Asia Pacific Management Review xxx (xxxx) xxx

H O S T E D BY Contents lists available at ScienceDirect

Asia Pacific Management Review

journal homepage: www.elsevier.com/locate/apmrv

Empirical analysis of the corporate social responsibility and financial performance

causal nexus: Evidence from the banking sector of Bangladesh
M. Adnan Kabir a, *, Sultana Sabina Chowdhury b
a
Caltech Global Consulting Limited, House 13, Block C, Main Road, Banasree, Rampura, Dhaka, 1219, Bangladesh
b
Sultana Sabina Chowdhury, Lecturer, ULAB School of Business (USB), University of Liberal Arts Bangladesh (ULAB), Bangladesh

a r t i c l e i n f o a b s t r a c t

Article history: In a culture like Bangladesh where social trust is low and corporate philanthropy is rare, corporate social
Received 8 November 2020 responsibility (CSR) engagements of the banking sector have been in a conspicuously upward trend. It
Received in revised form begs the question of whether the galvanization of CSR expenditure is validated by financial motivations
22 November 2021
or a genuine thirst for corporate philanthropy. The literature on the relationship between CSR and
Accepted 12 January 2022
Available online xxx
corporate financial performance (CFP) is without any overarching consensus. From an emotional view-
point, we revere corporate philanthropy without necessarily analyzing its financial merits. This study
extends the contemporary CSR literature horizon by examining 30 listed banks in Bangladesh from the
Keywords:
Corporate social responsibility
years 2006 through 2018, with particular emphasis on methodology that attempts to validate the CSR-
Banks CFP relationship. In addition to examining the bidirectional causality between CSR and financial
PVAR returns using Panel Vector Autoregression, the study examines the factor determinants of CSR. The study
Financial performance finds that better CFP leads to more CSR expenditure, but CSR expenditure does not necessarily influence
Bangladesh CFP. Moreover, net income, total deposits, return on asset, and previous year's CSR have a significant
positive relationship with CSR whereas firm age has a significant negative relationship.
© 2022 The Authors. Published by Elsevier B.V. on behalf of College of Management, National Cheng Kung
University. This is an open access article under the CC BY license (http://creativecommons.org/licenses/
by/4.0/).

1. Introduction anything tangible egged on by peer pressure? The academic anal-

ysis regarding CSR has primarily centered around these questions.
Corporate social responsibility (CSR) has been a crucial concept CSR is a way for firms to ensure mutual benefits between
in the fields of business and societal relations (Windsor, 2001). Over themselves and society (Nnaoma & Omotosho, 2017). It can work as
the past few decades, it has been considered in academic literature a rational strategy for corporate legitimacy and social control
as a mainstream and legitimate area of inquiry in understanding (Kamens, 1985). All corporate decisions are typically weighed
the relationship between corporations and their many stakeholders against its financial probity as a testament to its justification. There
in a societal context. Ever since its theoretical inception, many in- is a dichotomy of opinions concerning the undertaking of CSR.
vestors, corporations, consumers, and academicians have contem- Nobel Prize laureate Paul Samuelson (1971) vehemently supported
plated and evaluated the decision of businesses to engage in social the initiatives of corporations to engage in social responsibility. On
responsibilities (Moskowitz, 1972). Under this pretext, the notion of the contrary, fellow Nobel laureate Friedman (1970) wrote in a
CSR has been challenged with the questions: 1. Do businesses owe scathing article that espousing social responsibility to businesses in
something to the community on a moral and ethical level? 2. Does a free-market system is akin to ‘preaching pure and unadulterated
engaging in such socially responsible programs help firms finan- socialism’ and corporations should not be ‘spending someone else's
cially, at least in the long run? 3. Is CSR simply a social virtue lacking money for a general social interest’ (p. 33). Corporate policies and
practices are not without social consequences. Corporations are
vital centers of power and the decision they make have re-
percussions as it influences the lives of many people. Thus, there
* Corresponding author.
E-mail addresses: [email protected] (M.A. Kabir), sultana.sabina.chy@
are responsibilities that businesses are reasonably expected to as-
gmail.com (S.S. Chowdhury). sume (Bowen, 2013). The CSR practices reflect the values of an or-
Peer review under responsibility of College of Management, National Cheng ganization and paint an image in the minds of stakeholders - a
Kung University.

https://doi.org/10.1016/j.apmrv.2022.01.003
1029-3132/© 2022 The Authors. Published by Elsevier B.V. on behalf of College of Management, National Cheng Kung University. This is an open access article under the CC BY
license (http://creativecommons.org/licenses/by/4.0/).

Please cite this article as: M.A. Kabir and S.S. Chowdhury, Empirical analysis of the corporate social responsibility and financial performance
causal nexus: Evidence from the banking sector of Bangladesh, Asia Pacific Management Review, https://doi.org/10.1016/j.apmrv.2022.01.003
M.A. Kabir and S.S. Chowdhury Asia Pacific Management Review xxx (xxxx) xxx

perception the top management of organizations are well aware of always provide meaningful solutions to social issues, CSR initiatives
(Sturdivant & Ginter, 1977). This is evident in industries that are by organizations will go a long way in addressing such causes. CSR
more customer-centric such as banking, retail, telecom, etc. as they practices in Bangladesh are still in its inception (Alam et al., 2010).
have a higher proclivity to engage in corporate philanthropy (Xie, On top of that, prior research on Bangladeshi companies mostly
2015). fails to find any significant relation between CSR and financial
CSR refers to the acceptance of social obligation beyond the legal performance (Saha et al., 2013). Due to the scantiness of corporate
laws and regulations to maintain moral standards and ethical social engagements, the link between CSR and CFP in Bangladeshi
practices (Davis, 1973). Due to its varying nature and influence, banks has been unforeseeable to some extent (Islam et al., 2012).
market return from investing in CSR may be somewhat unpre- Against this backdrop, this study attempts to incorporate the
dictable. Researchers still debate whether a firm should engage in eclectic mix of CSR-CFP literature and investigate the possible
CSR unless it directly or indirectly increases the market return bidirectional relationship that may exist in the context of banks in
(Barnett, 2007). But to maintain the balance of the social ecosystem Bangladesh. Such a cross-tabulated representation of CSR factor
and long-term sustainability, CSR involvement is important for variability and implicit causal relationship between CSR and CFP in
corporations (Mintzberg, 1983). the context of Bangladesh is lacking in contemporary literature.
The relationship between corporate social performance (CSP) This is important since the banking sector in Bangladesh plays a
and corporate financial performance (CFP) has been analyzed for fundamental and dominant role as the financier of industrial and
decades. Despite these efforts, the results remain ambiguous commercial activities in the country (Khan, 2010). Moreover, the
(Schnippering, 2020; McWilliams & Siegel, 2000). For decades, paper also empirically investigates the factors that determine the
researchers have sought to divulge the mutual dependency of CSR progress or stifle of CSR and detail inferences. By considering Sys-
and firm performance. A significant contingent of the CSR literature tem dynamic panel-data estimations it becomes apparent that CSR
claims that corporate philanthropy influences market return and and CFP have a one-way causality as CFP influences CSR expendi-
numerous studies have established that CSR can leverage a firm's ture but not the other way round. CFP generally has a positive
profitability and market returns (Blasi et al., 2018; Suteja et al., inducing effect on the CSR expenditure of banks and follows a
2016; Ajide & Aderemi, 2014; Kempf & Osthoff, 2007). Blasi et al. quadratic relationship. Moreover, the study finds that net income,
(2018) found that there is a reduction in financial risk due to in- total deposits, and return on assets have a significant positive
vestment in almost all dimensions of CSR. A positive correlation relationship with CSR whereas the age of banks has a significant
between the CSR and CFP isomorphism is observed in growing inverse relationship. Earnings per share have limited capacity in
economies where the positive reputation of companies results in explaining the CSR expenditure dynamics. Additionally, the time
higher market return (Cheung et al., 2010; Ilhan-Nas et al., 2015). series of CSR expenditures of listed banks in Bangladesh follows a
Socially responsible companies gain a competitive edge as in- first-order autoregressive process [AR (1)]. CSR spending of year ‘t-
vestors prefer to invest in companies with higher social engage- 1’ has a highly significant positive effect on the CSR expenditure of
ment (Ramasamy et al., 2007). More investors are now considering year ‘t’.
fiduciary activities and CSR as business goals (Smith, 2005). The following section of the paper provides a theoretical review
Furthermore, increasing involvement in corporate philanthropy and empirical arguments of the CSR paradigm. The third section
can be exploited as a corporate advantage (Maqbool & Zameer, entails data and samples. The fourth section delves into the
2018). The CSR-CFP relationship can be thought of as a ‘virtuous empirical and methodological design. The fifth section details the
circle’ as it is anticipated that higher CSR disclosure will drive results and analysis. The final section discusses limitations and
performance upward as good social performance reduces a firm's sums up concluding remarks along with theoretical and practical
financial risk which in turn reinforces its commitment to CSR implications.
engagement (Chollet & Sandwidi, 2018; Waddock & Graves, 1997).
It must also be noted that there is still no overarching consensus 2. Literature review
when it comes to the CSR-CFP causal relation. Several researchers
found evidence that companies with higher returns are prone to 2.1. Review of CSR
contribute more to CSR (Vitezic, 2011; Cornett et al., 2016). When
firms have more resources available, especially cash, after paying The issue of corporate social responsibility has been debated in
out all operating expenses, the tendency of engaging in CSR the academic context for nearly a century. Even after years of
significantly increases (Seifert et al., 2003). Even the studies on CSR research, there is no single all-inclusive definition of CSR (Jackson &
and CFP relationship in emerging economies deliver mixed results Hawker, 2001). The challenge to businesses is not only defining CSR
(Senyigit & Shuaibu, 2017). The CSR and CFP correlation balance but also understanding the contextual variable constructions of CSR
diverges with changes in its financial stability (Ho et al., 2019). (Dahlsrud, 2008). Scholarly perceptions of CSR have also evolved
Moreover, dissenters of the CSR-CFP ethos argue that there is no over the years. CSR is a social construct molded by years of evolving
significant relationship between CSR and profitability (Saha et al., literature and such complex social constructs are difficult to define
2013; Goering, 2012; Seifert et al., 2003; Griffin & Mahon, 1997; under an umbrella of univocal consensus (Berger & Luckmann,
Alexander & Buchholz, 1978). Thus, due to the plethora of con- 1966). Rather than looking at an all size fits all definition of CSR,
flicting opinions, the reciprocity between CSR and firm perfor- a more context-based, historical, practical, and philosophical defi-
mance remains inconclusive. It could be that CSR increases profit, nition should be defined that is attuned with an organization's
or profits increase CSR initiatives, or a bidirectional causality be- awareness and ambition (van Marrewijk, 2003). If social initiatives
tween the two, or no discernible statistically significant relation- are not aligned with business objectives, then CSR investment may
ship whatsoever. become a liability for businesses in the long term (Lin et al., 2009).
Although CSR initiatives were largely thought of as an issue Be
nabou and Tirole (2010) detailed three views of CSR: (i) Win-
most prevalent in developed countries, its popularity has scaled up Win: a vision of simultaneously increasing profits and helping the
to a global context in recent years (Carroll & Shabana, 2010). That society, at least in the long run; (ii) Delegated philanthropy:
being said, research on CSR involvement in developing countries is engaging in philanthropy on behalf of stakeholders; (iii) Insider-
still quite inadequate and lacks empirical evidence (Visser, 2009). In initiated corporate philanthropy: a controversial yet common
a destitute country like Bangladesh where the government does not practice of giving to charities, institutions, political thinktanks
2
M.A. Kabir and S.S. Chowdhury Asia Pacific Management Review xxx (xxxx) xxx

