Professional Documents
Culture Documents
Abstract
doi: 10.1111/acfi.12560
The authors acknowledge the financial support of AFAANZ research grant, 2017
(No. G1701094) of $5000.
1. Introduction
1
Firms have the alternative of disclosing in their annual reports why the stipulated 2
percent has not been spent on the prescribed schedule VII CSR activities.
these contracts. Through CSR activities, firms intend to create and maintain a
standing for their obligations related to implicit contracts.
Supporting this theoretical concept, there is a stream of literature demon-
strating how CSR increases firms’ performance and stakeholder value. The
studies of Servaes and Tamayo (2013) and Fombrun (2005) in general find a
positive relation between CSR and firms’ stakeholder value. Barnett (2007) also
advocates that CSR spending has an effect on firm value that is determined by
the ability of CSR to influence the stakeholders of the firms. Further, several
studies document an explicit linkage between CSR expenditure and financial
performance. For example, higher levels of CSR initiatives and spending on
strategic CSR activities help firms to (i) attract and retain high quality skilled
employees (Greening and Turban, 2000), (ii) improve product and service
promotion efficiency (Fombrun, 2005), providing them with greater rights to
use scarce resources (Cochran and Wood, 1984), (iii) increase demand for
goods and services (Navarro, 1998) and (iv) create goodwill that can restore
performance in the event of a violation of a legal requirement by reducing
penalties (Godfrey, 2005). In contrast, a negative relation between increased
CSR undertakings or ratings with a firm’s return on assets (ROA) and stock
return is reported by Di Giuli and Kostovetsky (2014).
In addition, the impact of CSR spending may extend beyond firms’ financial
performance measures (Lys et al., 2015). For example, higher institutional
investor ownership, broader analyst coverage and reductions in firms’ cost of
capital may be achieved by voluntary CSR activities (Dhaliwal et al., 2011).
Firms with higher consumer responsiveness document a positive association
between spending on CSR activities and their financial performance although
firms with lower consumer responsiveness show no association (Servaes and
Tamayo, 2013). A firm’s legitimacy and reputation may be maintained through
CSR expenditure (Peloza, 2006) and possible legislative action or negative
regulatory views may be moderated (Hillman and Keim, 2001).
In contrast, Lys et al. (2015) assert that CSR activity may be undertaken by a
firm because it anticipates sustainable positive financial performance, rather
than the CSR activity itself causing this positive performance. The methods
used in prior studies do not allow for conclusions to be drawn about the direct
effect of CSR expenditure on firms’ current financial performance. A further
limitation of this literature is its mixed findings. Reviewing the literature on the
association between CSR activities and firms’ financial performance, Margolis
et al. (2009) note that a few studies report significant positive relations between
CSR activities and firm performance, while others report negative associations.
The context of this study is mandatory CSR expenditure in India, which may
have a detrimental effect on shareholders’ wealth rather than a beneficial one,
for a number of reasons. Consistent with Friedman (1970), if firms choose not
to spend on CSR, then mandating CSR activities and spending will reduce their
value. Furthermore, by directing CSR spending in a particular way, business
choices over the type of CSR activity they undertake are restricted. Karnani
(2003) argues that section 135 basically imposes an additional 2 percent tax on
companies, which may make India less attractive to foreign investors and large
multinational companies. If they take the option of disclosing why the
stipulated 2 percent has not been spent as directed, firms are likely to
implement a number of costly face-saving strategies in order to avoid any
potential adverse legal consequences. Associated compliance costs also are an
integral part of a mandated CSR policy.
Although mandated CSR expenditure regulations may impose certain costs on
shareholders, there may also be some benefits. A firm’s CSR policies become
more formalised and visible under a mandatory CSR expenditure rule, providing
a stronger signal to shareholders. Strategic CSR reporting and spending involves
more transparent CSR and financial disclosures, which have the potential to
improve a company’s reputation both locally and nationally. The literature
reports enhanced firm values resulting from the 1999 Korean governance reforms
(Black and Kim, 2012), significant increases in CSR performance after the
introduction of section 299(1)(f) of the Australian Corporations Act 2001 (Frost,
2007), and a positive stock market reaction to Indian governance reforms
(Clause 49, adopted in early 2000) (Black and Khanna, 2007).
In summary, the majority of the literature reports that CSR can have a
positive effect on firm performance. Thus we hypothesise for banks:
H1: In a mandatory setting, CSR expenditure is positively associated with financial
and market performance.
In developing countries, only a small proportion of the population can afford
banking products and services, thus excluding the majority who do not have
access to them. However, the unbanked population has the same entitlement to
banking services as those who can afford them, and banking services and
products could have a significant impact on their lives. Using the argument of
Brown and Deegan (1998) that there is an implicit contract between the firm
and those who are affected by its operations, we recognise that there is a social
contract between banks and the unbanked population. Hawkins (2006) and
Jamali (2008) find that the success of catering for the needs of primary
stakeholders depends on whether firms are fulfilling the needs of other
stakeholders. Following stakeholder theory, Freeman (2010) and Jo et al.
