SSRN Id4452395

Effect of CSR on the Overall and Components of Firm-Specific Earnings
(Running title: CSR and Firm-Specific Earnings)
JiangBo HuangFu
Berea College
Mark Kohlbeck *
Florida Atlantic University
Hanbing Xing
March 20, 2023
* Corresponding author
777 Glades Road
Boca Raton, FL 33431
561-297-1363
mkohlbec@fau.edu
We appreciate comments and suggestions on previous versions of our paper from workshop
participants at Florida Atlantic University, and participants at the 2021 American Accounting
Association Southwest Region Annual Meeting and the 2022 American Accounting Association
Southeast Region Annual Meeting.
Electronic copy available at: https://ssrn.com/abstract=4452395

ABSTRACT
We investigate whether firm corporate social responsibility (CSR) performance is associated
with firm-specific earnings or its components. Resource-based view suggests firms use their
unique resources, such as CSR performance, to develop competitive advantages. Competitive
advantages are then reflected in earnings that are specific to the firm and not the industry overall.
We therefore predict and find a positive association between CSR and firm-specific earnings.
When we disaggregate firm-specific earnings into its components, CSR is positively associated
with both the level of firm-specific cash flows and normal accruals, but is negatively associated
with discretionary accruals. We also examine the different CSR components and find that our
results are driven by CSR related to the environment and product characteristics and industry
concerns. Collectively, our findings contribute to the debate on CSR and firm performance. We
also provide insights for managers and investors about components of firm earnings related to
CSR activities and drivers of effective CSR applications.
Keywords: CSR performance, firm-specific accruals, firm-specific cash flows, firm-specific
earnings
Data Availability: Data are available from the public sources identified in the text.

INTRODUCTION
Global sustainable investment assets increased to $30.7 trillion at the beginning of 2018,
a 34% increase from 2016 (US SIF 2019). Increasing investments imply positive financial
returns, and studies confirm this implication by documenting that corporate social responsibility
(CSR) is positively associated with both firm financial performance and market valuation
(Flammer, 2015; Lins, et al. 2017). Therefore, CSR benefits not only social welfare but also
firms’ market value, and incentivizes firms to engage in CSR efforts and compete on CSR
performance.
However, CSR’s emphasis on social welfare deviates from standard economic theory
(Friedman, 1970; Benabou and Tirole, 2010; Kitzmueller and Shimshack, 2012). The main
argument from this perspective is that a corporation’s responsibility is to generate profit and
wealth for shareholders. Commitments to CSR imply additional costs incurred for social
purposes that are irrelevant to and even reduce shareholder wealth maximization, and hence
conflict with a firm’s shareholder orientated mission. Consistent with this argument, some
studies document a negative relationship between CSR engagement and firm value (Margolis, et
al. 2012; Di Giuli and Kostovetsky, 2014).
As a result, there is an ongoing debate on whether CSR enhances firm value or represents
an additional cost. One major contention is whether CSR or other characteristics and factors
determine financial performance. For example, profitable firms have a higher propensity to
engage CSR, and as a result, these firms’ relatively higher level of profitability rather than the
level of a firm’s CSR performance drives the seemingly positive relationship between CSR and
firm financial performance (Hong, et al. 2012). Critics argue that big and successful firms are

more likely to engage in CSR (Clarkson, et al. 2011), and that there is a lack of direct evidence
of the effect of CSR on firm success. At the heart of this debate is whether a firm’s success is
due to CSR or whether CSR is what a successful firm does (Huang and Watson, 2015).
Ultimately, whether, and if so how, CSR influences a firm’s financial performance is an open
empirical question.
We focus on firm-specific earnings to provide evidence for this question. Firm-specific
earnings refer to the portion of a firm’s earnings performance that is from firm-unique factors.
As an idiosyncratic component of firm-level earnings, firm-specific earnings are linked to a
firm’s unique capabilities and contribute to its market value (Brown and Kimbrough, 2011). One
such unique factor is CSR performance. Superior CSR performance enhances a firm’s reputation
(Du, et al. 2011; Lattanzio and Litov, 2020) and establishes a unique competitive edge
(Loikkanen and Hyytinen, 2011). Therefore, superior CSR performance enables a firm to
develop exclusive CSR resources and capabilities compared to its peers.
According to the resource-based view (RBV) theory (Barney, 1991; Amit and
Schoemaker, 1993), firms identify and develop their unique resources to differentiate themselves
from others and improve their competitiveness. Competing on CSR motivates firms to create and
increase their unique capabilities, reputation, and value differentiation that give them a
competitive advantage. CSR performance is therefore expected to be positively associated with
firm-specific earnings.
We first measure CSR performance using MSCI KLD ratings (Servaes and Tamayo,
2013; Cui, et al. 2018; Burke, et al. 2020). We develop three measures that include both the
conventional and a broader category of CSR activities as well as an industry-adjusted CSR
measure that addresses potential selection bias from high or low concentrated CSR industries.

We then follow Hui, et al. (2016) to measure firm-specific earnings as the difference between a
firm’s disclosed earnings and industry-wide earnings.
We regress firm-specific earnings on CSR performance and find that firms with higher
overall CSR performance are positively associated with firm-specific earnings. We then
disaggregate firm-specific earnings into firm-specific cash flows, normal accruals, and
discretionary accruals to identify relevant components of firm-specific earnings related to CSR
performance. We find a positive association between CSR performance and firm-specific cash
flows and normal accruals; however, CSR performance is negative associated with firm-specific
discretionary accruals. Our findings are consistent with RBV and support our hypotheses.
We perform a number of tests to better understand our results. We find that our results
are driven by CSR activities classified as strengths. We also investigate individual CSR
components. We find that a firm’s CSR performance on the environment, product quality and
safety, and industries with less controversial business involvement are significantly positively
associated with firm-specific earnings. These are consistent with environment and product
(controversial industry involvement) related CSR activities being value-enhancing (value-
destructive).
Our study contributes to the literature on firm-specific earnings and extends the CSR
literature by investigating whether CSR competition and performance are associated with firm-
specific earnings and thus increase firms’ economic differentiation. As firm-specific earnings
refer to the portion of earnings due to firm-specific factors instead of industry-wide and market-
level factors, the positive association between superior CSR performance and firm-specific
earnings implies CSR is an important driver of a firm differentiation strategy that creates
competitive and sustainable advantages over industry peer firms. We link CSR with firm-specific