which their top management favors or themselves chair. bidirectional in nature (Busch & Friede, 2018; Martínez-Ferrero &
The origin of CSR in the academic sphere can be traced to the Frías-Aceituno, 2015; Orlitzky et al., 2003). Moreover, Lougee and
seminal study of Sheldon (1924) where CSR was broadly articulated Wallace (2008) found CSR investment positively influences long-
as the voluntary engagement in environmental and social pro- term profitability that follows a ‘circular’ causality where profit-
grams. Ever since, numerous studies have investigated these re- able companies are more likely to invest in CSR initiatives to begin
lationships with CSR from many angles. Dahlsrud (2008) did an with, but then find their performance further improved by such
extensive review of CSR definitions from literature spanning from investment. The literature pertaining to this specific question may
1980 to 2003 and found five broad categories of dimensions: the diverge due to differences in selection bias, measurement
environmental dimension; the social dimension; the economic approach, or sampling limitation (Galant & Cadez, 2017).
dimension; the stakeholder dimension; and the voluntary dimen- The relationship between CSR and bank performance is subject
sion. In addition, Cochran and Wood (1984) argued that CSR mea- to a number of economic and institutional variables. It differs based
surement falls into two broad generally accepted methods - the on country, economic structure, and banking operations. Generally
first being a ‘reputation index’ and the second a ‘content analysis’. understood, a bank's commitment to CSR practices enhances
The reputation index ranks firms by their social involvement and the earnings and cash flow predictability (García-Sa
nchez & García-
content analysis analyses the extent of a firm's CSR activity Meca, 2017). In the case of Turkish banks, return on asset (ROA)
reporting in publications such as their annual reports. An early and return on equity (ROE) do not influence CSR. But a bidirectional
example of a reputation index was constructed by Moskowitz relationship exists between CSR and net interest margin (NIM),
(1972) who focused on racial diversity, gender diversity, and anti- where banks with higher NIMs are more likely to subsidize CSR
pollution efforts. Furthermore, Matten and Moon (2008) segre- activities (Taskin, 2015). From a global perspective, Wu and Shen
gated CSR into two distinctive categories: explicit and implicit. (2013) found that CSR has a positive correlation with ROA, ROE,
Explicit CSR demonstrates a company's responsibilities to net interest income, and non-interest income, but has a negative
contribute to social welfare while Implicit CSR is not a voluntary correlation with non-performing loans. A positive relation between
movement but one which relies on a reciprocal approach to the CSR and bank profitability is mostly prominent in developed
corporate environment. CSR must not be confused with Corporate countries due to high investor protection and stakeholder orien-
Sustainability (CS). There are considerable differences between the tation (Belasri et al., 2020; García-Sa
nchez & García-Meca, 2017).
two. While CSR involves a mutual association of people and orga- This is not necessarily the case for developing economies that have
nizations, CS is mainly concerned with the agency principle (van divergent economic structures compared to developed economies
Marrewijk, 2003). The best CSR practices are not confined to (Blowfield & Frynas, 2005).
monetary transactions as it must include measurable outcomes and
keeping track of records and progress on a regular basis (Porter & 2.3. CSR in Bangladesh
Kramer, 2006). Similar sentiments were echoed by Kurucz et al.
(2008) whose research indicated CSR is not about charity but The deficiency of literature on CSR in developing countries is
creating a mutual relationship where a firm delivers social welfare indicative of the fact that the practice of CSR is a recent phenom-
and society gives a positive return in kind. enon in these countries (Ramasamy et al., 2007). Unlike in the west
where a significant portion of CSR activities is directed towards
2.2. CSR and CFP social issues ranging from race to gender equality, firms in devel-
oping countries have a much narrower understanding of CSR. In a
A common understanding among stakeholders is that CSR has culture like Bangladesh, social trust is low and philanthropy is rare
positive externalities that improve the financial position of a firm. (Karim & Kasim, 2018). However, it is being promulgated with a
Kempf and Osthoff (2007), using a ‘Carhart four-factor model’ that change in mindset. Khan et al. (2009) examined the CSR reporting
entails positive and negative screenings, opined that investing in of banks in Bangladesh and found banks did little CSR reporting.
portfolios that extensively focus on CSR results in abnormally high The study also used a questionnaire to investigate the level of un-
returns. Even though the return varies depending on the area of derstanding and perception towards CSR reporting by managers,
investment, CSR involvement usually results in high stock returns shareholders, employees, suppliers, and customers, etc. to conclude
and low financial risk (Blasi et al., 2018). On the contrary, there is a that most stakeholders preferred more CSR disclosure. The CSR
range of studies accentuating the negative correlation between CSR sectoral expenditure is different in developing countries such as
and firm performance (Lin et al., 2019; Hirigoyen & Rehm, 2015; Bangladesh. It is mostly limited to contributions to natural disasters
Makni et al., 2009). The primary generic reason for the negative such as tornado or flooding, expenditure in education such as
correlation can be attributed to the fact that when firms act socially buying books for students, and scholarships (Khan, 2010). Social
responsibly the additional social expense curtails profit maximi- components such as racial or gender equality rarely fall under the
zation efforts. Although it must be noted there are those who CSR universe in Bangladesh. These issues rarely come up in the
disagree with this notion such as Kempf and Osthoff (2007), who national dialogue. Thus, corporations feel it makes little sense to
found that contrary to belief, socially responsible investing that incorporate such ethos into their organizational strategy. Having
includes community, diversity, employee relations, human rights, said that, the CSR expenditure of banks in Bangladesh is thriving
and environment, do not result in performance loss. Moreover, compared to the past. In 2018, the total CSR spending of banks in
there are also those who suggest that while there is a positive Bangladesh was around 21 percent more than the preceding year.
relationship between a firm's available cash resources and dona- The CSR expenditure in the banking sector is on an upward trend
tions, there is no significant relationship between corporate phi- due to changes in the market paradigm and a shift in the corporate
lanthropy and performance (Nelling & Webb, 2009; Seifert et al., culture of Bangladesh. The fact that banks get tax rebates for CSR
2003; Alexander & Buchholz, 1978). spending (Alo, 2019) could be adding to this upward trend.
Whether the CSR-CFP relationship follows a bidirectional cau- Developing countries are going through its infant stages of
sality is still a point of contention. Many studies found no such ‘bankization’ where low-income citizens are coming under the
evidence of bidirectional causality (Testa & D'Amato, 2017; Nelling umbrella of financial inclusion mainly through the advent of micro-
& Webb, 2009). But several research findings did demonstrate finance (Prior & Argandon ~ a, 2009). These new financial institutions
positive causality between CSR and CFP where the relationship is are themselves new to the CSR universe. Due to the lack of CSR
3
M.A. Kabir and S.S. Chowdhury Asia Pacific Management Review xxx (xxxx) xxx