(2015) point out that a firm can enhance its reputation and performance
positively by attending to the implicit rights of all stakeholders. Financial
inclusion activities widen the financial system’s resource base by engaging with
all of the country’s populace, thus playing a pivotal role in the process of
economic development.
The literature uses well-established country-level supply side measures of
financial inclusion, such as branch density, the number of ATMs, market
penetration of mobile phones, or the share of households or firms with access
to financial services. In addition, international organisations regularly survey
financial inclusion (IMF Financial Access survey, Fidex database).
government controls the resources that are vital for the firms’ sustained
operations and success (Pfeffer and Salancik, 1978). As a firm’s engagement
with financial inclusion activities and mandated CSR both fulfil the govern-
ment’s inclusive growth and poverty reduction goals and satisfy the demands of
all stakeholders, we argue that mandated CSR activity and spending will
moderate the association of financial inclusion with financial performance.
Therefore, we hypothesise that:
H3: Mandated CSR spending moderates the association of financial inclusion with
market performance.
The relations between the three hypotheses proposed in this study are shown
in Figure 1.
3. Method
3.1. Models
CSR
Expenditure
H1
Financial H3 Firm
Inclusion H2
Performance
X
Performancet ¼ ac þ b1b CSRt þ b2b FIt þ li;c Control variablesi;t þ e
ð3Þ
The above regression models are estimated using ordinary least squares
(OLS) as it is the most commonly used approach for examining a causal
inference in the case of a repeated cross-section sample (see, for example, Bose
et al., 2017; Chauvet and Jacolin, 2017; Alhaj-Ismail et al., 2019). The literature
also uses two-stage least squares (2SLS) to examine causal inference between
financial variables, particularly to address reverse causality and endogeneity
problems (Richardson et al., 2016; Ahmed and Ali, 2017; Ahmed et al., 2017).
Moreover, in order to address endogeneity, simultaneity and unobserved
heterogeneity, researchers also use generalised method of moment (GMM)
estimation in the corporate finance/governance literature (Sufian and Habibul-
lah, 2010; Ahmed et al., 2017; Safiullah and Shamsuddin, 2018). Since our
initial estimation shows that CSR expenditure or financial inclusion activities
have an insignificant (negative and significant) association with the accounting
performance (stock return) measure, we consider that our regression modelling
is unlikely to be characterised by reverse causality or endogeneity problems.
Although we do not address them in our baseline analysis,2 as a robustness
check, we estimate the above-mentioned regression models using both 2SLS
and GMM approaches.
Statistical inference with regard to CSR expenditure’s and financial inclusion’s
causal relation with firm performance is derived from both standard t-statistics
and heteroscedasticity-consistent robust t-statistics following the Huber–White
procedure. Since we find qualitatively similar results, only regression estimates
based on standard t-statistics are reported. However, we present our main results
based on Huber–White robust t-statistics in Appendix Table A1. We also
examine potential multicollinearity by calculating the variance inflation factor
(VIF) of each independent variable. Both the centred and uncentred VIFs are
below 5, indicating the absence of a high degree of multicollinearity (Wooldridge,
2015). In estimating the regression models, we include year-fixed effects to
remove the potential impact of any time trend.
Our entire analysis concentrates on the relation of banking firm performance
with CSR expenditure and financial inclusion after the introduction of the
mandatory CSR legislation. We are unable to examine the causal relation in
both the pre- and post-legislation periods because actual CSR expenditure data
before the introduction of the mandatory CSR legislation is mostly unavailable
for banking firms in India. Additionally, because the number of banking firms
listed on the Indian stock exchanges is not very large, segregating them into
treatment and control groups does not provide a statistically valid sample to
conduct a difference-in-difference analysis.
2
This issue is further discussed in Section 5.2.3.
3.2. Variables
Two measures of firm performance, return on assets (ROA) and stock return
(SR) are used. The former is an accounting-based measure that captures
historical financial performance whereas the latter is a market-based measure
that represents the market perception of future performance. ROA is the most
widely used indicator of firm performance with regard to accounting profit.
Stock returns are expected to capture all possible ways that the level of CSR
expenditure and financial inclusion can influence firm performance. According
to the semi-strong interpretation of the efficient markets hypothesis, stock
prices reflect all publicly available information, thereby providing a more
comprehensive picture of firm performance relative to the accounting-based
measures. These measures are also used in the literature. For example, Lys et al.
(2015) use ROA, cash flow from operations (CFO) and SR while Cheng et al.
(2015) use ROA and SR as indicators of firm performance.
To measure financial inclusion at the firm level, we consider five financial
inclusion indicators. They are the number of bank branches, the number of
ATMs, the number of accounts, the amount of bank credit and the amount of
bank deposits. The choice of these financial inclusion variables is consistent
with the previous literature (see, for example, Beck et al., 2007; Sarma and Pais,
2011).
To mitigate the omitted variable problem, we include several control
variables in the regression models. We include control variables that capture
board independence because the Companies Act 2013 includes several key
provisions on corporate governance. However, prior studies by Lys et al.