factors that contribute to earnings and address the selection bias (only profitable firms engage
CSR) issue in CSR research. Our study therefore responds to Huang and Watson’s (2015) call
for research to examine the CSR - earnings relationship and addresses the reverse causality issue
and provides evidence consistent with CSR being value-enhancing.
The results are useful to investors in understanding the effect of CSR performance on
firm-specific earnings and could help investors better decide on investing in CSR competitive
firms. Academia and society in general debate whether business performance benefits from CSR.
Our study joins this debate and provides evidence that certain CSR activities, especially those
related to environment and product, are positively associated with firm-specific earnings and its
components. Investors may therefore look not only to general CSR performance but also the
specific areas of CSR. The study results are also useful to managers by providing them guidance
when making decisions and determining strategies to engage in CSR competition, especially in
value-enhancing areas of CSR activities.
LITERATURE REVIEW AND HYPOTHESES DEVELOPMENT
Related Literature
CSR is defined as a firm’s voluntary efforts and commitments that surpass compliance
with law and regulations to further social good (McWilliams, et al. 2006). Both external
pressures and internal managerial motivations create the widespread popularity of CSR.
Shareholders and other external forces influence and could even pressure firms to
undertake CSR activities. Rodrigue, et al. (2013) use a case approach and find that shareholders
influence managers’ CSR decisions, strategies, and even performance. Pondeville, et al. (2013)
find that not only shareholders but also the market and community can significantly influence a
firm’s CSR control systems. Contrafatto (2014) identifies that public opinions and expectations

have a positive effect on the influential role of ordinary employees in institutionalizing CSR
within the firm.
Internal managerial motivations originate from managers’ focus on efficiency (Burnett
and Hansen, 2008; Burnett, et al. 2011; Figge and Hahn, 2013). Efficiency captures efficient use
of all resources, including environmental resources, for value creation. Managers engage in CSR
to create sustainable values from an eco-efficiency aspect. Managers use CSR to leverage firm
competitiveness and firm financial performance (Kemper, et al. 2013) or directly engage in CSR
for competitive advantage (Loikkanen and Hyytinen, 2011). Therefore, managers are motivated
to engage in CSR as long as CSR is tied with firm efficiency and competitiveness, both of which
are positively associated with firm financial performance.
The growing number of CSR-related investments implies that CSR is believed to be
associated with firm financial performance, stimulating increasing amounts of sustainable
investments. While research on CSR and firm financial performance is abundant, the results are
mixed.
Margolis, et al. (2012) use meta-analysis to review 251 academic works that examine the
relation between CSR and firm financial performance. They conclude a positive relation between
CSR and firm financial performance exists. Dowell, et al. (2000) find that firms who adopt
stringent global environmental standards, guiding firms to improve CSR practices, have much
higher afterward market returns than those who default or neglect the standards. CSR is found to
positively affect firm financial performance because CSR reduces the cost of equity capital (EI
Ghoul, et al. 2011), significantly lowers capital constraints (Cheng, et al. 2014), mitigates agency
costs and information asymmetry promotes product differentiation resulting in lower systematic
risk (Albuquerque, et al. 2019), and increases the likelihood of softer penalties in litigation

(Hong, et al. 2012).
Not all studies support that increasing sustainable investments are associated with
improved financial performance. The fundamental argument for CSR’s negative association is
rooted in classic economic theory. For example, Modigliani and Miller’s (1958) theory suggests
that CSR performance is irrelevant to firm value. CSR also conflicts with Friedman’s (1970)
neoclassical economic theory that suggests that corporations, different from individuals, have no
social conscience because, in the free capital economy, a firm’s primary responsibility is to the
owners (shareholders). CSR is voluntary and based on individual freedom; hence it is not the
firm’s responsibility.
Consistent with these theoretical arguments, some empirical studies document a
negative relationship between CSR and firm financial performance. Lys, et al. (2015) find firms
use CSR as a signaling apparatus and document evidence that firms increase current period CSR
spending when they anticipate promising future financial performance. CSR has no significant
association with firm-level social performance nor firm future financial performance directly,
and the study, therefore, highlights the signaling effect of CSR reporting.
Bhandari and Javakhadze (2017) find a negative association between CSR strategies and
firm-level resource allocation efficiency. They attribute their results to CSR requiring resources
that otherwise could be utilized for more efficient investments. CSR, in this case, is considered a
deviation from normal business operations in capital allocation efficiency and hence mitigates
firm financial performance. CSR, however, does not necessarily conflict with normal business
operations for maximizing shareholder values. CSR positively influences a firm’s reputation,
promotes trust, and gains shareholders as well as stakeholders’ acceptance and support, all of
which eventually lead to improved financial performance (Porter and Kramer, 2011; Clarkson, et

al. 2020; Lattanzio and Litov, 2020).
The challenge in studying the CSR - firm financial performance relationship is to identify
the costs and benefits of CSR (Huang and Watson, 2015). Without a clear identification and
inclusion of CSR’s costs and benefits, the estimation of the CSR - firm financial performance
relationship suffers a high likelihood of a spurious conclusion. One issue is reverse causality
where CSR performance is an outcome of firm financial performance because successful and
profitable firms tend to engage in CSR and score on CSR performance (Hong, et al. 2012; Lys, et
al. 2015). Flammer (2015) addresses this issue by studying CSR shareholder proposals that pass
or are defeated by small margins to create a quasi-experimental setting. Flammer finds that
adoption of CSR shareholder proposals is associated with superior firm financial performance
compared to those that are defeated. Bose and Yu (2023) take a different approach in their study
of CSR performance and earnings quality. Using Granger causality methods, they find that
changes in earnings quality lead to CSR, but not the opposite direction.
Huang and Watson (2015) call for future accounting research to further investigate the
mixed results and specifically address the endogeneity problem. They suggest that ‘studying
earnings is a promising avenue by which to begin to answer the CSR-financial performance
question because it avoids using market perceptions as performance measures as in studies that
use firm value or stock returns’ (Huang and Watson, 2015, p. 8).
Hypotheses Development
Resource-based view (RBV) theory suggests that firms maintain and develop core
resources and capabilities to achieve and sustain competitive advantages (Lippman and Rumelt,
1982; Hall, 1993). Specifically, RBV articulates that a firm’s core resources and capabilities are
drivers of firm differentiation strategies that help the firm gain competitive edges over rival

firms. The view of CSR performance as a source of competitive differentiation is consistent with
RBV theory for the following reasons.
CSR is an integral part of firm-level product differentiation strategies. For example,
Harley-Davidson introduced an all-electric motorcycle, LiveWire, in 2019. Harley-Davidson
even highlights this all-electric bike as a new prototype for its ‘Grow Business Without Growing
Environmental Impact’ business model in its 2019 annual sustainability report (Harley-
Davidson, 2019). Some consumers who prefer the electric bike to a conventional model are
willing to pay a premium because of the environmental-friendly features. These customers view
CSR as a signal of responsibility and honesty (Siegel and Vitaliano, 2007). CSR, therefore, can
be part of a firm’s differentiation strategy.
The accounting literature highlights that a firm’s internal resources and its unique
capabilities are sources of firm differentiation strategies. For example, Palepu, et al. (2007) point
out that internal resources and unique capabilities related to brand image, R&D, innovation, and
control systems are essential to a firm’s differentiation strategy to distinguish the firm from its
rivals in the industry and at the same time bring value to its customers.
Prior CSR literature provides evidence that CSR can repair and promote a firm’s
reputation and brand image (Chakravarthy, et al. 2013); hence it also creates a reputation
differentiation (Lattanzio and Litov, 2020). Firm CSR activities and performance can positively
influence shareholders and their willingness to buy from, invest in, or work for the firm
(Lattanzio and Litov, 2020).
Managers develop eco-control systems to manage CSR (Henri and Journeault, 2010). As
the eco-control system is integrated into existing management control systems, the newly
established control systems help the firm improve both environmental and financial performance.