awareness and seemingly no tangible benefit for Bangladeshi banks possibility of causality in this context can be statistically estimated.
in engaging in it, it is reasonable to surmise that a bidirectional The second involves sensitivity tests to examine the robustness of
causality between CSR and firm performance may not be present in the statistical applications. And the third is the set of panel re-
the Bangladeshi context. Moreover, based on the literature analysis gressions of CSR determinants and entrapments. The Autore-
and our understanding, CFP may play an influential role in deter- gressive process of order one [AR (1)] estimates the impact of time
mining the CSR expenditure, but CSR is unlikely to have any trends on a given year's CSR initiative. This also reverberates with
perceptible implication on the CFP. Thus, the study's research hy- the unit root tests.
potheses are stated as follows: The System dynamic panel AR (1) Model is denoted by the
formulation,
H1. The CFP and CSR expenditure of banks have no discernable
bidirectional causality. InCSRt ¼ a þ r InCSRt1 þ et
H2. CFP increases CSR spending. Here InCSR0 is the observed initial value, in this case, year 2006,
H3. CSR engagement does not positively influence CFP. and et is independent and identically distributed with 0 mean
under the pretext the estimated errors are uncorrelated with each
year's CSR. Thus, E (InCSR0) ¼ 0 and et ~ i.i.d, N (0, s2). Since et is
3. Data
assumed to be normally distributed, the OLS estimator, shown
below, concedes with the maximum likelihood estimator for q.
The study sample comprises 30 Private Commercial Banks
(PCBs) listed in the Dhaka Stock Exchange (DSE) in panel data from 1
PT
b T t¼1 LnCSRt1 LnCSRt
OLS ¼
the years 2006 through 2018. The compilation of this secondary Ø P
1 T LnCSR2
data was from the banks’ annual reports and the dataset can be T t¼1 t1
found in Kabir and Sunon (2021).
Next, the study considers a k-variate homogeneous panel VAR of
As per Bangladesh Bank's regulation, both PCSs and State-
order 1 with panel-specific effects demonstrated by a system of
Owned Commercial Banks (SOCBs) engage in CSR to some extent
matrix functional form. This VAR(1) allows us to contemplate the
but compared to PCBs, SOCBs' contribution to CSR is strikingly low
causal relationship between CSR and ROA that is discussed exten-
(Roy et al., 2017; Ndiweni et al., 2018). SOCBs in Bangladesh are
sively in the literature. It helps us to estimate and observe whether
mostly sunk in bad loans due to the lack of initiatives in recovering
changes in one of the variables are correlated with changes in other
repayments and substantial capital investment in classified loans.
variables either contemporaneously or over time. The functional
Most of the SOCBs are surviving this financial predicament by solely
equation form follows,
depending on taxpayers' money. As the board members and exec-
utive committees are under no pressure to answer for their InCSRt ¼ a11 InCSRt1 þ a12 ROAt1 þ εInCSR;t
mismanagement it is highly unlikely that SOCBs will display better InROAt ¼ a21 InCSRt1 þ a22 ROAt1 þ εROA;t
performance in the future (Ahmed, 2017). SOCBs use government
funds to recover their shortfall and the lack of accountability cre- This would translate to vector-matrix as
ates resistance to structural changes (Haroon, 2017). It is for these     
reasons that the study shies away from using SOCBs in our models LnCSRt C11 LnCSRt1 C12 ROAt1 εLnCSR; t
¼
and focuses on PCBs as they are the only contingent of the Ban- ROAt C21 LnCSRt1 C22 ROAt1 εROA;t
gladeshi banking sector truly driven by the profit motive. PCBs have
" # " #
a philanthropic and profit-maximizing angle to their CSR engage- a11 a12 LnCSRk;t
ments that make it a better test case for the study. A list of the yt ¼ þ εt
a21 a22 ROAk;t
sample banks can be found in Table 7 (Appendix) along with the
number of branches and employees as a point of reference.
The corporate social responsibility initiatives are the total CSR yt ¼ ak yt1 þ εt
expenses in a year and the dependent variable of concern. The
subsequent independent variables are net income (NI), total de- where yt is a (2  1) vector of dependent variables, ak is a (2  2)
posit (TD), earnings per share (EPS), return on asset (ROA), and matrix of coefficient parameter vectors to be estimated, and yt-1 is a
years of operation as a proxy for firm maturity (MAT). ROAt-1 is a (2  1) vector of 1 year lagged dependent variable transformed as
year-lagged ROA value. ROA is calculated in percentage, EPS is a proxies for the independent variable. εt is a (2  1) error vector of
ratio, MAT is in years, and all other variables are denoted in Ban- specific panel idiosyncratic errors encompassing exogenous
gladeshi taka1 (BDT). A description of data and the pairwise cor- covariates.
relation matrix are shown in Tables 1 and 2, respectively. To account
for the possibility of bias arising from the correlation of one or more
4.2. Unit-root test
of the independent variables, the Variance Inflation Factors (VIF)
are estimated. The VIFs from Table 2 show no persistent problem of
To add to the methodological consistency of VAR application, a
multicollinearity among the regressors.
test on the time trend unpredictability relationship between the

4. Research design Table 1

Descriptive statistics.
4.1. Dynamic panel estimations Variable Mean Std.Dev. Min Max

CSRa 121 157.2 2.39 1057.7

The study uses a multi-step methodological approach. The first NIa 2121.6 1469.2 1211.7 9684.8
employs the use of dynamic panel estimations that includes system TDa 164,000 104,000 16775.34 754,000
dynamic panel-data estimations and panel vector autoregression ROA 1.118 .617 -.362 4.4
(PVAR). This reveals the continuity of CSR dependency with a year MAT 20.8 11.718 4 58
a
lag and the direction of influence between CSR and ROA. Any Values in million BDT.

4
M.A. Kabir and S.S. Chowdhury Asia Pacific Management Review xxx (xxxx) xxx

Table 2
Pairwise correlation & variance inflation factor.

Pairwise Correlation VIF

Variables CSR NI TD EPS ROA ROAt-1 Mat VIF Tolerance

CSR 1.000
NI 0.359a 1.000 5.65 0.177
TD 0.812a 0.434a 1.000 3.44 0.290
EPS 0.119 0.080 0.224a 1.000 1.12 0.892
ROA 0.125 0.569a 0.304a 0.386a 1.000 6.36 0.157
ROAt-1 0.178 0.259a 0.297a 0.324a 0.656a 1.000 1.63 0.613
MAT 0.025 0.133 0.308a 0.102 0.168a 0.171a 1.000 1.13 0.884
Mean VIF 3.22
a
Denotes significance of 0.05.

two variables is adopted using the Fisher-type unit-root test. This omitted variable bias and model misspecification. Furthermore, the
unit root form factor is shown in Table 3. tests for heteroskedasticity further authenticate the regression in-
Comparing the test statistics with a benchmark p-value of 5 ferences. Moreover, the Skewness-kurtosis Test for assessing the
percent significance level validates that CSR, ROA, and NI are non- normality of fitted residuals confirms standard regression as-
stationary at the level and stationary in the first difference form. sumptions. This inference can also be observed in Fig. 1 (Appendix),
which shows the randomness of predictive errors, and Fig. 2 (Ap-
pendix), which shows the normality of residuals with no discern-
4.3. Cointegration & robustness able cyclical pattern. The results of all the aforementioned tests
show no imposing econometric deficiency or any discernible sta-
Table 4 shows the tests of the robustness of the models with tistical anomaly.
multiple alternative and contemporaneous specifications. The
panel VAR paradigm entails the composite validation of co- 4.4. Panel regressions of CSR determinants
integration, estimated here using the Westerlund (2005) test for
cointegration. The panel regressions of CSR determinant and entrapment
Stationary panel data have a time-invariant mean and variance. models (shown in Table 6) is a linear multivariate estimation
Non-stationary panel data have a time-variant mean or time- focusing on the predictors of CSR and demarcating the said models
variant variance or both. When the first difference of a non- with the following structured formulations,
stationary process is stationary, the process is said to be inte-
grated of order 1. And when a linear combination of several order 1 DLnCSRit ¼ a0 þ b1 DLnNIit þ b2 DLnTDit þ b3 DEPSit þ b4 DROAit
series is stationary, the series is said to be cointegrated. Cointe- þ b5 DROA2 þ Dmit
gration implies that the series move together in long-run equilib-
rium, although a group of that can wander arbitrarily. Westerlund (1)
(2005) assumes that panel-specific cointegrating vectors have in-
dividual slope coefficients. Here the statistic is used to test the null LnCSRit ¼ ai þ b1 LnNIit þ b2 LnTDit þ b3 EPSit þ b4 ROAit
hypothesis (H0) of no cointegration against the alternative hy- þ b5 ROAt1 þ b6 ROA2 (2)
pothesis (Ha) that all the panels are cointegrated. Thus, the high p-
þb7 ROA2t1 þ mit
value (>0.5 from Table 4) is indicative of the presence of cointe-
gration through the rejection of the null hypothesis. This implies
there is a short-term relationship between CSR and ROA, and not a LnCSRit ¼ ai þ b1 LnNIit þ b2 LnTDit þ b3 EPSit þ b4 ROAit
long-term one. þ b5 ROAt1 þ b6 ROA2 þ b7 ROA2t1 þ b8 MATit þ mit
Ramsey regression equation specification error test (RESET),
also shown in Table 4, substantiates the OLS estimation is free of (3)

Table 3
Fisher-type unit-root test.