(2015), Chauvet and Jacolin (2017), Manchiraju and Rajgopal (2017) and
Mukherjee et al. (2018) used only firm-related control variables, ignoring board
independence. We control for firm-related effects for size, risk and profit
margin. Larger firms may have additional resources for CSR spending and,
therefore, may be subject to more pressure to engage in CSR-related activities.
In general, firms with lower risk are more likely to spend on CSR activities
(Orlitzky and Benjamin, 2001).
Variable definitions for measures used in this study are provided in Table 1.
4.1. Sample
Banking firms and their industry groupings are chosen from Thompson
Reuters DataStream. The main source of data used in this paper is Prowess,
which is the largest database of financial performance of Indian business
entities. This database is maintained by the Centre for Monitoring Indian
Economy (CMIE), an independent Indian think-tank and business
Table 1
Variable description
Variable Description
ROA Profit before tax scaled by total assets at the end of financial year t
SR Annual stock return
SIZE Natural logarithm of total assets at the end of a given financial year t
LEVERAGE Total liabilities scaled by total assets at the end of a financial year t. Total
liability is calculated as the sum of current liabilities and non-current
liabilities
ATO Interest income scaled by total assets at the end of a given financial year t
PM Profit before tax scaled by interest income for the financial year t
DIRECTOR Number of independent directors as a proportion of total directors
3
More information about this database can be obtained from https://www.cmie.com/
kommon/bin/sr.php?kall=wapps&tabno=7030&page=about_us
To address the concern that a three-year sample period may not be enough to
capture the impact of CSR expenditure and financial inclusion on firm
performance, we conduct a robustness check. In addition to estimating the
models (Eqns 1–3) for the whole sample period (2015–2017) that generates our
main results, we also estimate the models for yearly sub-periods. We find that
the cross-sectional relation between mandatory CSR expenditure and financial
inclusion with firm performance is remarkably similar across the years and with
the whole sample period. Additionally, we observe that CSR expenditure as a
proportion of income, financial inclusion variables and our measures of firm
performance do not vary significantly across the years (t-statistics for mean
difference of the variables across the years are statistically insignificant). The
stability of our results through our sample period implies that our baseline
results are not sensitive to the length of sample period.4
Our initial sample includes 61 banking firms publicly listed on the Indian
stock exchanges across the sample period. However, we exclude those banks for
which none of the required variables are available. The number of excluded
banks due to data omission is not the same across the years. Therefore,
although our initial sample remains the same (61 banking firms) over the
sample period, in the final sample, the number of sample firms differs in
different years. More specifically, the final sample is 149 banking firm-years and
comprises 49, 52 and 48 banks, respectively, in the years 2015, 2016 and 2017.5
4
The results pertaining to these analyses are available from the corresponding author on
request.
5
However, we have also undertaken a robustness test of the results on a uniform sample
(44 firms over the period 2015–2017, total 132 firm-year observations). A few
observations were missing from our workable sample that we have imputed using
Markov Chain Monte Carlo assuming a multivariate normal distribution. This data
imputation process is particularly suitable in the case of a non-monotonous data pattern
which matches our case.
6
Although we take the natural logarithm of the actual amount of the CSR expenditure
variable (CSR), FI variables (BRANCH, ATM, ACCOUNT, CREDIT and DEPOSIT)
and one of the control variables (SIZE) for our empirical analysis, descriptive statistics
(reported in Table 2) are based on actual data before taking their natural log.
Table 2
Descriptive statistics
ROA, return on assets; SR, stock return; CSR, actual amount spent on CSR activities during
the year; BRANCH, number of bank branches; ATM, number of ATMs; ACCOUNTS,
number of accounts; CREDIT, amount of credit provided by each bank; DEPOSIT, amount
of deposit received by each bank; SIZE, firm size; LEVERAGE, level of debt in capital
structure; ATO, asset turnover; PM, profit margin; DIRECTOR, proportion of independent
directors. All variables are explained in Table 1.
is INR192 million or $US2.86 million (Panel B). However, this may be affected
by outliers as we can see that the 25th and 75th percentiles of CSR expenditure
are, respectively, INR5 million ($US0.075 million) and INR133 million
($US1.98 million). From Panel C, it is observed that in general, the banks
have large networks of branches and ATMs. The average number of branches
and ATMs is 2,258 and 3,938, respectively. This large network of branches has
enabled the banks to access a large customer base as indicated by the number of
accounts (an average of 5,332,143), and the amount of credit provided (an
average of 1,490,156) and deposits received (an average of 1,932,349). With
regard to the control variables (Panel D), it can be seen that banks are more
reliant on equity than debt, as on average only 6.43 percent of total capital
comes from debt. Average ATO and PM are about 8 percent each. The
governance variable (DIRECTOR) implies that on average about 30 percent of
the directors on the banks’ boards are independent.