Therefore, the unique blended control systems also differentiate firms from their peers.
CSR aims for a smaller impact on the natural environment and society, and is closely
related to firm efficiency. It requires innovation to improve efficiency strategically and to reduce
costs while increasing business growth. Bocquet, et al. (2017) classify CSR into strategic
(proactive efforts) and responsive (passive adoption) when estimating the CSR - innovation
relationship. They find that strategic CSR has a positive effect on both firm-level product and
process innovations.
In sum, CSR is a competitive tool that is part of product differentiation strategies,
representing a firm’s internal resources or unique capabilities that significantly influence firm
reputation differentiation, innovation, and control systems. A firm with superior CSR
performance can therefore differentiate itself. RBV theory posits that competition in the CSR
space should result in greater firm earnings that are not explained by industry factors, which we
refer to as firm-specific earnings. 1 Therefore, we expect a positive association between firm CSR
performance and firm-specific earnings.
We state the following hypothesis:
H1: Controlling for other factors, CSR performance is positively associated with firm-
specific earnings.
Earnings are composed of cash flows and accruals. While both cash flows and accruals
may explain firm performance, Desai, et al. (2004) find that the coefficient on cash flows
significantly subsumes that of accruals in predicting future returns. Since cash flows and
earnings generally move together, it is reasonable to expect that CSR performance is more likely
to be positively associated with cash flows than accruals.
1 Signaling theory provides an alternative argument that is also consistent with a positive association.

Firm-specific cash flows reflect a unique portion of cash flows associated with firm-
specific characteristics rather than industry-wide and market. RBV suggests that CSR firms will
utilize these unique competencies to create potential external opportunities and competitive
advantages for higher product and/or service values. Highly rated CSR firms also have more
opportunities and competitive advantages to create brand and product differentiation (Siegel and
Vitaliano, 2007; Chakravarthy, et al. 2013; Lattanzio and Litov, 2020) in which CSR firms could
establish their market separately from their peers enabling more repeat business from their loyal
customers. Also, higher product and/or service values are likely to increase customer
evaluations, which could translate into higher customer loyalty and more volume and repeat
purchases (Dodds, et al. 1991). Higher customer loyalty and more repeat business provide more
stable cash flows. Therefore, CSR is more likely to be positively associated with firm-specific
cash flows.
In summary, firms with superior CSR performance are more likely to differentiate
themselves from their industry peers, and because of this unique performance; these firms are
then more likely to have increased firm-specific cash flows. 2 This leads to our second
hypothesis.
H2: Controlling for other factors, CSR performance is positively associated with firm-
specific cash flows.
2 CSR firms are less likely to engage in opportunistic earnings manipulations (Kim, et al. 2012), and hence have higher quality
reporting. With higher quality reporting, CSR firms establish a trustworthy reputation that would enhance capital flow. For
example, CSR firms incur a lower cost of debt (Cooper and Uzun, 2015), and suppliers are more likely to extend payment or
offer more favorable terms for firms with a trustworthy reputation (Xu, et al. 2020). All of these also suggest CSR firms are more
likely to experience speedier cash inflows and favorable payment terms that not only cut costs but also possibly extend cash
outflows.
10

Accruals are the other component of earnings. Previous research on the relationship
between accruals and firm performance is mixed (e.g., Sloan, 1996; Hribar and Yehuda, 2015). It
is, therefore, difficult to predict the presence or direction of any association between CSR
performance and firm-specific accruals. Instead, we separately consider normal and abnormal
accruals. 3
Based on the RBV and our arguments for our first two hypotheses, CSR is assumed to be
associated with increased firm-specific earnings and cash flows. Normal accruals are the accruals
necessary to reflect the current level of operations. Since firm-specific cash flows change
persistently with changes in operations and CSR, firm specific normal accruals should also
change as a result. This leads to our expectation that CSR performance is positively associated
with normal accruals.
Some studies show a negative relationship between CSR and abnormal accruals (e.g.,
Chih, et al. 2008; Hong and Andersen, 2011; Kim et al. 2012), while other studies find a positive
association (Petrovits, 2006; Prior, et al. 2008). The mixed findings reflect the controversial
CSR-agency cost relationship. On the one hand, CSR is associated with reporting transparency to
avoid abnormal accruals; on the other hand, CSR could be for managers’ self-interests, leading to
increased abnormal accruals.
We argue that CSR firms are less likely to generate abnormal accruals for three reasons.
First, a high level of abnormal accrual is negatively associated with firm reputation (Kaplan and
Ravensroft, 2004); CSR firms care about reputation (Chakravarthy et al. 2013), and thus are less
likely to generate abnormal accruals. Second, CSR decreases systematic risks (Albuquerque et
al. 2019); however, abnormal earnings eventually increase systematic risks. Thus, CSR firms are
3 Total accruals can be separated into normal and abnormal accruals.
11

less likely to increase abnormal earnings. Third, there is also strong evidence that CSR mitigates
information asymmetry (Cui et al. 2018), and abnormal earnings increase information
asymmetry; thus, CSR firms may not have the incentive to generate more abnormal earnings.
Combined, we expect that CSR performance is associated with lower abnormal accruals.
Similar to our first two hypotheses, we focus on firm-specific accruals. Both firm-specific
normal and abnormal accruals reflect the unique portion of accruals associated with firm-specific
characteristics rather than industry-wide and market as a whole. We expect the unique
competencies from higher CSR performance to be associated with firm-specific normal and
abnormal accruals. Our discussion leads to the following hypotheses:
H3a: Controlling for other factors, CSR performance is positively associated with firm-
specific normal accruals.
H3b: Controlling for other factors, CSR performance is negatively associated with firm-
specific abnormal accruals.
Our expectation that CSR affects normal and abnormal accruals in opposite directions may help
explain why prior research is mixed on the CSR and accruals association.
RESEARCH DESIGN
Measurement of CSR Performance
We use the MSCI KLD dataset for our CSR data. 4 Although firms have increased the
extent of voluntary CSR disclosures, MSCI KLD is still widely used, especially for investigating
4 The MSCI KLD database evolves with time. It initially covers firms in the S&P 500 and the Domini 400 Social Index, and then
it expands its coverage of firms in Russell 1000 and Russell 3000. It started to include the human rights category in 2002. To
capture firm CSR performance given the above-mentioned changes in the database, we follow prior literature to measure a firm’s
CSR performance (Servaes and Tamayo, 2013; Cui, et al. 2018; Burke et al. 2020).
12