ADF regressions: 1 lag Inverse Chi2 Inverse normal Inverse logit t Modified inverse Chi2 Implication

Variable Form P Z L PM

CSR Level Statistics 0.7060 4.0370 4.6078 4.6152 Unit root non-stationary
p-value 1.000 1.000 0.9999 1.000
Differenced Statistics 15.0161 2.6328 2.9802 2.1230 Stationary
p-value 0.9953 0.0042 0.0050 0.9831

ROA Level Statistics 16.7028 1.8000 1.7271 3.4612 Unit root non-stationary
p-value 1.0000 0.9641 0.9553 0.9997
Differenced Statistics 92.0501 3.9204 8.6573 5.4609 Stationary
p-value 0.0000 0.0000 0.0000 0.0000

NI Level Statistics 18.59 0.3803 0.5719 3.2760 Unit root non-stationary

p-value 1.0000 0.6482 0.7151 0.9995
Differenced Statistics 48.1460 2.4179 3.7517 0.6706 Stationary
p-value 0.2381 0.0078 0.0003 0.2512

Ho: All panels contain unit roots Ha: At least one panel is stationary.

5
M.A. Kabir and S.S. Chowdhury Asia Pacific Management Review xxx (xxxx) xxx

Table 4
Robustness tests.

Tests Chi2/ df P H0 Implication

Statistic

Hausman specification test 14.27 7 0.0466 Difference in coefficient not Fixed effect is appropriate
systemic
White's test 43.76 42 0.3966 Homoskedasticity No unrestricted
heteroskedasticity
Cameron & Trivedi's decomposition of IM- Heteroskedasticity 35.29 33 0.3603 Normality Presence of Normality
test Skewness 10.94 7 0.1414
Kurtosis 1.38 1 0.2401
Joint 47.61 41 0.2216
Breusch-Pagan/Cook-Weisberg test for Variables: fitted values of lnCSR 3.47 1 0.0624 Constant variance Constant variance
heteroskedasticity
Westerlund test for cointegration Cointegrating vector & AR parameter: 4.7984 0.1469 No cointegration Panels are not
Panel specific cointegrated
RESET Y F
statistics
X Yh2 0.308 (1, 0.5824 Model is specified No model mis-
37) specification
X Yh2 Yh3 0.981 (2, 0.3848
36)
X Yh2 Yh3 Yh4 1.783 (3, 0.1684
35)

These are panel data regressions that delve into the factors that deduced through the Hausman (1978) specification test which
determine CSR progression. These include first difference (FD), revealed the fixed effect estimation to be the more appropriate
estimated by equation (1). Fixed effect (FE); fixed effect with be- statistical fit. Thus, the validation led us to focus on the standard
tween estimator (Febe); and fixed effect with Autoregressive order one-way fixed-effect (FE).
1 [FR,AR (1)], estimated by equation (2). Equation (3) estimates
pooled OLS (POLS), and random effect (RE). 5. Results and analysis
POLS lumps all our observations together and estimates the
model under the guise of one large cross-section. The first differ- 5.1. System dynamic panel-data estimations
enced equation is a cross-sectional equation with differenced var-
iables over time. The assumption here is that changes in the Table 5 shows two salient system dynamic panel-data estima-
idiosyncratic error, Dmi, is uncorrelated with Dxi given that mi and xi tions: Autoregressive Model of Order 1 [AR (1)] & Panel vector
are also uncorrelated in both the years. Here, each variable is dif- Autoregression (PVAR). AR (1), in its basic form, shows the influ-
ferenced over time. The fixed effects, on the other hand, carries ence of past CSR on current CSR. The result implies previous year's
with it the notion of removing the effects of unobserved hetero- CSR growth has a highly significant, linear, and positive effect on
geneity. More specifically, ai is the unobserved fixed effect, illus- the current year's CSR expenditure. The implication here is that a 1
trated without the transcript ‘t’ as it does not change over time. percent increase in CSR at year ‘t-1’ induces about a 0.8 percent
Moreover, the error, mit, is the idiosyncratic error that represents the increase of CSR in year ‘t’. This makes practical sense in the context
changing effect on LnCSRit over time. Both of these culminate as the of Bangladeshi banks as the upwards trend in CSR expenditure is an
composite error, hit, with the goal of eliminating ai if it is correlated indication of the banking sector's cultural shift towards corporate
with one or more of the Xitj. The severity of heterogeneity that philanthropy (Bangladesh Bank, 2015). Decisions of CSR investment
banks can exhibit due to any unseen exogenous forces can be are based, at least, partially on the previous year's CSR. It is likely
reduced by adopting a fixed-effect model that includes the entire managers choose to outspend their previous year's CSR expendi-
banking sector-specific factors. Moreover, it should also help in ture to showcase their commitment towards social responsibility to
dealing with possible problems of endogeneity due to unobserved all stakeholders involved in this competitive sector.
fixed effects. Equation (3) translates into a random effect, under the The Panel vector Autoregression (PVAR) is an assessment of the
Generalized Least Square (GLS) paradigm, with the benefit of esti- equilibrium dependency of CSR and ROA and the direction of cau-
mating the effect of the time constant variables on CSR under the sality. This is a protraction of CSR's relationship with its lagged
premise Cov(ai,Xit) ¼ 0 with a lower standard error from that of the value. The estimation shows CSR value in year ‘t’ is not necessarily
fixed effect estimations as se (b^RE) ˂ se (b^FE). Under the random dependent on ROA. ROA of year ‘t’, on the other hand, seems to be
effect formulation, hit are serially correlated across time as it con- independently determined from both CSR and its previous year
tains ai in each time period. a is the unobserved heterogeneity and ROA values. This implies no inferences can be drawn as the esti-
is assumed to be uncorrelated with the regressors. Here, the vari- mation fails to register any statistical significance. Thus, favorable
ations across banks are assumed to be random and uncorrelated past performance does not necessarily have an inducing effect on
with the explanatory variables. The assumption is that the error ROA. The absence of reciprocal causality between CSR and ROA in
differs stochastically. It is possible that the banks could exhibit the banking sector of Bangladesh may be indicative of the lack of
notions of heterogeneity with regards to factors outside our anal- socially aware investing. This is understandable since, culturally
ysis. These could very well be specific to the banking sector in speaking, the concept of socially responsible investment in
Bangladesh. This conundrum is resolved through the RE. Bangladesh is yet to take off.
The rationale behind the multi-angular dissection of CSR is to Though not specific to the context of Bangladesh, our findings
shed a richer light on the factors of CSR under varying statistical are in lockstep with studies that have walked along the same
anomalies. The appropriateness of fixed or random effect was empirical path. Studies such as Lin et al. (2019) found better
6
M.A. Kabir and S.S. Chowdhury Asia Pacific Management Review xxx (xxxx) xxx

Table 5
Autoregressive model of order 1 & panel vector autoregression.

AR (1) PVAR

Dependent Variable Dependent Variable

Variables lnCSRt (Ln)CSRt ROAt (Ln)NIt

(Ln)CSRt-1 0.797*** (0.156) 0.299 (0.286) 0.051 (0.077) 0.165 (0.187)

ROAt-1 1.23*** (0.334) .375*** (0.129) 0.278 (0.266)
(Ln)NIt-1 2.28*** (.718) 0.165 (0.261) 0.365 (0.375)
Constant 3.963 GMM Estimation. GMM weight matrix: Robust
Number of obs 102 70
No. of panels 26 25

Standard errors in parentheses.

*p < 0.1, **p < 0.05, ***p < 0.01.

financial performance leads to better CSR engagement, but CSR, in which some studies have posited does not exist in the banking
turn, does not necessarily improve financial performance. Inter- sector of Bangladesh.
estingly enough, ROA and lagged CSR have a negative relationship.
Although not statistically significant the orientation point towards
CSR of the past year as having a deracinating effect on ROA. This is at 5.2. CSR factor determinants
odds with literature that attest to the financial merits of CSR
engagement. One such qualification was espoused by Ramasamy Table 6 focuses on the analysis of the explanatory factors that
et al. (2007) who found that the positive impression CSR engage- affect CSR. It is worth noting that financial performance indicators
ments create on consumers, employees and other stakeholders generally exhibit a significant relationship with CSR, with some
enables CSR active firms to outperform their counterparts. exceptions. Among the significant estimations is the strength of NI.
Furthermore, lagged CSR has limited capacity in influencing year The estimations show that NI is a significant predictor of CSR across
t's ROA and NI, thereby confirming there is no presence of bidi- the board, a phenomenon consistent with prior literature such as
rectional causality. Although we fail to establish a bidirectional Sun (2012). This makes economic and intuitive logic as an increase
relationship between CSR and CFP, there is still proof of a one- in the net income induces an increase in CSR expenditure. On the
dimensional relationship. Thus, the concept of a ‘virtuous circle’ other hand, ROA, an additional indicator of a firm's control over its
finances, has a statistically significant non-linear effect. The fixed

Table 6
CSR determinants & panel regressions.