Table 3
Correlation matrix
ROA SR CSR BRANCH ATM CREDIT DEPOSIT ACCOUNTS SIZE LEVERAGE ATO PM DIRECTOR
Entries are the Pearson’s bivariate correlation between variables. ROA, return on assets; SR, stock return; CSR, actual amount spent on CSR
A. Bhattacharyya et al./Accounting & Finance
activities during the year; BRANCH, number of bank branches; ATM, number of ATMs; ACCOUNTS, number of accounts; CREDIT, amount
of credit provided by each bank; DEPOSIT, amount of deposit received by each bank; SIZE, firm size; LEVERAGE, level of debt in capital
structure; ATO, asset turnover; PM, profit margin; DIRECTOR, proportion of independent directors. All variables are explained in Table 1. ***,
** and * indicate statistical significance at the 1, 5 and 10 percent level, respectively.
5. Empirical results
7
The term r represents bivariate correlation.
Table 4
CSR expenditure and firm performance
(1) (2)
Variable ROA SR
C 0.01*** 1.28
(6.94) (1.38)
CSR 0.00 0.38***
(0.23) (6.64)
SIZE 0.00 0.27*
(1.53) (1.71)
LEVERAGE 0.02*** 1.48
(8.89) (0.99)
ATO 0.13*** 10.78
(7.74) (1.03)
PM 0.09*** 0.66
(43.56) (0.51)
DIRECTOR 0.00 0.83
(0.31) (1.15)
Adjusted R2 0.97 0.33
Year fixed effect Included Included
Dependent variable: Firm performance measured by ROA and SR. Entries are the regression
coefficients and t-statistics (standard) within parentheses derived from estimating Equation
(1) using the OLS method. The sample includes 149 banking firm-years over the period 2015–
2017. CSR, actual amount spent on CSR activities during the year; ROA, return on assets;
SR, stock return; SIZE, firm size; LEVERAGE, level of debt in capital structure; ATO, asset
turnover; PM, profit margin; DIRECTOR, proportion of independent directors. All variables
are explained in Table 1. ***, ** and * indicate statistical significance at the 1, 5 and 10
percent level, respectively.
8
A similar approach is used by Baker and Wurgler (2006). While the principal
components of a list of variables are derived from estimating the eigenvalue
decomposition of observed variance, the first principal component is the original
variables’ unit-length linear combination with maximum variance (for more details, see
Johnson and Wichern, 2013).
Table 5
Financial inclusion and firm performance
(1) (2)
Variable ROA SR
C 0.01*** 0.55
(7.44) (0.62)
FI 0.00 0.56***
(0.33) (5.56)
SIZE 0.00 0.48***
(1.00) (2.33)
LEVERAGE 0.02*** 0.00
(9.00) (0.00)
ATO 0.13*** 18.81*
(7.91) (1.77)
PM 0.09*** 0.39
(44.46) (0.30)
DIRECTOR 0.00 0.75
(0.32) (1.00)
Adjusted R2 0.97 0.28
Year fixed effect Included Included
Dependent variable: Firm performance measured by ROA and SR. Entries are the regression
coefficients and t-statistics (standard) within parentheses derived from estimating Equation
(2) using the OLS method. The sample includes 149 banking firm-years over the period 2015–
2017. ROA, return on assets; SR, stock return; FI, financial inclusion; SIZE, firm size;
LEVERAGE, level of debt in capital structure; ATO, asset turnover; PM, profit margin;
DIRECTOR, proportion of independent directors. All variables are explained in Table 1. ***,
** and * indicate statistical significance at the 1, 5 and 10 percent level, respectively.
9
The results pertaining to these analyses are available from the corresponding author on
request.
Table 6
CSR expenditure, financial inclusion and firm performance
(1) (2)
Variable ROA SR
C 0.01*** 1.75
(5.70) (1.59)
CSR 0.00 0.49***
(0.18) (3.38)
FI 0.00 0.20
(0.30) (0.81)
SIZE 0.00 0.53
(0.40) (1.47)
LEVERAGE 0.02*** 1.88
(8.33) (1.19)
ATO 0.13*** 9.29
(7.55) (0.87)
PM 0.09*** 0.88
(42.54) (0.66)
DIRECTOR 0.00 0.83
(0.31) (1.14)
Adjusted R2 0.97 0.33
Year fixed effect Included Included
Dependent variable: Firm performance measured by ROA and SR. Entries are the regression
coefficients and t-statistics (standard) within parentheses derived from estimating Equation
(3) using the OLS method. The sample includes 149 banking firm-years over the period 2015–
2017. ROA, return on assets; SR, stock return; CSR, actual amount spent on CSR activities
during the year; FI, financial inclusion; SIZE, firm size; LEVERAGE, level of debt in capital
structure; ATO, asset turnover; PM, profit margin; DIRECTOR, proportion of independent
directors. All variables are explained in Table 1. ***, ** and * indicate statistical significance
at the 1, 5 and 10 percent level, respectively.
legislation rather than being voluntarily chosen by the firms, is not associated with
better firm performance, and that FI inclusion, being more a strategy for societal
benefit than corporate profit, also is not associated with better firm performance.