CSR - financial performance relationship (Khan, et al. 2016). The dataset is culled from a wide
range of sources, including surveys, global media publications, corporate disclosures and filings,
and academic journals. MSCI KLD evaluates a firm’s CSR based on eight categories:
community, diversity, employees, the environment, human rights, product quality and safety,
corporate governance, and controversial business involvement. Individual CSR attributes are then
classified as strengths and concerns. MSCI KLD rates the first seven categories with both
strengths and concerns but only rates controversial business involvement with concerns. The
controversial business involvement category includes alcohol, gaming, firearms, military,
nuclear, and tobacco activities, all of which are not typical CSR practices.
Our first measure is a narrow CSR performance score that captures the firm’s overall
CSR performance across community, diversity, employee relations, environment, human rights,
and product quality and safety categories used by MSCI / KLD. 5 We first compute CSR strength
and concern ratios for each social rating category, where the ratio is the number of strengths
(concerns) divided by the maximum possible number of strengths (concerns) in each category for
that year. Then we deduct the concerns ratio from the strengths ratio to obtain a measure of net
CSR engagement in each social rating category. Finally, we sum the scores of net CSR
engagement over the six categories to obtain CSR_NARROW, which ranges from -6 to +6 for
each firm-year observation.
Our second CSR measure is a broad CSR performance score and captures an overall CSR
performance in seven rating categories (the above six plus controversial business involvement
5 Compared with the other rating categories that focus more broadly on social objectives, corporate governance is about
evaluating monitoring mechanism placed on agents (the management) as an assurance for principals’ (shareholders) investments;
therefore, we exclude corporate governance from our CSR measure.
13

concerns). We compute the concern ratio for controversial business involvement as the sum of
all concern scores on controversial business involvement divided by the maximum possible
scores in this category. We then subtract this ratio from the narrow CSR performance score to
obtain CSR_BROAD, which ranges from -7 to +6 for each firm-year observation.
CSR activities may be industry-specific, meaning that CSR could be more concentrated
in certain industries more than others, and could create a selection bias issue. To address this
issue, we construct an industry-adjusted CSR measure. Our third measure (CSR_IDX) captures
overall CSR performance adjusted for industry variations. CSR_IDX is measured as the ratio of
1) CSR_NARROW less the minimum industry CSR_NARROW and 2) maximum industry
CSR_NARROW less the minimum industry CSR_NARROW.
Measurement of Firm-specific Earnings, Cash Flows, and Accruals
Following Hui and Yeung (2013) and Hui et al. (2016), we partition earnings into
industry-wide and firm-specific components. We use Equation (1) to compute firm-specific
earnings (firm and year subscripts are omitted for brevity unless necessary to understand this and
later equations).
FirmE = EARN - ∑ (EARNj *1/Nj) (1)
where FirmE is firm-specific earnings; EARN denotes firm earnings measured as operating
income after depreciation divided by average assets; j refers to industry using two-digit SIC
codes, and N refers to the number of firms in industry j.
In the same way, we partition operating cash flows into its industry-wide and firm-
specific components, using Equation (2).
FirmCF = OCF - ∑ (OCFj *1/Nj) (2)
where FirmCF is firm-specific cash flow; OCF denotes firm cash flow from operations divided
14

by average assets; j refers to industry using two-digit SIC codes, and N refers to the number of
firms in industry j.
Assuming earnings are composed of operating cash flows and accruals, we rearrange this
identity to show that accruals are the difference between earnings and cash flows. We use
Equation (3) to compute firm-specific accruals as follows.
FirmAcc = FirmE - FirmCF (3)
where FirmAcc is firm-specific accruals and other variables are as previously defined.
We further separate firm-specific accruals into its non-discretionary and discretionary
components. We estimate a performance-adjusted discretionary accrual model (Kothari, et al.
2005) by industry-year to compute firm-specific normal accruals (NORMAcc). The residual
from these estimations provides our estimates of the firm-specific abnormal accruals (DISAcc).
Empirical Specification
We modify the model from Brown and Kimbrough (2011) and include our variable of
interest to capture CSR performance. Our model to examine the association between firm-
specific earnings (or the earnings components) and CSR performance is as follows:
FirmE = β0 + β1 CSR_METRIC + β2 SIZE+ β3 MB + β4 MktShare+
β5 StdROA + β6 DIVERSE + β7 HERF + β8 LEVERAGE + β9 NINDF +
Year Fixed Effects + Firm Fixed Effects + ɛ (4)
where variables are defined in the Appendix. We include year and firm fixed effects to control
15

for time variant and firm fixed effects. 6 We cluster standard errors by both firm and year level. 7
The dependent variable is firm-specific earnings, FirmE, to test H1. FirmE is replaced by
firm-specific cash flows (FirmCF) to test H2. We also use NORMAcc and DISAcc as the
dependent variable to test H3a and H3b, respectively. CSR_METRIC takes the value of
CSR_NARROW, CSR_BROAD, or CSR_IDX. H1 is supported if the estimated coefficient on
the CSR variable is positive and significant when the dependent variable is firm-specific
earnings (β1 > 0). If the dependent variable is firm-specific cash flow, H2 is supported when the
estimated coefficient on the CSR variables is positive and significant (β1 > 0). Finally, for H3a
(H3b), we expect a positive (negative) coefficient on the CSR variable when the dependent
variable is NORMAcc (DISAcc).
We include several control variables that are potentially associated with firm-specific
earnings or their components. 8 Based on Durnev,et al. (2004) and Piotroski and Roulstone
(2004), we control for firm size and market share. Large firms and firms with market power may
operate differently, implying that the profitability of these firms might be independent of the
effect of industry. We expect the coefficient on firm size (SIZE) to be positive. However, the
direction of the influence of having larger market shares for firm-specific earnings is less clear.
Greater firm-specific earnings could result from the dominance of larger market shares or from
niche market players. Therefore, we leave our prediction for market share (MktShare) unsigned.
6 We use firm fixed effects instead of industry fixed effects because firm specific earnings are already adjusted for industry
factors.
7 Standard errors of firm specific earnings could be correlated across time within a firm, but are not likely to be correlated
industry effects. So, we cluster standard errors by firm and year level.
8 Consistent with prior research, our expectations for the control variables when the dependent variable is firm-specific cash
flows or accruals are the same as firm-specific earnings as discussed in the remainder of this section.
16