Models (POLS) (FD) (FEbe) (FE) (RE) FE,AR (1)

(Ln)NI .905*** .595** 1.422 .876*** .962*** .87***

(.161) (.232) (.851) (.169) (.713) (.197)

(Ln)TD .893*** .338* .302 .624*** .713*** .651***

(.177) (.192) (.92) (.196) (.191) (.213)

EPS -.037 -.041** -.126* -.017 -.022* -.029

(.029) (.018) (.068) (.014) (.013) (.048)

ROA 1.376* -.826 2.572 2.753*** 2.36*** 2.722***

(.806) (.593) (3.713) (.679) (.670) (.844)

ROAt-1 -.322 2.939 -.753* -.482 -.837

(.388) (2.234) (.381) (.367) (.569)

ROA2 .28 .147 1.172 .735*** .597*** .74***

(.22) (.117) (1.042) (.19) (.188) (.235)

ROA2t-1 .032 .653 .131 .070 .177

(.073) (.526) (.084) (.081) (.193)

MAT -.03*** -.029***

(.01) (.011)

Constant 22.156*** 18.663 14.191*** 15.461*** 14.725***

(4.607) (11.682) (5.206) (4.684) (5.053)

N 114 99 114 114 114 85

R2 Within 0.2419 0.5704 0.5311 0.4846
Between 0.6190 0.2233 0.4907 0.4891
Overall 0.603 0.179 0.3706 0.4694 0.5639 0.5078

Standard errors are in parentheses.

***p < 0.01, **p < 0.05, *p < 0.1.

7
M.A. Kabir and S.S. Chowdhury Asia Pacific Management Review xxx (xxxx) xxx

effect estimation shows that the relationship between ROA and CSR and business ethics contingent which takes precedence over a
is a quadratic one. In other words, banks do more in terms of CSR financial motivation. This is evident by the fact that CSR and CFP
activities only when it experiences a ROA of more than 1.87 percent. have a one-way causality where CFP improves CSR, but CSR has no
Moreover, a 1-year lagged ROA follows a similar non-linear rela- influence on CFP. Thus, banks look at their financial health in
tionship. The CSR expenditure only increases when the previous determining CSR expenditure but do not necessarily get any
year's ROA is 2.87 percent or more. The implication here is that financial reward from it.
banks view previous year's ROA of less than 2.87 percent as too low
to warrant any increase in their current CSR expenditure. Thus, CSR 6. Discussions
expenditure partially takes its cue from the previous year's ROA.
Another noteworthy observation is the negative relationship Do CSR and CFP have a bidirectional causality? This is the basic
between CSR expenditure and firm age. The relationship between premise the study attempts to analytically assess. The research used
CSR and financial performance cannot be accurately evaluated a multidimensional approach to dissect the CSR and CFP relation-
without taking asset age into consideration as the extent of CSR is ship into an exercise of formal logic. The aim was to integrate the
moderated by firm age (Hu et al., 2018). This theory is also sup- empirical literature and validate the theoretical predictions of CSR.
ported by Sun (2012) who concluded that the age of long-term The study found no discernible bidirectional causality between CSR
assets significantly impacts CSR. Cochran and Wood (1984) postu- and CFP. Additionally, CFP has an element of influence on CSR, but
lated that firms with older assets have lower engagement in CSR CSR, in turn, does not influence CFP. More specifically, CFP has a
activities. Our estimation is in line with such analysis as evident by quadratic effect on CSR. The study is further expanded by investi-
the statistically significant negative coefficient of MAT. Thus, in the gating the factor determinants of CSR. The findings show that NI,
context of Bangladesh, all else held constant, each additional year ROA, and TD have a significant positive impact on CSR expenditure.
the bank is in operation, CSR expenditure is reduced by a little less Bank maturity (MAT), on the other hand, has a statistically signif-
than 3 percent. It must be stressed that the relationship between icant inverse relationship with CSR. Overall, consistent with prior
CSR and firm age remains a point of contention. It is believed that literature, the results reveal the presence of a generally positive
small firms dawdle in their CSR initiatives and portray a weaker association between CSR and financial performance, with the
proclivity to engage in philanthropic activities due to inadequate exception of MAT and EPS. On top of financial performance being a
financial resources, lack of corporate experience, and unstable possible guiding determinant of CSR expenditure, banks use their
market positioning (D'Amato & Falivena, 2019; Badulescu et al., past CSR engagement as a reference in determining their current
2018). Even though there is a general understanding that firm CSR engagement. More specifically, the study finds that the previ-
maturity increases the extent of CSR disclosure, firm age is an ous year's CSR has an inducing effect on the current year's CSR
inconclusive indicator to predict the overall CSR involvement expenditure as the CSR expenditures of listed banks in Bangladesh
(Bidari, 2016). The CSR-Firm age dichotomy must be analyzed on a follow a first-order autoregressive process. The study's overarching
case-by-case basis. Depending on firm age, the causality of CSR and empirical estimations are generally in lockstep with theoretical
CFP also shifts. Research has shown that a firm's age induces CSE predictions. In summation, the results orient us to the narrative
practices (Basuony et al., 2014; Fahad & Nidheesh, 2020). In general, that the banks in Bangladesh engage in CSR primarily for social
matured firms have a higher tendency to invest in CSR even though responsibility as financial motivations may not be the sole driver of
firm age affects different categories of CSR in varying ways CSR initiatives.
(Withisuphakorn & Jiraporn, 2015). Although this may be true in a In a historical context, CSR has shown remarkable long-term
broader global context, the CSR perception dynamic in Bangladesh advantages (Lin et al., 2009). Generally understood, CSR has many
is different. In a culture where CSR awareness is low but growing, soft advantages beyond financial payoffs such as boosting social
mature and established banks are less motivated to allocate funds media campaigns and helping attract socially responsible investors
for CSR as they are more secure of their brand image and market both at home and abroad. It is an unfortunate epilogue that such is
positioning. For older firms, CSR turns into more of an obligation not the case in the banking sector of Bangladesh as CSR engage-
rather than adding value to the firm's performance (Han & Kim, ment shows no discernible financial gain. It seems that banks are
2020). Arora and Soni (2017) found that new firms are more influenced more by social cues than financial health when to comes
prone to make CSR investments in order to create a brand image. to CSR engagement. Regardless, pragmatism should be the func-
This could also hold true in the Bangladesh banking sector where a tioning ethos. A ‘one solution fits all’ definition should be avoided,
bank that has many years of operation under its belt does not feel and situation-based CSR awareness should be the goal (van
much of a pressing need to announce its presence through CSR Marrewijk, 2003). Banks in Bangladesh should strike a balance
engagement. without comprising their returns or their social responsibility. In
Furthermore, the total deposit (TD) coefficients show expected many companies, management truly gives an effort to conduct their
signs. TD has an alleviating effect, implied by the positive and business ensuring social and environmental welfare (Vogel, 2005).
significant relationship with CSR. This relationship is not hard to Individual managers can make changes in policies to accommodate
fathom as more deposits help banks to amass more funds for CSR social and benevolent work ethic (Hemingway & Maclagan, 2004).
engagement. Additionally, Table 6 shows no satisfactory results Moreover, research has found that customer citizenship behavior is
were obtained for EPS. The earnings per share have a deracinating positively influenced by the customer's perception of CSR (Hur
effect on CSR expenditure. This was the sole exception to our pos- et al., 2018; Swaen & Chumpitaz, 2008). With the advent of time
itive CSR-CFP relationship. It could be that when banks allocate and more CSR awareness, these impressions can develop consumer
their funds towards outstanding shares of their common stocks, trust and help banks in gaining a competitive advantage.
they do so at the expense of their CSR funds. The study has some limitations which may offer some insights
Collectively, the findings imply that CSR engagement decisions into opportunities for future research. One major challenge is the
of banks in Bangladesh are based more on social cues than financial measure of CSR. There are inconsistencies in findings and in-
health. This is understandable as financial betterment has a natural terpretations that often occur in the literature due to operational
disposition towards more open-handed philanthropy. The results and methodological discrepancies (Griffin & Mahon, 1997). Social
of this analysis have practical applications as we come to under- performance, which is subjective may be difficult to measure.
stand that banks’ engagement in CSR has a corporate stewardship While a company may be doing well in one aspect, such as
8
M.A. Kabir and S.S. Chowdhury Asia Pacific Management Review xxx (xxxx) xxx