Column 2 shows that CSR expenditure has a statistically significant negative
association with SR, the market-based measure of firm performance. This
result is consistent with our previous findings presented in Table 4, and
supports the interpretation that investors perceive mandatory CSR to be an
additional and unwelcome cost imposed by legislation. In contrast to our
previous findings of a statistically significant negative relation between SR and
FI presented in Table 5, we observe that SR is not associated with FI. This
result indicates that mandated CSR expenditure has a moderating effect on the
relation between SR and FI.10
As indicated earlier, relevant results in the previous literature may be subject to
model misspecification bias. This bias may arise from several sources, such as (i)
the true functional form of the relation between economic and financial variables
being generally unknown, (ii) a single econometric model being unable to
incorporate all the relevant explanatory variables, and (iii) variables themselves
being subject to measurement error. We address the model misspecification bias in
several ways. First, as indicated already, the causal inference of CSR expenditure
and financial inclusion with firm performance is measured by running three
separate regression models (Eqns 1, 2 and 3) with different model specifications.
Second, in estimating each regression model (Eqns 1, 2 and 3), we check the
robustness of results derived from different model specifications such as (i)
including only point-of-interest variable(s) (for example, CSR, FI or both), (ii)
including both point-of-interest and control variables, and (iii) incorporating/
ignoring year-fixed effects.11 As we find consistent results across the models, it may
indicate the absence of model misspecification bias. Third, we use a set of control
variables supported by economic arguments and the previous literature. Fourth,
we use different definitions of the explanatory variables. For example, financial
inclusion is represented by five individual measures and a composite measure of
them. Moreover, although SIZE and LEVERAGE are, respectively, defined as
natural logarithm of total assets and total liabilities scaled by total assets, we also
measure them as the natural logarithm of market capitalisation and non-current
liabilities scaled by non-current assets.12 Finally, to check the general functional
10
While our baseline results are derived from a sample of 149 banking firms (49, 52 and
48 banks respectively in the years 2015, 2016 and 2017), we also estimate our baseline
regression models using a common sample of 44 banking firms each year. We find
remarkably similar results for these two samples. The regression results based on
common samples are presented in Appendix Table A2.
11
Results relating to all these model specifications are presented in Appendix Table A3.
12
Results pertaining to these alternative measures of SIZE and LEVERAGE are not
collated in this paper. However, they can be obtained from the corresponding author on
request.
13
The RESET test, a general specification test, examines whether the response variables
can be explained by non-linear combinations of the fitted values. This test is particularly
applicable for the linear regression model and based on the notion that a model is mis-
specified when non-linear combinations of explanatory variables exhibit explanatory
power with regard to the response variable. Under this circumstance, a polynomial or
other non-linear functional form may better explain the data-generating process.
Table 7
Financial inclusion and firm performance
(continued)
Table 7 (continued)
Dependent variable: Firm performance measured by ROA and SR. Entries are the regression
coefficients and t-statistics (standard) within parentheses derived from estimating Equation
(2) using the OLS method. The sample includes 149 banking firm-years over the period 2015–
2017. Models 1, 2, 3, 4 and 5, respectively, include BRANCH, ATM, CREDIT, DEPOSIT
and ACCOUNTS as financial inclusion measures. ROA, return on assets; SR, stock return;
BRANCH, number of bank branches, ATM, number of ATMs; CREDIT, amount of credit
provided by each bank; DEPOSIT, amount of deposit received by each bank; ACCOUNTS,
number of accounts; SIZE, firm size; LEVERAGE, level of debt in capital structure; ATO,
asset turnover; PM, profit margin; DIRECTOR, proportion of independent directors. All
variables are explained in Table 1. ***, ** and * indicate statistical significance at the 1, 5 and
10 percent level, respectively.
Table 8
CSR expenditure, financial inclusion and firm performance
(continued)
Table 8 (continued)
Control variables
SIZE 0.41 0.23 0.18 0.67* 0.46**
(1.25) (0.90) (0.27) (1.84) (2.10)
LEVERAGE 1.73 1.43 1.38 2.05 1.82
(1.09) (0.93) (0.83) (1.31) (1.20)
ATO 10.43 10.94 10.86 8.48 7.82
(0.99) (1.04) (1.03) (0.80) (0.73)
PM 0.80 0.64 0.61 0.94 0.93
(0.61) (0.49) (0.46) (0.72) (0.71)
DIRECTOR 0.89 0.84 0.84 0.79 0.80
(1.21) (1.15) (1.15) (1.09) (1.10)
Adjusted R2 0.32 0.32 0.32 0.33 0.33
Year fixed effect Included Included Included Included Included
Dependent variable: Firm performance measured by ROA and SR. Entries are the regression
coefficients and t-statistics (standard) within parentheses derived from estimating Equation
(3) using the OLS method. The sample includes 149 banking firm-years over the period 2015–
2017. Models 1, 2, 3, 4 and 5, respectively, include BRANCH, ATM, CREDIT, DEPOSIT
and ACCOUNTS as financial inclusion measures. CSR, actual amount spent on CSR
activities during the year; ROA, return on assets; SR, stock return; BRANCH, number of
bank branches, ATM, number of ATMs; CREDIT, amount of credit provided by each bank;
DEPOSIT, amount of deposit received by each bank; ACCOUNTS, number of accounts;
SIZE, firm size; LEVERAGE, level of debt in capital structure; ATO, asset turnover; PM,
profit margin; DIRECTOR, proportion of independent directors. All variables are explained
in Table 1. ***, ** and * indicate statistical significance at the 1, 5 and 10 percent level,
respectively.