The market-to-book value of common equity ratio is included to capture growth potential where
we expect higher growth (MB) to be associated with greater firm-specific earnings.
The standard deviation of ROA captures the volatility in a firm’s earnings. A high
earnings volatility is not likely to be correlated with long-term industry and/or market trends
(Piotroski and Roulstone, 2004), but a high level of firm-level earnings volatility over five-years
could signal risk (Graham, et al. 2005) as a result of market forces / shocks. Also, a high level of
earnings volatility could introduce an increasing level of earnings smoothing (Graham et al.
2005; Khurana, et al. 2018), which would decrease firm-specific earnings. Firms are likely to
manage the negative impact of earnings volatility risks by minimizing the impact on firm-
specific earnings. Therefore, we predict a positive sign of StdROA.
Firms with more diverse operations could be less sensitive to market and industry-wide
factors, so the sign of DIVERSE is expected to be positive. As the earnings fundamentals of
firms in a highly concentrated industry are strongly correlated (Morck, et al. 2000), we predict a
negative sign on our measure of industry-level competition (HERF). Studies also document a
positive association between leverage and earnings volatility (Long and Malitz, 1985; Hall,
2002), given the above discussion of earnings volatility and firm-specific earnings (Piotroski and
Roulstone, 2004), we would expect a positive association. However, profitability and cash flows
could decline as leverage increases. Therefore, the association of LEVERAGE with firms’
specific-earnings is uncertain.
When more firms are in an industry, these firms could be homogenous resulting in less
firm-specific earnings (Durnev et al. 2004). However, more firms in one industry could imply
more competition. To survive and to grow in a competitive market, firms push to differentiate
their products and/or service from each other. Differentiated products and/or services in an
17

industry could lead to more firm-specific earnings. Increased competition when there are more
firms in an industry may also increase motivation to innovate to meet diverse yet changing
demands. This will result in diverse and changing features of products and services that different
firms can offer. Therefore, we predict a positive sign of NIND.
EMPIRICAL RESULTS
Sample Selection
We construct our sample starting with an initial sample from MSCI KLD of 55,551 firm-
year observations over the period from 1991 to 2018. 9 We retrieve firm earnings data and
financial controls from the North American Compustat database. We eliminate observations
missing either CSR performance ratings or financial data for the empirical tests. Our final sample
is 27,070 firm-year observations.
Table 1 reports our sample distribution by year. The difference between the number of
firms in the 1990s and later periods reflects the change in MSCI KLD’s coverage over time.
During the 1990s, the database primarily includes S&P 500 firms and those in the Domini 400
Social Index. Starting in 2001, MSCI KLD expanded coverage to the Russell 1000 firms, and
after 2003, expanded it further to the Russell 3000.
[INSERT TABLE 1 HERE]
Descriptive Statistics
Table 2 presents the descriptive statistics for the variables included in the primary
empirical model. We winsorize all continuous variables at the 1st and 99th percentiles to
minimize the influence of outliers. The means (medians) of FirmE, FirmCF, NORMAcc, and
9 Our sample ends in 2018 to avoid confounding effects from the Pandemic.
18

DISAcc (our dependent variables) are 0.694 (0.385), 0.287 (0.216), 0.398 (0.139), and 0.015
(0.018) respectively. The means (medians) for the dependent variables are comparable with prior
studies (e.g. Hui et al. 2016). 10
Our first two CSR performance metrics, CSR_NARROW and CSR_BROAD, have
means of 0.022 and 0.004, respectively, and medians near zero. The third measure, CSR_IDX, is
an industry-adjusted CSR performance indicator and has a mean (0.412) that is roughly equal to
its median (0.382). Since this proxy measures CSR performance as the difference between a firm
and the lowest-ranked CSR firm in an industry, the value is almost always greater than zero, so
the distribution of CSR_IDX differs from those of the other two CSR proxies. Our CSR
measures are in line with prior studies (e.g. Cui et al. 2018; Burke et al. 2020). 11 The values of
the control variables are also consistent with those reported in prior literature (e.g. Brown and
Kimbrough, 2011).
Table 3 reports Pearson pairwise correlation coefficients between all variables in
Equation (4). As expected, there are strong correlations between the firm-specific earnings
variables and between the CSR performance variables. However, this doesn’t affect our
estimations as only one specific-earnings variable and one CSR variable are included in the
model at a time. Consistent with our hypotheses, both CSR_NARROW and CSR_BROAD are
positively correlated with all firm-specific earnings variables except DISCAcc. However, the
correlations between CSR_IDX and the two firm-specific earnings proxies are negative. This
10 Descriptive statistics of the dependent variables before merging with MSCI KLD data are consistent with prior work. Since
MSCI KLD consists of primarily larger firms, our final sample consists of larger firms.
11 CSR_BROAD is less than that of CSR_NARROW because CSR_BROAD is reduced for controversial business involvement.
19

could be due to the unique distribution of the industry-adjusted CSR_IDX. Also, firm-specific
earnings exclude industry effects, whereas CSR_IDX are industry adjusted, which could reverse
correlations between CSR_IDX and firm-specific earnings variables.
There are several significant correlation coefficients among the control variables above
0.50. The correlation between SIZE and MktShare is expected as larger firms have greater
market shares. There are also a number of higher correlations among MktShare, NIND, and
HERF. Since all of these are industry-based measures, the higher correlations are expected. We
also conduct multicollinearity tests and find that the variance inflation factors for the
independent variables are well below 5. Combined, we do not expect multicollinearity to be a
problem.
Multivariate Regression Results
Table 4 reports the results of estimating Equation (4) where the dependent variable is
firm-specific earnings and the definition of CSR is varied across the three columns. 12 The
adjusted R2 for each estimation is approximately 35%. The estimated coefficients for control
variables are in general, as predicted. The estimated coefficients for the CSR metrics in each
model are positive and significant (0.043, p-value < 0.01; 0.044, p-value < 0.01; and 0.154, p-
value < 0.10, respectively). Whether we measure CSR over the six categories, include the
influence of controversial businesses, or industry adjust a measure of CSR performance, we find
a positive association. Our first hypothesis is supported.
In Table 5, we consider the association of CSR with firm-specific cash flows using
12 Robust t-statistics for all estimations are based on correction for heteroskedasticity with clustering at the firm and year level.
20

Equation (4). The explanatory power of the three models are approximately 56%. The significant
control variables are also consistent with our expectations. The estimated coefficients on the
CSR metrics are again positive and significant in all three specifications (0.013 in model (1), p-
value < 0.01; 0.014 in model (2), p-value < 0.01; and 0.033 in model (3), p-value < 0.10). These
results show that firms with higher CSR performance have a higher level of firm-specific cash
flows and are consistent with the argument that firms gain a competitive advantage through their
unique CSR resources and report increased cash flow when the CSR performance is higher. This
evidence also supports RBV that firms with unique resources are more likely to have a
competitive advantage in terms of cash inflows among their rival firms. H2 is supported.
Before we provide evidence on our final hypotheses, we estimate Equation (4) where the
dependent variable is firm-specific accruals (untabulated). The estimated coefficients on the CSR
metrics are positive and significant in all three specifications (0.032 in model (1), p-value < 0.05;
0.032 in model (2), p-value < 0.05; and 0.132 in model (3), p-value < 0.10). We find consistent
evidence that higher CSR performance is associated with a higher level of firm-specific accruals.
We explore this in our tests of our final hypotheses.
Next, we estimate Equation (4) with firm-specific normal accruals as the dependent
variable (Table 6, Panel A). The explanatory power for each model is approximately 45% and
the significant coefficients on the control variables are consistent with our predictions. The
estimated coefficients on the CSR metrics are positive and significant supporting H3a. The
estimated coefficients on the CSR metrics are 0.021 in model (1), p-value < 0.01; 0.021 in model
(2), p-value < 0.01; and 0.067 in model (3), p-value < 0.01).
When firm-specific discretionary accruals is the dependent variable, the estimated
21