pollution control, it may be behind in another aspect such as mi- policies, and sound management system (Jain et al., 2014). Sus-
nority hiring (Moskowitz, 1972). In the Bangladeshi context, the tainable and responsible investment ensures long-term financial
CSR disclosure of different companies remains questionable due to growth, keeping social, environmental, and corporate governance
the lack of statutory support (Ullah & Rahman, 2015). A more in- strategies in check (Bihari & Pradhan, 2011). Finally, it must also be
depth look at the quality of disclosure unveils that almost 75 noted that despite the vast body of research on CSR and financial
percent of disclosures are qualitative in nature without any data to performance, additional analysis is warranted.
support the claim (Azim et al., 2009). It is also important that CSR
reporting be perceived as credible (Miras-Rodríguez et al., 2019). Notes
Specific emphasis on corporate laws and regulations of Bangladesh
have revealed that legal protocols do not have well-formulated As a point of reference, 1 USD is equivalent to BDT 84.73 (Yahoo!
strategies to develop responsible corporate social culture (Khatun, Finance, 2020).
2014; Yesmine & Bhuiyah, 2015). Such considerations should be
taken into account to further encompass the external validity of any Declarations of interest
future study.
The CSR ethos has an ethical conundrum connotation. CSR We have no conflicts of interest to disclose with respect to our
spending has to come from somewhere and it is mainly the cus- article.
tomers who are footing the bill. When it comes to lending, banks
with high CSR concerns charge a premium to borrowers as they try Funding
to offset the financial costs involved with social responsibility en-
gagements (Goss & Roberts, 2011). In addition, CSR initiatives that This research did not receive any specific grant from funding
surreptitiously engage in advertisement and promotional schemes agencies in the public, commercial, or not-for-profit sectors.
raise many eyebrows and question the sincerity of corporate phi-
lanthropy. CSR may just be a way for corporations to distinguish Appendix
themselves from others without any meaningful altruistic motive.
Moreover, corporate political campaign contributions under the
guise of CSR come dangerously close to corporate illegality.
As financial markets are not well-formed in developing econo-
mies, financial institutions have to bear a disproportionate burden
of the CSR load. Kitzmueller and Shimshack (2012) state that under
certain economic and non-economic conditions, CSR may produce
higher welfare than public and even some private channels. In
developing countries where funds allocated for development pur-
poses are inadequate, many social and economic responsibilities
are shouldered by financial institutions (Prior & Argandon ~ a, 2009).
In Bangladesh, contribution to CSR activities is relatively insignifi-
cant in proportion to the bank's profit (Saha et al., 2013). Even
though the intended consequences of CSR may not be in lockstep
with its intended purpose, Bangladesh could still benefit if banks
increase and improve their CSR engagement. It is optimistic that
banks in Bangladesh emphasize on disclosure of CSR activities with
a growing concentration on poverty alleviation, safety, security,
public health, and community facilities (Khan et al., 2009). In
Fig. 1. Pattern of Predictive Errors
developing countries, contributions from banks and large corpo-
rations in the form of CSR can help to eradicate existing social
problems like illiteracy, environmental catastrophe, healthcare in-
adequacy, etc. (Sharif & Rashid, 2014).
Over the years, activist organizations have become more vocal
and effective in galvanizing public pressure on corporations to
engage in social responsibility and ethical standards. Many consider
it a noble way for corporations to behave (Mintzberg, 1983). Thus, it
is important we seamlessly blend our theoretical and empirical
understanding of CSR with reasoning while untangling the ca-
cophony of viewpoints present in the contemporary literature. It is
a good sign that banks in Bangladesh are making their CSR
disclosure more visible (Khan et al., 2009). For maximum effect,
CSR must be correctly communicated and have public visibility. The
quality of CSR disclosure in banks has conspicuously ameliorated
over the years (Krasodomska, 2015), and multi-level factors have to
be taken into account for banks to prove the legitimacy of CSR
disclosure. This includes, but is not limited to, maintaining sus-
Fig. 2. Normality of Residuals
tainability standards, ethical reporting, adoption of environmental

9
M.A. Kabir and S.S. Chowdhury Asia Pacific Management Review xxx (xxxx) xxx

Table 7
List of Banks used in models

Banks Number of Branches Number of employees As of

AB Bank Limited 105 2210 2019

Al-Arafah Islami Bank Limited 129 2810 2019
Bank Asia Limited 125 2463 2020
BRAC Bank Limited 187 8160 2019
Dhaka Bank Limited 87 1524 2019
Dutch-Bangla Bank Limited 208 10,022 2020
Eastern Bank Limited 85 3044 2020
Export Import Bank of Bangladesh Limited 130 2956 2019
First Security Islami Bank Limited 178 3565 2019
ICB Islamic Bank Limited 33 e 2020
IFIC Bank Limited 154 3011 2020
Islami Bank Bangladesh Limited 357 16,807 2019
Jamuna Bank Limited 132 e 2020
Mercantile Bank Limited 150 2428 2020
Mutual Trust Bank Limited 118 2340 2019
National Bank Limited 209 4740 2019
National Credit and Commerce Bank Limited 121 2156 2020
One Bank Limited 103 2414 2019
Premier Bank Limited 120 1992 2020
Prime Bank Limited 146 3090 2020
Pubali Bank Limited 482 8153 2019
Rupali Bank Limited 583 5490 2020
Shahjalal Islami Bank Limited 132 2657 2020
Social Islami Bank Limited 161 2947 2019
Southeast Bank Limited 135 2885 2019
Standard Bank Limited 138 2353 2019
The City Bank Limited 121 4356 2020
Trust Bank Limited 114 1940 2019
United Commercial Bank Limited 195 4988 2019
Uttara Bank Limited 241 3801 2020

Source: Annual Reports and bank websites

References Economica, 77(305), 1e19. https://doi.org/10.1111/j.1468-0335.2009.00843.x

Berger, P. L., & Luckmann, T. (1966). The Social Construction of Reality. Penguin
Group.
Ahmed, A. (2017, October 11). SoCB business: Giving away people's money! the
Bidari, G. (2016). Factors Affecting CSR Disclosure in Nepalese Banks: A Global
Financial Express. https://thefinancialexpress.com.bd/views/columns/socb-
Reporting Initiative Perspective. Research Online. https://ro.ecu.edu.au/theses/
business-giving-away-peoples-money-1507737521.
1803.
Ajide, F. M., & Aderemi, A. A. (2014). The effects of corporate social responsibility
Bihari, S. C., & Pradhan, S. (2011). CSR and performance: The story of banks in India.
activity disclosure on corporate profitability: Empirical evidence from Nigerian
Journal of Transnational Management, 16(1), 20e35. https://doi.org/10.1080/
commercial banks. IOSR Journal of Economics and Finance, 2(6), 17e25. https://
15475778.2011.549807
doi.org/10.9790/5933-0261725
Blasi, S., Caporin, M., & Fontini, F. (2018). A multidimensional analysis of the rela-
Alam, S. S., Hoque, S. M., & Hosen, M. Z. (2010). Corporate social responsibility (CSR)
tionship between corporate social responsibility and firms' economic perfor-
of MNCs in Bangladesh: A case study on GrameenPhone ltd. Journal of Potua-
mance. Ecological Economics, 147, 218e229. https://doi.org/10.1016/
khali University of Science and Technology, 1e11. https://ssrn.com/
j.ecolecon.2018.01.014
abstract¼1639570.
Blowfield, M., & Frynas, J. G. (2005). Setting new agendas: Critical perspectives on
Alexander, G. J., & Buchholz, R. A. (1978). Corporate social responsibility and stock
corporate social responsibility in the developing world. International Affairs,
market performance. Academy of Management Journal, 21(3), 479e486. https://
81(3), 499e513. https://doi.org/10.1111/j.1468-2346.2005.00465.x
doi.org/10.5465/255728
Bowen, H. R. (2013). Social responsibilities of the businessman. University of Iowa
Alo, J. N. (2019, February 22). Banks' CSR spending soars despite lower profits. The
Press.
Daily Star. https://www.thedailystar.net/business/banking/news/banks-csr-
Busch, T., & Friede, G. (2018). The robustness of the corporate social and financial
spending-soars-despite-lower-profits-1705639.
performance relation: A second-order meta-analysis: Corporate social and
Arora, A., & Soni, T. K. (2017). Corporate social responsibility and firm characteris-
financial performance. Corporate Social Responsibility and Environmental Man-
tics: Evidence from BSE 500. International Journal of Information, Business and
agement, 25(4), 583e608. https://doi.org/10.1002/csr.1480
Management, 9(1), 119e140. https://ijibm.elitehall.com/IJIBM_Vol9No1_
Carroll, A. B., & Shabana, K. M. (2010). The business case for corporate social re-
Feb2017.pdf.
sponsibility: A review of concepts, research and practice. International Journal of
Azim, M. I., Ahmed, S., & Islam, M. S. (2009). Corporate social reporting practice:
Management Reviews, 12(1), 85e105. https://doi.org/10.1111/j.1468-
Evidence from listed companies in Bangladesh. Journal of Asia-Pacific Business,
2370.2009.00275.x
10(2), 130e145. https://doi.org/10.1080/10599230902885556
Cheung, Y. L., Tan, W., Ahn, H., & Zhang, Z. (2010). Does corporate social re-
Badulescu, A., Badulescu, D., Saveanu, T., & Hatos, R. (2018). The relationship be-
sponsibility matter in Asian emerging markets? Journal of Business Ethics, 92(3),
tween firm size and age, and its social responsibility actionsdfocus on a
401e413. https://doi.org/10.1007/s10551-009-0164-3
developing country (Romania). Sustainability, 10(3), 805. https://doi.org/
Chollet, P., & Sandwidi, B. W. (2018). CSR engagement and financial risk: A virtuous
10.3390/su10030805
circle? Int. Evid. Global Finance J., 38, 65e81. https://doi.org/10.1016/
Bangladesh Bank. (2015). We shall overcome: Review of CSR activities of Bangladesh
j.gfj.2018.03.004
Bank, Commercial Banks & financial institutions-2014. Dhaka: Bangladesh Bank.
Cochran, P. L., & Wood, R. A. (1984). Corporate social responsibility and financial
https://www.bb.org.bd/pub/annual/csr_activities/2014.pdf.
performance. Academy of Management Journal, 27(1), 42e56. https://doi.org/
Barnett, M. L. (2007). Stakeholder influence capacity and the variability of financial
10.2307/255956
returns to corporate social responsibility. Acad. Manag., 32(3), 794e816. https://
Cornett, M. M., Erhemjamts, O., & Tehranian, H. (2016). Greed or good deeds: An
doi.org/10.2307/20159336
examination of the relation between corporate social responsibility and the
Basuony, M. A., Elseidi, R. I., & Mohamed, E. K. (2014). The impact of corporate social
financial performance of U.S. commercial banks around the financial crisis.
responsibility on firm performance: Evidence form a MENA country. Corporate
Journal of Banking & Finance, 70, 137e159. https://doi.org/10.1016/
Ownership and Control, 12(1), 761e774. https://doi.org/10.22495/cocv12i1c9p1
j.jbankfin.2016.04.024
Belasri, S., Gomes, M., & Pijourlet, G. (2020). Corporate social responsibility and
D'Amato, A., & Falivena, C. (2019). Corporate social responsibility and firm value: Do
bank efficiency. Journal of Multinational Financial Management, 54, 100612.
firm size and age matter? Empirical evidence from European listed companies.
https://doi.org/10.1016/j.mulfin.2020.100612

nabou, R., & Tirole, J. (2010). Individual and corporate social responsibility. Corporate Social Responsibility and Environmental Management, 27(2), 909e924.
Be
https://doi.org/10.1002/csr.1855