Relevant results are reported in Table 9. Panels A and B show the results
from a model with dependent variables of ROA and SR, respectively. We find
that although lagged CSR expenditure does not have a statistically significant
association with firm performance when performance is measured as ROA
(Panel A), it has a significant negative association with market-based firm
performance (SR) (Panel B). This result implies that mandatory CSR
expenditure is not associated with firms’ one-year ahead performance, but is
associated with lower subsequent performance (when stock return is a measure
of firm performance). This is consistent with the results from the contempo-
raneous regressions. Overall, this result supports the idea that financial
inclusion programs are driven more by social and political motives than
economic ones. As such, financial inclusion activities are negatively perceived
by shareholders. Nonetheless, we find evidence of a moderating effect of CSR
expenditure and financial inclusion on firm performance. Neither CSR
expenditure nor financial inclusion has a statistically significant association
with future firm performance when they are combined in a model (column 3 in
Panels A and B).
Table 9
CSR expenditure, financial inclusion and firm performance
Dependent variable: Firm performance measured by ROAt+1 and SRt+1. Entries are the
regression coefficients and t-statistics (standard) within parentheses. Models 1, 2 and 3 are,
respectively, estimated based on Equations (4), (5), and (6) using the OLS method. The
sample includes 149 banking firm-years over the period 2015–2017. CSR, actual amount
spent on CSR activities during the year; FI, financial inclusion; ROA, return on assets; SR,
stock return; SIZE, firm size; LEVERAGE, level of debt in capital structure; ATO, asset
turnover; PM, profit margin; DIRECTOR, proportion of independent directors. All variables
are explained in Table 1. ***, ** and * indicate statistical significance at the 1, 5 and 10
percent level, respectively.
Table 10
CSR expenditure, financial inclusion and firm performance
2SLS GMM
C 0.01*** 0.01**
(5.16) (3.26)
CSR 0.00 0.00
(0.48) (0.43)
FI 0.00 0.00
(0.11) (0.07)
SIZE 0.00 0.00
(0.38) (0.23)
LEVERAGE 0.02*** 0.02**
(6.62) (2.82)
ATO 0.13*** 0.13***
(5.14) (3.21)
PM 0.09*** 0.09***
(39.32) (18.62)
DIRECTOR 0.00 0.00
(0.64) (0.55)
Adjusted R2 0.97 0.97
Year fixed effect Included Included
C 3.50** 3.50**
(2.41) (2.23)
CSR 0.50** 0.50*
(2.54) (2.27)
FI 0.07 0.06
(0.18) (0.20)
SIZE 0.50 0.50
(1.11) (1.24)
LEVERAGE 0.66 0.66
(0.30) (0.37)
ATO 9.90 9.89
(0.57) (0.70)
PM 0.89 0.89
(0.61) (0.81)
DIRECTOR 0.39 0.39
(0.49) (0.46)
Adjusted R2 0.33 0.46
Year fixed effect Included Included
Dependent variable: Firm performance measured by ROA and SR. Entries are the regression
coefficients and t-statistics (standard) within parentheses. The models are alternatively
estimated using 2SLS and GMM approaches. The sample includes 132 banking firm-years
over the period 2015–2017 (44 banks each year). CSR, actual amount spent on CSR activities
during the year; FI, financial inclusion; ROA, return on assets; SR, stock return; SIZE, firm
size; LEVERAGE, level of debt in capital structure; ATO, asset turnover; PM, profit margin;
DIRECTOR, proportion of independent directors. All variables are explained in Table 1. ***,
** and * indicate statistical significance at the 1, 5 and 10 percent level, respectively.
are presented in Table 10. Panels A and B display results derived from the
models with ROA and SR as alternative dependent variables.14
We find that results from both 2SLS and GMM approaches are consistent in
terms of direction and strength of the relation of CSR expenditure and financial
inclusion with firm performance. ROA is found to be invariant to the level of
CSR expenditure and financial inclusion while SR responds negatively to CSR
expenditure. The sign, magnitude and statistical significance of the regression
coefficients are in line with the results derived from the OLS estimation
presented in Section 5.1. Overall, this result indicates that our key findings
remain robust to the use of these alternative model estimation approaches.
6. Conclusion
14
These results are based on a common sample of 44 banking firms each year.
References
Abdmoulah, W., and R. B. Jelili, 2013, Access to finance thresholds and the finance-
growth nexus, Economic Papers 32, 522–534.
Ahmed, A., and S. Ali, 2017, Boardroom gender diversity and stock liquidity: evidence
from Australia, Journal of Contemporary Accounting and Economics 13, 148–165.
Ahmed, A., R. Monem, D. Delaney, and C. Ng, 2017, Gender diversity in corporate
boards and continuous disclosure: evidence from Australia, Journal of Contemporary
Accounting and Economics 13, 89–107.
