPEX is used as a proxy for the efficiency of the firm in the use of its assets and sales growth to measure increased resources.
22 (Hassan and Romilly, 2017; Luo and Tang, 2014). Other signs cannot be assigned, due to mixed results (Luo and Tang, 2014; Cooper et al., 2008; Walls et al., 2012; Wang et al., 2019).
For board characteristics two proxies will be used, namely board size (BSI) and board gender diversity (BGD). Board characteristics have been related to environmental disclosure and environmental performance in previous studies with mixed results, (Freedman and Jaggi, 2011; Hassan and Romilly, 2017; Luo and Tang, 2014; Walls et al., 2012) so the sign could either be positive or negative. Variable descriptions can be found in Appendix B.
From my initial sample of the top 700 companies from the MSCI World Index, imported from Thompson Reuters Eikon database, only 356 companies (50.86%) disclosed their GHG emissions. I will use these companies for further research. 225 of these companies are from the USA, followed by 17 from Japan. Country of origin and industry distribution of the dataset can be found in Appendix C. My dataset will consist of data for the period 2013-2019, because of the importance of current knowledge about the relationship between environmental disclosure and performance, and the effect of institutional pressures
3.3 Descriptive analysis
23
Variable Obs. Mean Standard deviation
Median Min Max
GHG (ln) 2030 13.73 2.17 13.64 9.46 18.21 ED (1-100 score) 2435 66.23 21.06 70.74 10.72 98.46 AIO (ratio) 2457 0.561 0.227 0.622 0.010 0.928 PIO (ratio) 2457 0.085 0.054 0.078 0.005 0.248 LEV (ratio) 2483 0.275 0.182 0.250 0 2.44 SIZE (ln) 2486 24.63 1.49 24.46 21.92 28.26 ROA (ratio) 2484 0.056 0.059 0.046 -0.325 0.480 CAPEX (in billions USD) 2400 2.79 4.33 1.16 0.05 22 BSIZE (person) 2435 12.24 3.15 12 4 25 BGD (ratio) 2435 0.227 0.111 0.214 0 0.646 SALESG (ratio) 2259 0.049 0.139 0.037 -0.295 0.501
Table 1. The table shows the descriptive statistics of all key variables including the scale used in the models. All
variables are defined Appendix B. The sample is based on data from 2012 to 2018.
Table 2. Pearson correlation matrix of the variables in equation 1 and 2. Significance level at 95%, shown by *.
GHG ED AIO PIO LEV SIZE ROA CAPEX BS BGD SALESG
24 and 8.5% of total shares outstanding are owned by active and passive institutional investors respectively. The percentage of active institutional investors is quite high compared to the study of Almazan et al. (2005), which had an average of 31%. The BGD mean shows that in a firm’s board, 22.7% are female on average. This low percentage is probably caused by the large amount of firms from the USA. In 2016, board gender diversity was between 15% and 20% in the USA (Harvard University).
For the univariate analysis, the Pearson’s correlation coefficients are given in table 2. In general, all correlation coefficients are well below the critical value of 0.8, so multicollinearity will not be a problem in this model. Only CAPEX and SIZE correlate highly and significantly with a coefficient of 0.595. The table shows that GHG and ED are positively correlated, which means that environmental disclosure is negatively related to environmental performance. The coefficients of GHG-AIO and GHG-PIO are both negative, where the coefficient of the former is twice the size of the coefficient of the latter. This indicates that higher active and passive institutional ownership are associated with higher environmental performance (lower GHG emissions). Coefficients of ED-AIO and ED-PIO are also negative, meaning a lower disclosure score for higher active and passive institutional ownership. Furthermore, correlation signs of GHG and ED with control variables SIZE, CAPEX, BS and SALESG do not show unexpected results. Important to note is that the correlation coefficient of GHG-BGD and GHG-ROA are insignificant on the 95% significance level.