10
M.A. Kabir and S.S. Chowdhury Asia Pacific Management Review xxx (xxxx) xxx

Dahlsrud, A. (2008). How corporate social responsibility is defined: An analysis of reporting: A study of selected banking companies in Bangladesh. Social Re-
37 definitions. Corporate Social Responsibility and Environmental Management, sponsibility Journal, 5(3), 344e357. https://doi.org/10.1108/17471110910977276
15(1), 1e13. https://doi.org/10.1002/csr.132 Khatun, M. (2014). Corporate social responsibility in Bangladesh: The law and
Davis, K. (1973). The case for and against business assumptions of social re- practices. Journal of Law, Policy and Globalization, 31, 10e17. https://www.iiste.
sponsibilities. Academy of Management Journal, 16(2), 312e322. https://doi.org/ org/Journals/index.php/JLPG/article/view/17485.
10.5465/255331 Kitzmueller, M., & Shimshack, J. (2012). Economic perspectives on corporate social
Fahad, P., & Nidheesh, K. B. (2020). Determinants of CSR disclosure: An evidence responsibility. Journal of Economic Literature, 50(1), 51e84. https://www.jstor.
from India. J.Indian. Bus. Res., 13(1), 110e133. https://doi.org/10.1108/jibr-06- org/stable/23269969.
2018-0171 Krasodomska, J. (2015). CSR disclosures in the banking industry. Empirical evidence
Friedman, M. (1970, September 13). The social responsibility of business is to increase from Poland. Social Responsibility Journal, 11(3), 406e423. https://doi.org/
its profits. New York Times Magazine. https://www.nytimes.com/1970/09/13/ 10.1108/srj-02-2013-0019
archives/a-friedman-doctrine-the-social-responsibility-of-business-is-to.html. Kurucz, E., Colbert, B., & Wheeler, D. (2008). The Business Case for Corporate Social
Galant, A., & Cadez, S. (2017). Corporate social responsibility and financial perfor- Responsibility. Oxford University Press. https://doi.org/10.1093/oxfordhb/
mance relationship: A review of measurement approaches. Econ. Res. Eko- 9780199211593.003.0004
nomska Istrazivanja, 30(1), 676e693. https://doi.org/10.1080/ Lin, W. L., Law, S. H., Ho, J. A., & Sambasivan, M. (2019). The causality direction of the
1331677x.2017.1313122 corporate social responsibility e corporate financial performance nexus:
García-Sa
nchez, I. M., & García-Meca, E. (2017). CSR engagement and earnings Application of panel vector Autoregression approach. The North American
quality in banks. The moderating role of institutional factors. Corporate Social Journal of Economics and Finance, 48, 401e418. https://doi.org/10.1016/
Responsibility and Environmental Management, 24(2), 145e158. https://doi.org/ j.najef.2019.03.004
10.1002/csr.1405 Lin, C. H., Yang, H. L., & Liou, D. Y. (2009). The impact of corporate social re-
Goering, G. E. (2012). Corporate social responsibility and marketing channel coor- sponsibility on financial performance: Evidence from business in Taiwan.
dination. Research in Economics, 66(2), 142e148. https://doi.org/10.1016/ Technology in Society, 31(1), 56e63. https://doi.org/10.1016/
j.rie.2011.10.001 j.techsoc.2008.10.004
Goss, A., & Roberts, G. S. (2011). The impact of corporate social responsibility on the Lougee, B. A., & Wallace, J. S. (2008). The corporate social responsibility (CSR) trend.
cost of bank loans. Journal of Banking & Finance, 35(7), 1794e1810. https:// The Journal of Applied Corporate Finance, 20(1), 96e108. https://doi.org/10.1111/
doi.org/10.1016/j.jbankfin.2010.12.002 j.1745-6622.2008.00172.x
Griffin, J. J., & Mahon, J. F. (1997). The corporate social performance and corporate Makni, R., Francoeur, C., & Bellavance, F. (2009). Causality between corporate social
financial performance debate: Twenty-five years of incomparable research. performance and financial performance: Evidence from Canadian firms. Journal
Business & Society, 36(1), 5e31. https://doi.org/10.1177/000765039703600102 of Business Ethics, 89(3), 409e422. https://doi.org/10.1007/s10551-008-0007-7
Han, H., & Kim, Y. (2020). The impact of firm age on corporate social responsibility. Maqbool, S., & Zameer, M. N. (2018). Corporate social responsibility and financial
EMAC 2020 Annual Conference. Proc. Eur. Market. Acad., 49, 62801 http:// performance: An empirical analysis of Indian banks. Future Business Journal,
proceedings.emac-online.org/index.cfm?abstractid¼A2020-62801&The% 4(1), 84e93. https://doi.org/10.1016/j.fbj.2017.12.002
20impact%20of%20firm%20age%20on%20corporate%20social%20respons. Martínez-Ferrero, J., & Frías-Aceituno, J. V. (2015). Relationship between sustainable
Haroon, J. (2017, December 2). SoCBs see profit drop or count loss. The Financial development and financial performance: International empirical research.
Express, 1512185508 https://www.thefinancialexpress.com.bd/trade/socbs-see- Business Strategy and the Environment, 24(1), 20e39. https://doi.org/10.1002/
profit-drop-or-count-loss. bse.1803
Hausman, J. A. (1978). Specification tests in econometrics. Econometrica: Journal of Matten, D., & Moon, J. (2008). “Implicit” and “explicit” CSR: A conceptual framework
the econometric society, 1251e1271. https://doi.org/10.2307/1913827 for a comparative understanding of corporate social responsibility. Academy of
Hemingway, C. A., & Maclagan, P. W. (2004). Managers' personal values as drivers of Management Review, 33(2), 404e424. https://doi.org/10.5465/
corporate social responsibility. Journal of Business Ethics, 50(1), 33e44. https:// amr.2008.31193458
doi.org/10.1023/b:busi.0000020964.80208.c9 McWilliams, A., & Siegel, D. (2000). Corporate social responsibility and financial
Hirigoyen, G., & Rehm, T. P. (2015). Relationships between corporate social re- performance: Correlation or misspecification? Strategic Management Journal,
sponsibility and financial performance: What is the causality? Journal of Busi- 21(5), 603e609. https://doi.org/10.1002/(sici)1097-0266 (200005)21:5<603::
ness Management, 4(1), 18e43. https://doi.org/10.12735/jbm.v4i1p18 aid-smj101>3.0.co;2-3.
Ho, A. Y., Liang, H., & Tumurbaatar, T. (2019). The impact of corporate social re- Mintzberg, H. (1983). The case for corporate social responsibility. Journal of Business
sponsibility on financial performance: Evidence from commercial banks in Strategy, 4(2), 3e15. https://doi.org/10.1108/eb039015
Mongolia. Advances in Pacific Basin Business, Economics and Finance, 109e153. Miras-Rodríguez, M., Martínez-Martínez, D., & Escobar-Pe
rez, B. (2019). Which
https://doi.org/10.1108/s2514-465020190000007006 corporate governance mechanisms drive CSR disclosure practices in emerging
Hu, Y., Chen, S., & Wang, J. (2018). Managerial humanistic attention and CSR: Do countries? Sustainability, 11(1), 61. https://doi.org/10.3390/su11010061
firm characteristics matter? Sustainability, 10(11), 4029. https://doi.org/10.3390/ Moskowitz, M. R. (1972). Choosing socially responsible stocks. Business and Society
su10114029 Review, 1, 71e75.
Hur, W., Kim, H., & Kim, H. K. (2018). Does customer engagement in corporate social Ndiweni, E., Haque, F., & Kamal Hassan, M. (2018). Corporate social responsibility
responsibility initiatives lead to customer citizenship behaviour? The mediating practices of banks in Bangladesh: A structuration theory perspective. Investment
roles of customer-company identification and affective commitment. Corporate Management and Financial Innovations, 15(1), 350e360. https://doi.org/
Social Responsibility and Environmental Management, 25(6), 1258e1269. https:// 10.21511/imfi.15(1).2018.29
doi.org/10.1002/csr.1636 Nelling, E., & Webb, E. (2009). Corporate social responsibility and financial perfor-
Ilhan-Nas, T., Koparan, E., & Okan, T. (2015). The effects of the CSR isomorphism on mance: The “virtuous circle” revisited. Review of Quantitative Finance and Ac-
both CSP and CFP. Journal of Asia Business Studies, 9(3), 251e272. https://doi.org/ counting, 32(2), 197e209. https://doi.org/10.1007/s11156-008-0090-y
10.1108/jabs-11-2014-0086 Nnaoma, F. C., & Omotosho, O. (2017). Impact of corporate social responsibility cost
Islam, M. Z., Ahmed, S., & Hasan, I. (2012). Corporate social responsibility and on investment policy and performance of firms in Nigeria. European Journal of
financial performance linkage: Evidence from the banking sector of Bangladesh. Business and Management, 9(4), 94e129. https://www.iiste.org/Journals/index.
Journal of Organizational Management, 1(1), 14e21. https://papers.ssrn.com/ php/EJBM/article/view/35317.
sol3/papers.cfm?abstract_id¼1999140. Orlitzky, M., Schmidt, F. L., & Rynes, S. L. (2003). Corporate social and financial
Jackson, P., & Hawker, B. (2001). Is corporate social responsibility here to stay? performance: A meta-analysis. Organization Studies, 24(3), 403e441. https://
Communication Directors’ Forum, 18e04, 2016. doi.org/10.1177/0170840603024003910
Jain, A., Keneley, M., & Thomson, D. (2014). Bank reputation in Australia: A view Porter, M. E., & Kramer, M. R. (2006, December 1). Strategy and society: The link
from the inside. Journal of Applied Science in Southern Africa: The Finsia Journal of between competitive advantage and corporate social responsibility. Harvard
Applied Finance, (2), 29e34. https://search.informit.org/doi/10.3316/INFORMIT. Business Review. https://hbr.org/2006/12/strategy-and-society-the-link-
674504743181447. between-competitive-advantage-and-corporate-social-responsibility.
Kabir, M. A., & Sunon, K. H. (2021). Financial data of banks listed in the Dhaka stock Prior, F., & Argandon ~ a, A. (2009). Credit accessibility and corporate social re-
Exchange (DSE). Mendeley data, V2. https://data.mendeley.com/datasets/ sponsibility in financial institutions: The case of microfinance. Business Ethics: A
vd4g9xjt4b/2. European Review, 18(4), 349e363. https://doi.org/10.1111/j.1467-
Kamens, D. H. (1985). A theory of corporate civic giving. Sociological Perspectives, 8608.2009.01568.x
28(1), 29e49. https://doi.org/10.2307/1389073 Ramasamy, B., Ting, H. W., & Yeung, M. C. (2007). Does it pay to be good in
Karim, R. A., & Kasim, A. (2018). Urgent need for CSR in Bangladesh with an developing countries? The relationship between CSR and financial performance
emphasis on environmental issues. In Knowledge Management International in Malaysia. Asian Academy of Management Journal of Accounting and Finance,
Conference (KMIC) 2018 (pp. 235e241). https://repo.uum.edu.my/id/eprint/ 3(1), 21e36. http://web.usm.my/journal/aamjaf/vol%203-1-2007/3-1-2.pdf.
25230. Roy, I., Sarker, A. K., & Chowdhury, S. (2017). Corporate social responsibility prac-
Kempf, A., & Osthoff, P. (2007). The effect of socially responsible investing on tices in Bangladesh: A statistical analysis on state-owned & private commercial
portfolio performance. European Financial Management, 13(5), 908e922. banks. Economics World, 5(4), 322e332. https://doi.org/10.17265/2328-7144/
https://doi.org/10.1111/j.1468-036x.2007.00402.x 2017.04.005
Khan, H. (2010). The effect of corporate governance elements on corporate social Saha, A. K., Dey, S. K., & Khan, A. (2013). Corporate social responsibility in
responsibility (CSR) reporting. International Journal of Law and Management, Bangladesh: A comparative study of commercial banks of Bangladesh. The
52(2), 82e109. https://doi.org/10.1108/17542431011029406 Bangladesh Accountant, 69e80.
Khan, H., Halabi, A. K., & Samy, M. (2009). Corporate social responsibility (CSR) Samuelson, P. A. (1971). Love that corporation. Mountain Bell Magazine, 24(1).