Alhaj-Ismail, A., S. Adwan, and J. Stittle, 2019, Share-option based compensation
expense, shareholder returns and financial crisis, Journal of Contemporary Accounting
and Economics 15, 20–35.
Ardic, O. P., M. Heimann, and N. Mylenko, 2011, Access to financial services and the
financial inclusion agenda around the world: a cross-country analysis with a new data
set, World Bank Policy Research Working Paper No. 5537. Available at: https://ssrn.c
om/abstract=1747440
Association of Chartered Certified Accountants, 2008, Going Concern? A Sustainability
Agenda for Action (ACCA, Glasgow).
Baker, M., and J. Wurgler, 2006, Investor sentiment and the cross-section of stock
returns, The Journal of Finance 61, 1645–1680.
Barnett, M. L., 2007, Stakeholder influence capacity and the variability of financial returns
to corporate social responsibility, The Academy of Management Review 32, 794–816.
Baron, D. P., 2001, Private politics, corporate social responsibility, and integrated
strategy, Journal of Economics and Management Strategy 10, 7–45.
Beck, T., A. Demirg€ ußc-Kunt, and V. Maksimovic, 2005, Financial and legal constraints
to growth: does firm size matter?, Journal of Finance 60, 137–177.
Beck, T., A. Demirg€ußc-Kunt, and M. S. Martinez Peria, 2007, Reaching out: access to and
use of banking services across countries, Journal of Financial Economics 85, 234–266.
Black, B. S., and V. S. Khanna, 2007, Can corporate governance reforms increase firm
market values? Event study evidence from India, Journal of Empirical Legal Studies 4,
749–796.
Black, B., and W. Kim, 2012, The effect of board structure on firm value: a multiple
identification strategies approach using Korean data, Journal of Financial Economics
104, 203–226.
Bose, S., A. Bhattacharyya, and S. Islam, 2016, Dynamics of firm-level financial
inclusion: empirical evidence from an emerging economy, Journal of Banking and
Finance Law and Practice 27, 47–68.
Bose, S., J. Podder, and K. Biswas, 2017, Philanthropic giving, market-based
performance and institutional ownership: evidence from an emerging economy, The
British Accounting Review 49, 429–444.
Brown, N., and C. Deegan, 1998, The public disclosure of environmental performance
information – a dual test of media agenda setting theory and legitimacy theory,
Accounting and Business Research 29, 21–41.
Burgess, R., and R. Pande, 2005, Do rural banks matter? Evidence from the Indian
social banking experiment, American Economic Review 95, 780–795.
Chauvet, L., and L. Jacolin, 2017, Financial inclusion, bank concentration, and firm
performance, World Development 97, 1–13.
Cheng, S., K. Z. Lin, and W. Wong, 2015, Corporate social responsibility reporting and
firm performance: evidence from China, Journal of Management and Governance 20,
503–523.
Coase, R. H., 1937, The nature of the firm, Economica 4, 386–405.
Cochran, P. L., and R. A. Wood, 1984, Corporate social responsibility and financial
performance, Academy of Management Journal 27, 42–56.
Dhaliwal, D., Z. Li, A. Tsang, and G. Yang, 2011, Voluntary non-financial disclosure
and the cost of equity capital: the case of corporate social responsibility reporting, The
Accounting Review 87, 723–759.
Di Giuli, A., and L. Kostovetsky, 2014, Are red or blue companies more likely to go
green? Politics and corporate social responsibility, Journal of Financial Economics 111,
158–180.
Dupas, P., and J. Robinson, 2009, Savings constraints and microenterprise development:
evidence from a field experiment in Kenya, NBER Working Paper No. 14693 (National
Bureau of Economic Research).
Fombrun, C. J., 2005, Building corporate reputation through CSR initiatives: evolving
standards, Corporate Reputation Review 8, 7–12.
Margolis, J. D., H. A. Elfenbein, and J. P. Walsh, 2009, Does it pay to be good . . . and
does it matter? A meta-analysis of the relationship between corporate social and
financial performance. Available at: http://ssrn.com/abstract=1866371
Masulis, R. W., and S. W. Reza, 2014, Agency problems of corporate philanthropy, The
Review of Financial Studies 28, 592–636.
McWilliams, A., and D. S. Siegel, 2001, Corporate social responsibility: a theory of the
firm perspective, Academy of Management Review 26, 117–127.
Mukherjee, A., R. Bird, and G. Duppati, 2018, Mandatory Corporate Social
Responsibility: The Indian Experience, Journal of Contemporary Accounting &
Economics 14, 254–265.
Navarro, P., 1998, Why do corporations give to charity?, The Journal of Business 61, 65–
93.
Orlitzky, M., and J. D. Benjamin, 2001, Corporate social performance and firm risk: a
meta-analytic review, Business and Society 40, 369–396.
Park, C.-Y., and R. Mercado Jr., 2015, Financial inclusion, poverty, and income
inequality in developing Asia, Asian Development Bank Economics Working Paper
Series 426.