25
4. Results and discussion
4.1 Environmental disclosure, active and passive institutional ownership in relationship with environmental performance
26 Model 1 GHG Model 2 GHG Model 3 GHG Model 1, 2 and 3 variables combined Model 4 ED Model 5 ED Model 4 and 5 variables combined Constant 5.837 (1.676)*** 6.965 (1.704)*** 7.141 (1.667)*** 6.970 (1.702)*** -192.078 (46.514)*** -193.561 (45.199)*** -192.011 (46.496)*** ED -0.0020 (0.0014) -0.0024 (0.0015) -0.0025 (0.0015)* -0.0024 (0.0015) AIO (-1) -0.159 (0.160) -0.169 (0.169) 2.945 (4.111) 2.622 (6.018) PIO (-1) -0.081 (0.368) -0.143 (0.389) -5.309 (10.769) -3.909 (10.908) LEV -0.528 (0.126)*** -0.633 (0.150)*** -0.629 (0.148)*** -0.630 (0.150)*** 5.582 (4.111) 5.691 (4.094) 5.680 (4.125) SIZE 0.350 (0.072)*** 0.310 (0.074)*** 0.300 (0.071)*** 0.311 (0.074)*** 10.220 (1.973)*** 10.363 (1.877)*** 10.237 (1.979)*** ROA -0.063 (0.176) -0.310 (0.183)* -0.336 (0.187)* -0.308 (0.184)* 14.549 (6.849)** 14.925 (6.736)** 14.617 (6.855)** CAPEX 0.0075 (0.006) 0.0084 (0.006) 0.0086 (0.006) 0.0084 (0.006) 0.190 (0.232) 0.186 (0.231) 0.188 (0.232) BS -0.003 (0.008) -0.0050 (0.009) -0.0045 (0.008) -0.0049 (0.009) 0.0036 (0.218) 0.0026 (0.216) 0.004 (0.218) BGD -0.589 (0.162)*** -0.475 (0.173)*** -0.498 (0.176)*** -0.474 (0.174)*** 19.609 (6.006)*** 20.068 (5.847)*** 19.663 (5.998)*** SalesG -0.339 (0.078)*** -0.292 (0.076)*** -0.295 (0.078)*** -0.291 (0.786)*** -7.127 (2.230)*** -6.938 (2.195)*** -7.094 (2.226)*** IND fixed effects
Yes Yes Yes Yes Yes Yes Yes
Year fixed effects
Yes Yes Yes Yes Yes Yes Yes
N of Obs. 1764 1520 1521 1520 1842 1844 1842
F-statistic 6.48*** 4.30*** 4.60*** 4.11*** 7.39*** 6.74*** 6.57***
Adj. R-sq 0.1650 0.1569 0.1549 0.1566 0.1189 0.1192 0.1190
Table 3. This table shows the fixed effects regression results of environmental disclosure (ED), active institutional
ownership (AIO) and passive institutional ownership (PIO) on GHG emissions in model 1, 2 and 3 respectively, and of AIO and PIO on ED in model 4 and 5 respectively. The robust standard errors are in brackets. Significance levels at 99%, 95% and 90%, shown by ***, ** and * respectively.
27 stakeholders. One important group of stakeholders are institutional stakeholders, who can potentially pressure firms to increase environmental engagement.
In model 2, active institutional ownership shows a negative coefficient of 0.159, which is twice the size of the coefficient of passive institutional ownership in model 3. The difference in effect of active and passive institutional ownership originates from the fact that it is easier for active institutional investors to pressure firms, because of the lower monitoring costs and voting power. It is expected that these investors and managers suffer less from the agency problem, because they have better alignment with eachother. However, coefficients of active and passive institutional ownership are insignificant, so the implications do not sustain. In the combined model the coefficient of passive institutional ownership increases with 76%, but still remains insignificant. Regarding passive institutional ownership, H3 is accepted, which indicates that the monitoring power and ability to influence decision-making of institutional shareholders is insufficient to affect environmental performance. Furthermore, H2 is rejected, implying that active institutional shareholders are also unable to push firms into more sustainable practices. This implies that the monitoring power of institutional investors, even for active institutional investors, is insignificant, which is consistent with Brown et al. (2006), but also with Calza et al. (2012), who found no association between short-term/long-term institutional ownership and environmental practices.