11
M.A. Kabir and S.S. Chowdhury Asia Pacific Management Review xxx (xxxx) xxx

Schnippering, M. (2020). R&D: The missing link between corporate social perfor- practices in banking companies in Bangladesh. Journal of Financial Reporting &
mance and financial performance? Management Review Quarterly, 70(2), Accounting, 13(2), 200e225. https://doi.org/10.1108/jfra-05-2013-0038
243e255. https://doi.org/10.1007/s11301-019-00166-5 van Marrewijk, M. (2003). Concepts and definitions of CSR and corporate sustain-
Seifert, B., Morris, S. A., & Bartkus, B. R. (2003). Comparing big givers and small ability: Between agency and Communion. Journal of Business Ethics, 44, 95e105.
givers: Financial correlates of corporate philanthropy. Journal of Business Ethics, https://doi.org/10.1023/A:1023331212247
45(3), 195e211. https://www.jstor.org/stable/25075066. Visser, W. (2009). Corporate social responsibility in developing countries. The Ox-
Senyigit, Y. B., & Shuaibu, M. A. (2017). The effect of corporate social responsibility ford Handbook of Corporate social Responsibility. https://doi.org/10.1093/
on financial performance: Evidence from the banking industry in emerging oxfordhb/9780199211593.003.0021
economies. Journal of Business Studies Quarterly, 9(1), 22e31. Vitezic, N. (2011). Corporate reputation and social responsibility: An analysis of
Sharif, M., & Rashid, K. (2014). Corporate governance and corporate social re- large companies in Croatia. International Business & Economics Research Journal,
sponsibility (CSR) reporting: An empirical evidence from commercial banks 10(8), 85e95. https://doi.org/10.19030/iber.v10i8.5380
(CB) of Pakistan. Quality and Quantity, 48(5), 2501e2521. https://doi.org/ Vogel, D. J. (2005). Is there a market for virtue?: The business case for corporate
10.1007/s11135-013-9903-8 social responsibility. California Management Review, 47(4), 19e45. https://
Sheldon, O. (1924). The philosophy of management. Pitman. doi.org/10.1177/000812560504700401
Smith, T. (2005). Institutional and social investors find common ground. Journal of Waddock, S. A., & Graves, S. B. (1997). The corporate social performance-financial
Investing, 14(3), 57e65. https://doi.org/10.3905/joi.2005.580550 performance link. Strategic Management Journal, 18(4), 303e319. https://
Sturdivant, F. D., & Ginter, J. L. (1977). Corporate social responsiveness management doi.org/10.1002/(sici)1097-0266 (199704)18:4<303::aid-smj869>3.0.co;2-g.
attitudes and economic performance. California Management Review, 19(3), Westerlund, J. (2005). New simple tests for panel cointegration. Econometric Re-
30e39. https://doi.org/10.2307/41164709 views, 24(3), 297e316. https://doi.org/10.1080/07474930500243019
Sun, L. (2012). Further evidence on the association between corporate social re- Windsor, D. (2001). The future of corporate social responsibility. International
sponsibility and financial performance. International Journal of Law and Man- Journal of Organizational Analysis, 9(3), 225e256. https://doi.org/10.1108/
agement, 54(6), 472e484. https://doi.org/10.1108/17542431211281954 eb028934
Suteja, J., Gunardi, A., & Mirawati, A. (2016). Moderating effects of earnng man- Withisuphakorn, P., & Jiraporn, P. (2015). The effect of firm maturity on corporate
agement on the relationship between corporate social responsibility disclosure social responsibility (CSR): Do older firms invest more in CSR? Applied Eco-
and profitability of banks in Indonesia. International Journal of Economics and nomics Letters, 23(4), 298e301. https://doi.org/10.1080/13504851.2015.1071464
Financial Issues, 6(4), 1360e1365. https://www.econjournals.com/index.php/ Wu, M., & Shen, C. (2013). Corporate social responsibility in the banking industry:
ijefi/article/view/2779. Motives and financial performance. Journal of Banking & Finance, 37(9),
Swaen, V., & Chumpitaz, R. C. (2008). Impact of corporate social responsibility on 3529e3547. https://doi.org/10.1016/j.jbankfin.2013.04.02
consumer trust. Recherche et Applications en Marketing, 23(4), 7e34. https:// Xie, S. (2015). Corporate social responsibility and firm performance: OLS and
doi.org/10.1177/205157070802300402 granger causality analysis. SSRN Electronic Journal. https://doi.org/10.2139/
Taskin, D. (2015). The relationship between CSR and banks' financial performance: ssrn.2598384
Evidence from Turkey. Journal of Yasar University, 10(39), 21e30. https:// Yahoo! Finance [online] Available at: https://finance.yahoo.com/quote/usdbdt¼X/,
dergipark.org.tr/en/download/article-file/179558. (2020)e. (Accessed 24 October 2020).
Testa, M., & D'Amato, A. (2017). Corporate environmental responsibility and Yesmine, S., & Bhuiyah, M. S. (2015). Corporate social responsibility practices in
financial performance: Does bidirectional causality work? Empirical evidence Bangladesh: Evidence from engineering and pharmaceutical & chemical in-
from the manufacturing industry. Social Responsibility Journal, 13(2), 221e234. dustry. International Journal of Business and Management Invention, 4(10), 41e51.
https://doi.org/10.1108/srj-02-2016-0031 https://www.ijbmi.org/papers/Vol(4)10/F0401041051.pdf.
Ullah, M. H., & Rahman, M. (2015). Corporate social responsibility reporting

12