coefficients on the CSR variables are negative and significant (Table 6, Panel B). The estimated
coefficients on the CSR metrics are -0.003 in model (1), p-value < 0.01; -0.003 in model (2), p-
value < 0.01; and -0.005 in model (3), p-value < 0.10). The explanatory power for each model is
approximately 28% and the significant coefficients on the control variables are consistent with
our predictions. H3b is supported.
Overall, our accrual results imply a positive CSR and firm-specific earnings relationship,
and this positive relationship is driven by firm-specific cash flow and an outweighed CSR-
normal accruals relationship over the CSR-abnormal accruals relationship. The consistent
findings on the effect of CSR performance suggest that CSR benefits exceed its costs in terms of
increasing not just firm-specific earnings but high-quality earnings. 13
Difference-In-Differences Test
Firm-specific earnings can be determined by not only CSR performance but by
unobservable factors that affect CSR performance. This omitted variable bias presents an
endogeneity concern that we address by implementing a difference-in-difference design using
the first or initial MSCI KLD coverage as an exogenous shock (Cheng, et al. 2014).
The initiation of the MSCI KLD’s rating is unlikely to be relevant to a firm’s financial
information or reporting, or even its CSR disclosure. Since MSCI KLD expanded its coverage
from the fortune 500 to Russell 1000 and 3000 Index, firms are not able to influence MSCI
KLD’s coverage decision. Therefore, the initiation of MSCI KLD coverage serves as an
13 We also use the difference between the number of total strength and the number of total concerns for all the rated six CSR
categories as an alternative CSR performance measure and we re-estimate Equation (4) in each case. The estimation results (not
tabulated) are consistent with those using the three main CSR measures.
22

exogenous shock to firms. After beginning coverage, these firms are more likely to improve their
CSR performance to maintain their reputation. The initial MSCI KLD coverage provides a quasi-
experiment setting because the initial coverage allows a comparison of firm-specific earnings
between the pre-and-post-MSCI KLD period for firms with subsequent CSR ratings (treated
firms) compared with those (control firms) never covered by MSCI KLD. We modify Equation 4
to provide a difference-in-difference model as follows:
FirmE = β0 + β1 KLDt-1 + β2 POST + β3 (KLDt-1 × POST) + β4 SIZE+ β5 MB +
β6 MktShare + β7 StdROA + β8 DIVERSE + β9 HERF + β10 LEVERAGE + β11 NIND +
Year Fixed Effects + Firm Fixed Effects + ɛ (5)
where variables are defined in the Appendix. 14 The variable of interest is the interaction term
KLD × POST. A positive coefficient indicates CSR performance is associated with increases in
firm-specific earnings providing additional evidence to support our hypothesis. We also vary the
dependent variable to estimate firm-specific cash flows and accruals.
We match each MSCI KLD rated firm in our sample (without replacement) with a non-
MSCI KLD rated firm in the same industry (two-digit SIC codes) by the nearest firm size in the
year of initial MSCI KLD coverage. 15 We estimate Equation (5) and report our results in Table
7. The estimated coefficient of the interaction term is consistent with our earlier results of the
14 We include year fixed effects and firm fixed effects to control for time variant and firm fixed effects and cluster standard errors
by both firm and year level.

15 The sample size in Table 7 (n = 66,758) is much larger than that in Table 4 (n = 27,070). This is because for difference-in-
difference test we match companies with KLD ratings with those are not rated. Moreover, for KLD rated companies we do not
require CSR measures to test the difference-in-difference model. Since there are no missing values to construct CSR measures,
we have a more complete CSR related sample to match with non-rated firms to execute the difference-in-difference test. Further,
matching on size limits any biases that may result caused by how the MSCI KLD database was expanded.
23

positive association between CSR performance and firm-specific earnings (0.061 in model (1), p-
value < 0.05), firm-specific cash flows (0.030 in model (2), p-value < 0.01), firm-specific normal
accruals (0.027 in model (3), p-value < 0.01), and a negative association between CSR
performance and firm-specific discretionary accruals (-0.008 in model (4), p-value < 0.01).
Therefore, results reported in Table 7 further support our hypotheses.
Additional Analysis
Analyzing a comprehensive CSR performance metric presents an interpretability issue
because of the heterogeneous economic implications among the different CSR categories. It is
possible that different dimensions of CSR performance have different and even reversal effects
on firm-specific earnings or its components, or that an individual CSR category or a few CSR
categories are the primary CSR determinants. Therefore, we use the individual net CSR scores
for the categories: the environment (ENV_NET), community (COM_NET), diversity
(DIV_NET), employee relations (EMP_NET), human rights (HUM_NET), product
characteristics (PRO_NET), and industry concerns (IND_CONCERN). 16
We re-estimate Equation (4) as modified and report our results in Table 8 for firm-
specific earnings, cash flows, and accrual components, respectively. In model (1), we find that
the firm-specific earnings are significantly and positively associated with CSR performance on
environmental issues. That is, firm-specific earnings are greater when the strengths of
environmental issues are increased.
Similar results are found for firm-specific cash flows in model (2) and firm-specific
16 IND_CONCERN is an inverse measure of the number of controversial business involvement concerns in the industries to
allow for interpretation that is directionally consistent with the other dimensions.
24

normal accruals in model (3). In addition, firm-specific cash flows are also significantly
increasing in net strengths of product issues, and number of industry concerns. In model (3), we
find that the firm-specific normal accruals are also significant and positively associated with
CSR performance on environmental issues. For firm-specific discretionary accruals in model (4),
we find significant negative coefficients for both net diversity and human rights issues,
consistent with our main findings in Table 6, Panel B; however, we also find a positive
association for industry concerns.
Our analysis of individual CSR categories provides further information about the
association between different dimensions of CSR performance on the overall and components of
firm-specific earnings. Some interesting and meaningful findings are that CSR performance
regarding the environment issues is important in explaining firm-specific earnings, cash flows,
and normal accruals. Further, industry concerns have contradicting associations with the positive
association with both firm-specific cash flows and discretionary accruals.
Next, we separate our narrow CSR measure into strengths and concerns because prior
research suggests strengths and concerns are individually important and could have different
effects and re-estimate our models (Table 9). CSR strengths contribute to the increase of firm-
specific earnings, cash flows, and normal accruals but are associated with lower discretionary
accruals, consistent with our overall results. There is no significant association between CSR
concerns and any of the firm-specific earning proxies except for the firm-specific discretionary
accruals. We conclude that CSR strengths drive our results on the positive associations between
CSR performance and firm-specific earnings as well as firm-specific cash flow and normal
accruals, and both strengths and concerns drive the negative association for firm-specific
25