28 This means that a higher leverage ratio, woman percentage on board and sales growth percentage are associated with lower GHG emissions. The coefficient of ROA with a negative sign are significant on the 90% level in model 2 and 3, while in model 1 this coefficient is insignificant. CAPEX and board size (BS) have an insignificant coefficient in all models. The adjusted R-squared of the first 3 models is between 0.1549-0.1650, which implies a good explanatory value of the independent variables on the variance of the dependent variable.
4.2 Active and passive institutional ownership on environmental disclosure
Model 4 and 5 show the effect of active and passive institutional ownership on environmental disclosure. Noteworthy is that active institutional ownership has a positive coefficient of 2.945 and passive institutional ownership a negative coefficient of 5.309. But again, both coefficients are insignificant on the 90% significance level, so both types of institutional ownership do not seem to have a relationship with environmental disclosure. This means that H4 is rejected and H5 is accepted. Similar to findings in model 2 and 3, active and passive institutional investors also do not have the capability and monitoring power to affect environmental disclosure. The combined model of 4 and 5 does not present any significant changes. Another implication can be that institutional investors are not interested in environmental disclosure and environmental performance, but instead use their monitoring power and ability to affect decision-making in a firm for other purposes, for example by increasing pay-per-performance sensitivity of managers (Almazan et al., 2005).
29 or percentage of woman on the board are associated with an increase in environmental disclosure, which is also consistent with previous literature (Hassan and Romilly, 2017; Wang et al., 2019). The adjusted R-squared experienced a small decrease to 0.1189-0.1192. One possibility is that the link between institutional investors and environmental engagement differs between countries, because of regulations and levels of ownership and control. The following part investigates the relationship of institutional ownership and environmental performance and environmental disclosure in Anglo-Saxon countries versus non Anglo-Saxon countries
4.3 Anglo-Saxon vs. Non-Anglo-Saxon countries
30 environment disclosure. Implications made for the entire sample can also be used here. Hence, the ineffectiveness of the monitoring power is evident for institutional investors from Anglo-Saxon countries as well as non Anglo-Anglo-Saxon countries.
Model 6A GHG Model 6B GHG Model 7A ED Model 7B ED Wilcoxon rank-sum test Z (probability) Country type
Anglo-Saxon Non Anglo-Saxon Anglo-Saxon Non Anglo-Saxon
Constant 7.207 (2.027)*** 7.196 (2.270)*** -181.749 (53.101)*** -218.033 (76.037)*** GHG -3.190 (0.0014) ED -0.0025 (0.0016) -0.002 (0.003) 7.280 (0.0) AIO (-1) -0.217 (0.210) -0.053 (0.271) 5.628 (8.161) -1.285 (6.835) -31.643 (0.0) PIO (-1) -0.145 (0.483) -0.514 (0.662) -6.481 (12.952) -4.396 (14.673) -22.699 (0.0) LEV -0.584 (0.177)*** -1.095 (0.509)** 6.277 (4.388) 0.467 (9.303) -13.909 (0.0) SIZE 0.300 (0.087)*** 0.306 (0.100)*** 9.460 (2.258)*** 11.964 (3.216)*** 10.806 (0.0) ROA -0.141 (0.180) -1.588 (0.819)* 14.820 (7.658)* 9.872 (13.687) -8.832 (0.0) CAPEX 0.0091 (0.009) 0.0053 (0.005) 0.364 (0.318) -0.283 (0.208) 6.128 (0.0) BS 0.0063 (0.0079) -0.019 (0.015) 0.295 (0.252) -0.376 (0.297) 11.506 (0.0) BGD -0.518 (0.177)*** -0.403 (0.368) 30.295 (7.704)*** -10.476 (6.105)* -5.021 (0.0) SalesG -0.362 (0.095)*** -0.132 (0.085) -7.714 (3.250)** -3.422 (2.025)* -2.690 (0.0071) IND fixed effects
Yes Yes Yes Yes
Year fixed effects
Yes Yes Yes Yes
N 1104 416 1352 490
F-statistic 3.46*** 3.73*** 6.08*** 3.16***
Adj R-sq 0.1885 0.1377 0.1334 0.1364
Table 4. This table shows the fixed effects regression results of environmental disclosure (ED), active institutional