Peloza, J., 2006, Using corporate social responsibility as insurance for financial
performance, California Management Review 48, 52–72.
Pfeffer, J., and G. R. Salancik, 1978, The External Control of Organizations: A Resource
Dependence Perspective (Stanford Business Books, Stanford, CA).
Ramsey, J. B., 1969, Tests for specification errors in classical linear least-squares
regression analysis, Journal of the Royal Statistical Society: Series B (Methodological)
31, 350–371.
Richardson, G., B. Wang, and X. Zhang, 2016, Ownership structure and corporate tax
avoidance: evidence from publicly listed private firms in China, Journal of Contem-
porary Accounting and Economics 12, 141–158.
Safiullah, M., and A. Shamsuddin, 2018, Risk-adjusted efficiency and corporate
governance: evidence from Islamic and conventional banks, Journal of Corporate
Finance 55, 105–140.
Sarma, M., and J. Pais, 2011, Financial inclusion and development, Journal of
International Development 23, 613–628.
Servaes, H., and A. Tamayo, 2013, The impact of corporate social responsibility on firm
value: the role of customer awareness, Management Science 59, 1045–1061.
Shane, P. B., and B. H. Spicer, 1983, Market response to environmental information
produced outside the firm, The Accounting Review 58, 521–538.
Sufian, F., and M. S. Habibullah, 2010, Does economic freedom fosters banks’
performance? Panel evidence from Malaysia, Journal of Contemporary Accounting and
Economics 6, 77–91.
United Nations Conference on Trade and Development (UNCTAD), 2014, Bangladesh
Country Paper: Impact of Access to Financial Services. Available at: http://unctad.org/
meetings/en/Presentation/ciem6_2014_Bangladesh_en.pdf
van Beurden, P., and T. Gossling, 2008, The worth of values – a literature review on the
relation between corporate social and financial performance, Journal of Business
Ethics 82, 407–424.
Vig, V., 2013, Access to collateral and corporate debt structure: evidence from a natural
experiment, The Journal of Finance 68, 881–928.
Wang, H., and C. Qian, 2011, Corporate philanthropy and corporate financial
performance: the roles of stakeholder response and political access, Academy of
Management Journal 54, 1159–1181.
Wintoki, M. B., J. S. Linck, and J. M. Netter, 2012, Endogeneity and the dynamics of
internal corporate governance, Journal of Financial Economics 105, 581–606.
Appendix
Table A1
CSR expenditure, financial inclusion and firm performance
(continued)
Table A1 (continued)
Dependent variable: Firm performance measured by ROA and SR. Entries are the regression
coefficients and t-statistics (Huber-White) within parentheses derived from estimating the
models using the OLS method. The sample includes 149 banking firm-years over the period
2015–2017. Models 1, 2 and 3 are, respectively, estimated based on Equations (1), (2) and (3).
CSR, actual amount spent on CSR activities during the year; FI, financial inclusion; ROA,
return on assets; SR, stock return; SIZE, firm size; LEVERAGE, level of debt in capital
structure; ATO, asset turnover; PM, profit margin; DIRECTOR, proportion of independent
directors. All variables are explained in Table 1. ***, ** and * indicate statistical significance
at the 1, 5 and 10 percent level, respectively.
Table A2
CSR expenditure, financial inclusion and firm performance (common sample)
(continued)
Table A2 (continued)
Dependent variable: Firm performance measured by ROA and SR. Entries are the regression
coefficients and t-statistics within parentheses derived from estimating the models using the
OLS method. The results are based on a common sample of 44 firms over a period 2015–2017
(132 firm-year observations). Models 1, 2 and 3 are, respectively, estimated based on
Equations (1), (2) and (3). CSR, actual amount spent on CSR activities during the year; FI,
financial inclusion; ROA, return on assets; SR, stock return; SIZE, firm size; LEVERAGE,
level of debt in capital structure; ATO, asset turnover; PM, profit margin; DIRECTOR,
proportion of independent directors. All variables are explained in Table 1. ***, ** and *
indicate statistical significance at the 1, 5 and 10 percent level, respectively.
Table A3
CSR expenditure, financial inclusion and firm performance
(continued)
Table A3 (continued)
Dependent variable: Firm performance measured by ROA and SR. Entries are the regression
coefficients and t-statistics (standard) within parentheses derived from estimating the models
using the OLS method. The sample includes 149 banking firm-years over the period 2015–
2017. Models 1 and 2, respectively, include CSR expenditure and financial inclusion as
explanatory variables; models 3 and 4, respectively, include CSR expenditure and financial
inclusion along with the control variables; and models 5 and 6, respectively, include and
exclude year fixed effects incorporating CSR expenditure, financial inclusion and control
variables. CSR, actual amount spent on CSR activities during the year; FI, financial
inclusion; ROA, return on assets; SR, stock return; SIZE, firm size; LEVERAGE, level of
debt in capital structure; ATO, asset turnover; PM, profit margin; DIRECTOR, proportion
of independent directors. All variables are explained in Table 1. ***, ** and * indicate
statistical significance at the 1, 5 and 10 percent level, respectively.