discretionary accruals.
Although we control for firm-fixed effects, there is a potential concern that the effect of
CSR performance on firm-specific earnings may be spurious if past firm-specific earnings
significantly explain CSR performance variation. To address this potential endogeneity issue, we
include the firm-specific earnings metric from the prior year in Equation (4) and re-estimate the
model. Results (untabulated) are consistent with those previously reported, providing evidence
that the association is not spurious.
MSCI KLD coverage increased over time. We therefore limited the sample to 2003 –
2018 when the coverage definition appeared to stabilize and re-estimate our models. Our results
(untabulated) over this shorter period are similar to those reported.
Finally, we separately consider profit and loss firms to determine if the association holds
for both types of firms. Our subsamples consist of 21,086 profitable and 5,985 loss firms,
showing that most CSR firms are profitable. Our results (untabulated) report that the positive
association between CSR performance and firm-specific earnings is primarily evident among
profitable CSR firms.
CONCLUSIONS
Prior studies examine the relationship between CSR performance and firm performance
with mixed findings. One concern contributed to the mixed results is an endogeneity issue (i.e.
whether CSR performance increases firm performance or firms with better performance engage
in CSR). Our study minimizes this issue by examining the effect of CSR performance on firm-
specific earnings and its components. Relying on RBV, we expect that firms with unique
resources (i.e. CSR performance) gain a competitive advantage over their rivals and thus have
26

more firm-specific earnings and cash flows.
Our study design uses multiple CSR measures including narrow and broad measures of
CSR as well as an industry adjusted CSR measures and applies both OLS and quasi-experimental
difference-in-difference models to strengthen our results. The estimated results from these tests
are consistent, implying a high likelihood of a strong association between CSR performance and
firm-specific earnings and cash flows. However, using MSCI KLD data is a limitation because of
the dataset’s limited coverage and the method of classification.
Our findings provide support for the RBV that CSR is an effective business strategy to
establish competitive advantages as reflected in higher firm-specific earnings and cash flows.
The competitive advantage created from CSR activities helps firms generate extra value such as
unique product features integrated with sustainability and lower energy consumption, higher
quality, different aspects of production safety including data security, favorable employee
relations ideal for efficiency and innovation., Eventually, all these created values help CSR firms
constitute both a unique capability and a trustworthy reputation that enable firms to maximize
their firm-specific net cash inflow.
27

REFERENCES
Albuquerque, R., Y. Koskinen, and C. Zhang (2019), ‘Corporate social responsibility and firm
risk: Theory and empirical evidence’, Management Science, Vol. 65, No. 10, pp. 4451-4469.
Amit, R., and P. J. H. Schoemaker (1993), ‘Strategic assets and organizational rent’, Strategic
Management Journal, Vol. 14, No. 1, pp. 33-46.
Barney, J. (1991), ‘Firm Resources and Sustained Competitive Advantage’, Journal of
Management, Vol. 17, No. 1, pp. 99-121.
Benabou, R., and J. Tirole (2010), ‘Individual and corporate social responsibility’, Economica,
Vol. 77, No. 305, pp. 1-19.
Bhandari, A., and D. Javakhadze (2017), ‘Corporate social responsibility and capital allocation
efficiency’, Journal of Corporate Finance, Vol. 43, pp. 354-377.
Bocquet, R., C. Le Bas, C. Mothe, and N. Poussing (2017), ‘CSR, innovation, and firm
performance in sluggish growth contexts: A firm-level empirical analysis’, Journal of
Business Ethics, Vol. 146 No. 1, pp. 241-254.
Bose, S., and C. Yu (2023), ‘Does earnings quality influence corporate social responsibility
performance? Empirical evidence of the causal link’, Abacus, forthcoming.
Brown, N. C., and M. D. Kimbrough (2011), ‘Intangible investment and the importance of firm-
specific factors in the determination of earnings’, Review of Accounting Studies, Vol. 16,
No. 3, pp. 539-573.
Burke, Q. L., P. C. Chen, , and G. J. Lobo (2020), ‘Is corporate social responsibility performance
related to conditional accounting conservatism?’, Accounting Horizons, Vol 34, No. 2, pp.
19-40.
Burnett, R. D., and D. R. Hansen (2008), ‘Ecoefficiency: Defining a role for environmental cost
28

management’, Accounting, Organizations and Society, Vol. 33, No. 6, pp. 551-581.
Burnett, R. D., C. J. Skousen, and C. J. Wright (2011), ‘Eco-effective management: An empirical
link between firm value and corporate sustainability’, Accounting & the Public Interest, Vol.
11, pp. 1–15.
Chakravarthy, J., E. deHaan, and S. Rajgopal (2013), ‘Reputation repair after a serious
restatement’, The Accounting Review, Vol. 89, No. 4, pp. 1329-1363.
Cheng, B., I. Ioannou, and G. Serafeim (2014), ‘Corporate social responsibility and access to
finance’, Strategic Management Journal, Vol. 35, No. 1, pp. 1-23.
Chih, H. L., C. Shen, and F. Kang, (2008), ‘Corporate social responsibility, investor protection,
and earnings management: Some international evidence’, Journal of Business Ethics, Vol.
79, No. 1-2, pp. 179-198.
Clarkson, P. M., J. Ponn, G. D. Richardson, F. Rudzicz, A. Tsang, and J. Wang (2020), ‘A
textual analysis of US corporate social responsibility reports’, Abacus, Vol. 56, No. 1, pp. 3–
34.
Clarkson, P. M., M. B. Overell, and L. Chapple (2011), ‘Environmental reporting and its relation
to corporate environmental performance’, Abacus, Vol. 47, No. 1, pp. 27–60.
Contrafatto, M. (2014), ‘The institutionalization of social and environmental reporting: An
Italian narrative’, Accounting, Organizations and Society, Vol. 39, No. 6, pp. 414-432.
Cooper, E. W., and H. Uzun (2015), ‘Corporate social responsibility and the cost of debt’,
Journal of Accounting & Finance, Vol. 15, No. 8, pp. 2158-3625.
Cui, J., H. Jo, and H. Na (2018), ‘Does corporate social responsibility affect information
asymmetry?’, Journal of Business Ethics, Vol. 148, No. 3, pp. 1–24.
Desai, H., S. Rajgopal, and M. Venkatachalam (2004), ‘Value‐glamour and accruals mispricing:
29