31 Continuing with the control variables for model 6, the coefficient of LEV in model 6A is -0.584, compared to a coefficient of -1.095 in model 6B. The coefficients are both significant on the 99% significance level. This indicates that firms from non Anglo-Saxon countries emit a lower amount of GHG for the same leverage ratio compared to firms from Anglo-Saxon countries. It may be that firms from non Anglo-Saxon countries face more financial constraints for the same leverage ratio and have to improve sustainability practices, such as environmental performance, to reduce these constraints. It can also be said that in non Anglo-Saxon countries, it is more expensive to gain higher environmental performance, which involves higher investment (more debt). The coefficient of ROA shows an insignificant small result for Anglo-Saxon, while this is significant for non Anglo-Saxon countries. The result for non Anglo-Saxon is also significantly bigger (-1.588) than the result of the full sample (-0.308). This implies that firms from non Anglo-Saxon countries increase environmental performance by means of a reduction of GHG emissions, when they have higher firm performance. Both negative coefficients BGD and SalesG in Anglo-Saxon countries are significant and present similar results compared to the full sample, while in non Anglo-Saxon countries no significance for a link can be found.
32 board gender diversity is associated with higher environmental disclosure, where in non countries the opposite is true.
5. Conclusion
The goal of this study was to investigate the relationship between corporate environmental performance and corporate environmental disclosure as well as to investigate the effect of active and passive institutional ownership on these environmental variables. From a model based on panel data of 356 companies from the MSCI global index for the period 2013-2019, I obtained the following findings.
First, I find no evidence for a relationship between environmental disclosure, measured by using ASSET4’s environmental disclosure score, and environmental performance, measured by GHG emissions. A possible explanation might be that firms, and managers in particular, do not have the incentive to engage in GHG reduction, but instead focus on other environmental protection activities to increase their environmental disclosure rating and to satisfy stakeholders (Yu and Lee, 2017). Instead, firm size, board gender diversity, sales growth and the leverage ratio proved to be important determinants for environmental performance and environmental disclosure.
33 does not have an influence on both environmental disclosure and environmental performance. Evidence of this study shows that these institutions are not able to pressure businesses into more sustainable practices, assumably because their monitoring power is not sufficient enough. This means that there remains an agency problem between institutions and managers. Another reason could be that these institutions are not interested in environmental disclosure and performance of the firm, but prefer to use their influence and voting power to focus on other, in their eyes more important, firm activities. Furthermore, I also did not find a relationship of passive institutional ownership with either environmental performance and environmental disclosure. Among passive institutions are banks and trusts, insurance companies and hedge funds. This confirms that, in general, institutional investors do not play a crucial role in influencing companies’ environmental disclosure and performance.
Finally, investigation of firms from Anglo-Saxon countries and non Anglo-Saxon countries did not lead to differences between the two in terms of the link between environmental disclosure and environmental performance on one side, and a relationship of active and passive institutional ownership with environmental disclosure and environmental performance on the other. One interesting finding in this study is that, in Anglo-Saxon countries, a higher percentage of women on board is related with a higher environmental disclosure score, while the opposite holds true in non Anglo-Saxon countries.
34 decisions. On top of that, indifference of pressure of active and passive institutional investors on environmental engagement helps policy-makers to construct environmental guidelines. Development of mandatory disclosure policies could increase the alignment of environmental disclosure and performance and an increase of alignment between managers and institutional investors, which could decrease agency costs.
With this thesis, I cover all three aspects in the field of international financial management. I used a global sample of firms from 19 different countries including Anglo-Saxon and non Anglo-Saxon countries. Furthermore, institutional investors gain an improved understanding of the relationship between environmental disclosure and performance and the influence of their monitoring power, which helps them in making well-informed investment choices. At last, this thesis contributes to literature focused on environmental management activities and findings help managers understand the effect institutional pressures.
36
Appendix A
Hausman test for equation 1: GHG emissions as dependent variable Coefficients
(b) (B) (b-B) Sqrt(diag(V_b-V_B))
Fixed effects Random effects Difference S.E.
ED 0.021 0.019 0.002 0.001 AIO -5.392 -5.801 0.409 0.533 PIO -4.292 -4.413 0.121 0.328 LEV -1.734 -1.673 -0.061 0.154 SIZE 1.540 1.755 -0.215 0.180 ROA -4.847 -5.635 0.788 - CAPEX 0.172 0.281 -0.109 0.013 BS -0.228 -0.203 -0.025 - BGD -2.802 -2.476 -0.324 - SALESG -1.963 -2.130 0.167 -
b = consistent under Ho and Ha; obtained from xtreg
B = inconsistent under Ha, efficient under Ho; obtained from xtreg Test: Ho: difference in coefficients not systematic
Chi2(10) = (b-B)*[(V_b-V_B)^(-1)](b-B) = 550.62
Prob>chi2 = 0.0000
37 Hausman test for equation 2: ED as dependent variable
Coefficients
(b) (B) (b-B) Sqrt(diag(V_b-V_B))
Fixed effects Random effects Difference S.E.
AIO 2.622 1.385 1.237 2.321 PIO -3.909 -6.077 2.168 3.255 LEV 5.680 4.164 1.516 1.124 SIZE 10.237 6.949 3.288 0.709 ROA 14.617 14.970 -0.353 1.538 CAPEX 0.188 0.253 -0.065 0.093 BS 0.004 0.134 -0.130 0.052 BGD 19.663 24.867 -5.204 1.348 SALESG -7.094 -6.884 -0.210 0.242
b = consistent under Ho and Ha; obtained from xtreg
B = inconsistent under Ha, efficient under Ho; obtained from xtreg Test: Ho: difference in coefficients not systematic
Chi2(10) = (b-B)*[(V_b-V_B)^(-1)](b-B) = 22.08
Prob>chi2 = 0.0086
38
Appendix B
Variable descriptions. SIZE and CAPEX are in USD.
Variables Descriptions Code (EIKON)
GHG GHG emissions scope 1 and scope 2, measured by emissions in millions of tonnes. EIKON has data of CO2 and six largest other equivalent emissions combined in the datasets.
TR.CO2DirectScope1 TR.CO2IndirectScope2
ED The Environmental strategy disclosure score (0-100), measured by self-reported information on 51 environmental disclosure indicators
TR.EnvironmentPillarScore
AIO Active institutional ownership, measured by total active institutional investor shares to total shares outstanding in percentage.
Category investor percent of shares outstanding – TR.CategoryOwnershipPct
PIO Passive institutional ownership, measured by total passive institutional investor shares to total shares outstanding in percentage
Category investor percent of shares outstanding – TR.CategoryOwnershipPct
LEV Leverage ratio, measured by total debt to total assets TR.TotalDebtOutstanding TR.TotalAssetsReported SIZE Firm size, measured by taking the natural logarithm
of total assets
TR.TotalAssetsReported
ROA Return on assets, measured as net income after taxes to total assets
TR.NetIncomeAfterTaxes TR.TotalAssetsReported CAPEX Capital expenditure, measured as the sum of
purchase of fixed assets, purchase/acquisition of intangibles and software development costs. (in billions of USD)
TR.CapitalExpendituresCFStmt
BS Board size, measured as the total number of full-time directors on board
TR.CGBoardSize
BGD Board gender diversity, measured as the ratio of females on board
TR.AnalyticBoardFemale
SALESG Sales growth, measured by sales t to sales t-1. TR.TotalRevenue
IND Industry dummy variables. TR.GICSIndustry
39
Appendix C
Country of origin and industry sector for 356 firms used in this study.
Country N Industry sector N
Argentina 1 Energy 23
Belgium 1 Basic materials 19
Bermuda 1 Industrials 55
Brazil 8 Cyclical consumer goods and services 19
Canada 12 Non-cyclical consumer goods and services 74
China 11 Financials 30
France 17 Healthcare 56
Germany 17 Technology 12
Hong Kong 1 Telecommunications services 23
India 1 Utilities 41 Ireland 7 Unknown 4 Japan 18 Netherlands 8 Russia 1 South Korea 1 Spain 2 Switzerland 16 United Kingdom 14
United States of America 218
40
Acknowledgements
41
References
Almazan, A., Hartzell, J., Starks, L., 2005. Active Institutional Shareholders and Costs of Monitoring: Evidence from Executive Compensation. Financial Management 34(4), 5-34.
Aouadi, A., Marsat, S., 2017.Do ESG Controversies Matter for Firm Value? Evidence from International Data. Journal of Business Ethics 151(4), 1027-1047.
Baraibar, E., Odriozola, M., 2019. CSR committees and their effect on ESG performance in UK, France, Germany, and Spain. Sustainability 11(18), 1-20.
Bennet, J., Sias, R., Starks, L., 2003. Greener Pastures and the Impact of Dynamic Institutional Preferences. Review of Financial Studies 16(4), 1203-1238.
Bernardi, C., Stark, A., 2016. Environmental, social and governance disclosure, integrated reporting, and the accuracy of analyst forecasts. The British Accounting Review 50, 16-31.
Brooks, C., 2008. Introductory Econometrics for Finance. Cambridge University Press, Cambridge.
Brown, W., Helland, E., Smith, J., 2006. Corporate philanthropic practices. Journal of Corporate Finance 12(5), 855-877.
Brundtland, G., 1987. Our Common Future. Oxford University Press, Oxford
Cadez, S., Czerny, A., Letmathe, P., 2018. Stakeholder pressures and corporate climate change mitigation strategies. Business Strategy and the Environment 28, 1-14.
Calza, F., Profumo, G., Tutore, I., 2016. Corporate Ownership and Environmental Proactivity. Business Strategy and the Environment 25(6), 369-389.
Chen, X., Harford, J., Li, K., 2007. Monitoring: Which institutions matter? Journal of Financial Economics 86(2), 279-305.
Cheng, B., Ioannou, I., Serafeim, G., 2014. Corporate social responsibility and access to finance. Strategic Management Journal 35, 1-23.
Cho, C., & Patten, D., 2007. The role of environmental disclosures as tools of legitimacy: A research note. Accounting, Organizations and Society 32(7-8), 639–647.
Choi, J., Jo, H., Kim, J., Kim, M., 2018. Business groups and corporate social responsibility. Journal of Business Ethics 153(4), 931-954.
Clarkson, P., Li, Y., Richardson, G., Vasvari, F., 2008. Revisiting the relation between environmental performance and environmental disclosure: An empirical analysis. Accounting, Organizations and Society 33, 303-327.
Cooper, S., Raman, K., Yin, J., 2018. Halo effect or fallen angel effect? Firm value consequences of greenhouse gas emissions and reputation for corporate social responsibility? Journal of Accounting and Public Policy 37, 226-240. Córdova, C., Zorio-Grima, A., Merello, P., 2018. Carbon Emissions by South American Companies: Driving Factors for Reporting Decisions and Emissions Reduction. Sustainability 10(7), 2411-2427.
Cotter, J., Najah, M., 2012. Institutional investor influence on global climate change disclosure practices. Australian Journal of Management 37, 169-187.
42 Faisal, F., Andiningtyas, E., Achmad, T., Haryanto, H., Meiranto, W., 2018. The content and determinants of
greenhouse gas emission disclosure: Evidence from Indonesian companies. Corporate Social Responsibility and Environmental Management 25(6), 1397-1406.
Fekrat, M., Inclan, C., Petroni, D., 1996. Corporate environmental disclosures: Competitive disclosure hypothesis using 1991 annual report data. The International Journal of Accounting 31(2), 175-195.
Ferreira, M., Matos, P., 2008. The colors of investor’s money: the role of institutional investors around the world. Journal of Financial Economics 88(3), 499-533.
Freedman, M., Jaggi, B., 2005. Global warming, commitment to the Kyoto protocol, and accounting disclosures by the largest global public firms from polluting industries. International Financial Management & Accounting 40(3), 215-232.
Freedman, M., Jaggi, B., 2010. Global warming disclosures: Impact of Kyoto Protocol across countries. Journal of International Financial Management & Accounting 22, 46-90.
Giannarakis, G., Zafeiriou, E., Sariannidis, N., 2017. The Impact of Carbon Performance on Climate Change Disclosure. Business strategy and the Environment 26, 1078-1094.
Graafland, G., Smid, H., 2014. Does Corporate Social Responsibility Really Make a Difference? An Explorative Analysis for Chinese Companies. China & World Economy 22(2), 102-124.
Graves, S., Waddock, S., 1994. Institutional owners and corporate social performance. Academy of Management Journal 37(4), 1034-1046.
Hassan, O., Romilly, P., 2017. Relations between corporate economic performance, environmental disclosure and greenhouse gas emissions: New insights. Business Strategy and the Environment 27, 893-909.
Herold, D., Farr-Wharton, B., Lee, K., Groschopf, W., 2018. The interaction between institutional and stakeholder pressures: Advancing a framework for categorising carbon disclosure strategies. Business Strategy and
Development 2, 77-90.
Ho, P., Tower, G., 2011. Ownership structure and voluntary disclosure in corporate annual reports of Malaysian listed firms. Corporate Ownership and Control 8, 296-312.
Jackson, G., Apostolakou, A., 2010. Corporate social responsibility in Western Europe: An institutional mirror or substitute? Journal of Business Ethics 94, 371-394.
Keeling, R., Piper, S., Bollenbacher, A., Walker, S. Atmospheric CO2 values (ppmv) derived from situ air samples collected at Mauna Loa, Hawaii, USA. Scripps Institution of Oceanography, University of California, La Jolla, California 92093-0444, USA. Retrieved on 02-2020.
Konadu, R., 2017. Gender Diversity Impact on Corporate Social Responsibility (CSR) and Greenhouse Gas Emissions in the UK. Economics and Business Review 3(17), 127-148.
Lo, K., 2014. China's low-carbon city initiatives: The implementation gap and the limits of the target responsibility system. Habitat International 42, 236–244.
Lu, J., Herremans, I., 2019. Board gender diversity and environmental performance: an industries perspective. Business Strategy Environment 28, 1449-1464.
43 Manning, B., Braam, G., Reimsbach, D., 2019. Corporate governance and sustainable business conduct— Effects of board monitoring effectiveness and stakeholder engagement on corporate sustainability performance and disclosure choices. Corporate Social Responsibility and Environmental Management 26(2), 351-366.
Matisoff, D., 2013. Different rays of sunlight: understanding information disclosure and carbon transparency. Energy Policy 55, 579–592.
Neubaum, D., Zahra, S., 2006. Institutional owner ship and corporate social performance: the moderating effects of investment horizon, activism, and coordination. Journal of Management 32, 108-131.
Niedertscheider, M., Haas, W., & Görg, C., 2018. Austrian climate policies and GHGemissions since 1990: What is the role of climate policy integration? Environmental Science and Policy 81, 10-17.
Nollet, J., Filis, G., Mitrokostas, E., 2016. Corporate social responsibility and financial performance: A non‐linear and disaggregated approach. Economic Modelling 52, 400–407.
Shleifer, A., Vishny, R., 2012. A survey of Corporate Governance. The Journal of Finance 52(2), 737-783.
Van Aarle, B., Tielens, J., Van Hove, J., 2015. The financial crisis and its aftermath: the case of Ireland. International Economics and Economic Policy 12, 393-410.
Yu, J., Lee, S., 2017. The Impact of Greenhouse Gas Emissions on Corporate Social Responsibility in Korea. Sustainability 9(7), 1135-1150.
Walls, J., Berrone, P., Phan, P., 2012. Corporate governance and environmental performance: is there really a link? Strategic Management Journal 33, 885–913.
Wang, F., Yang, J., Liu, S., 2019. Institutional pressure, ultimate ownership, and corporate carbon reduction engagement: Evidence from China. Journal of Business Research 104, 14-26.