One anomaly or two?’, The Accounting Review, Vol. 79, No. 2, pp. 355-385.
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, Vol. 111, No.
1, pp. 158-180.
Dodds, W. B., K. B. Monroe, and D Grewal (1991), ‘Effects of price, brand, and store
information on buyers’ product evaluations’, Journal of Marketing Research, Vol. 28, No. 3,
pp. 307-319.
Dowell, G., S. Hart, and B. Yeung (2000), ‘Do corporate global environmental standards create
or destroy market value?’, Management Science, Vol. 46, No. 8, pp. 1059-1074.
Du, S., C. B. Bhattacharya, and S. Sen (2011), ‘Corporate social responsibility and competitive
advantage: Overcoming the trust barrier’, Management Science, Vol. 57, No. 9, pp. 1528-
1545.
Durnev, A., R. Morck, and B. Yeung (2004), ‘Value-enhancing capital budgeting and firm-
specific stock return variation’, Journal of Finance Vol. 59, No. 1, pp. 65–105.
El Ghoul, S., O. Guedhami, C. C. Y. Kwok, and D. R. Mishra (2011), ‘Does corporate social
responsibility affect the cost of capital?’, Journal of Banking and Finance, Vol. 35, No. 9,
pp. 2388-2406.
Figge, F., and T. Hahn (2013), ‘Value drivers of corporate eco-efficiency: Management
accounting information for the efficient use of environmental resources’, Management
Accounting Research, Vol. 24, No. 4, pp. 387-400.
Flammer, C. (2015), ‘Does corporate social responsibility lead to superior financial
performance? A regression discontinuity approach’, Management Science, Vol. 61, No. 11,
pp. 2549-2568.
30

Friedman, M. (1970), ‘The social responsibility of business is to maximise its profits. New York
Times Magazine (September).
Graham, J., C. Harvey, and S. Rajgopal (2005), ‘The economic implications of corporate
financial reporting’, Journal of Accounting and Economics, Vol. 40, No. 1/3, pp. 3–73
Hall, B. H. (2002), ‘The financing of research and development’, Oxford Review of Economic
Policy, Vol. 18, No. 1, pp. 35-51.
Hall, R. (1993), ‘A framework linking intangible resources and capabilities to sustainable
competitive advantage’, Strategic Management Journal, Vol. 14, No. 8, pp. 607-618.
Harley-Davidson (2019), ‘2019 Harley-Davidson Annual Sustainability Report. Available at:
https://investor.harley-davidson.com/static-files/651eaa8c-28f3-43cc-a1ff-3e1e913c85ce.
Henri, J. F., and M. Journeault (2010), ‘Eco-control: The influence of management control
systems on environmental and economic performance’, Accounting, Organizations and
Society, Vol. 35, No. 1, pp. 63-80.
Hong, H. G., J. D. Kubik, and J. A. Scheinkman (2012), ‘Financial constraints on corporate
goodness. National Bureau of Economic Research. Available at:
https://www.nber.org/papers/w18476.
Hong, Y., and M. L Andersen (2011), ‘The relationship between corporate social responsibility
and earnings management: An exploratory study’, Journal of Business Ethics, Vol. 104, No.
4, pp. 461-471.
Hribar, P., and N. Yehuda (2015), ‘The mispricing of cash flows and accruals at different life‐
cycle stages’, Contemporary Accounting Research, Vol. 32, No. 3, pp. 1053-1072.
Huang, X. B., and L. Watson (2015), ‘Corporate social responsibility research in accounting’,
Journal of Accounting Literature Vol. 34, pp. 1-16.
31

Hui, K. W., and P. E. Yeung (2013), ‘Underreaction to industry-wide earnings and the post-
forecast revision drift’, Journal of Accounting Research, Vol. 51, No. 4, pp. 701-737.
Hui, K. W., K. K. Nelson, and P. E. Yeung (2016), ‘On the persistence and pricing of industry-
wide and firm-specific earnings, cash flows, and accruals’, Journal of Accounting and
Economics, Vol. 61, No. 1, pp. 185-202.
Kaplan, S. E., and S. P. Ravenscroft (2004), ‘The reputation effects of earnings management in
the internal labor market’, Business Ethics Quarterly, Vol. 14, No. 3, pp. 453-478.
Kemper, J., O. Schilke, M. Reimann, X. Wang, and M. Brettel (2013), ‘Competition-motivated
corporate social responsibility’, Journal of Business Research, Vol. 66, No. 10, pp. 1954-
1963.
Khan, M., G. Serafeim, and A. Yoon (2016), ‘Corporate sustainability: First evidence on
materiality’, The Accounting Review, Vol. 91, No. 6, pp. 1697-1724.
Khurana, I. K., R. Pereira., and E. Zhang (2018), ‘Is real earnings smoothing harmful? Evidence
from firm‐specific stock price crash risk’, Contemporary Accounting Research, Vol. 35, No.
1, pp. 558-587.
Kim, Y., M. S. Park, and B. Wier (2012), ‘Is earnings quality associated with corporate social
responsibility?’, The Accounting Review, Vol. 87, No. 3, pp. 761–796.
Kitzmueller, M., and J. Shimshack (2012), ‘Economic perspectives on corporate social
responsibility’, Journal of Economic Literature, Vol. 50, No. 1, pp. 51-84.
Kothari, S., A. Leone, and C. Wasley (2005), ‘Performance matched discretionary accrual
measures’, Journal of Accounting and Economics, Vol. 39, No. 1, pp. 163-197.
Lattanzio, G., and L. P. Litov (2020), ‘Does competing through corporate social responsibility
engagements lead t superior financial performance?’, Working paper, available at SSRN:
32

https://ssrn.com/abstract=3449259.
Lins, K. V., H. Servaes, and A. Tamayo (2017), ‘Social capital, trust, and firm performance: The
value of corporate social responsibility during the financial crisis’, Journal of Finance, Vol.
72, No. 4, pp. 1785-1824.
Lippman, S. A., and R. P. Rumelt (1982), ‘Uncertain imitability: An analysis of interfirm
differences in efficiency under competition’, The Bell Journal of Economics, Vol. 13, pp.
418-438.
Loikkanen, T., and K. Hyytinen (2011), ‘Corporate social responsibility and competitiveness –
empirical results and future challenges’ In Environmental Management Accounting and
Supply Chain Management (pp. 151-170).
Long, M. S., and I. B. Malitz (1985), ‘Investment patterns and financial leverage’ In Corporate
Capital Structures in the United States (pp. 325-352).
Lys, T., J. P. Naughton, and C. Wang (2015), ‘Signaling through corporate accountability
reporting. Journal of Accounting and Economics, Vol. 60, No. 1, pp. 56-72.
Margolis, J. D., H. A. Elfenbein, and J. P. Walsh. 2012), ‘Does it pay to be good...and does it
matter? A meta-analysis of the relationship between corporate social and financial
performance’, Working paper, available at: https://ssrn.com/abstract=1866371 or
http://dx.doi.org/10.2139/ssrn.1866371.
McWilliams, A., D. S. Siegel, and P. M. Wright (2006), ‘Corporate social responsibility:
Strategic implications’, Journal of Management Studies, Vol. 43, No. 1, pp. 1–18.
Modigliani, F., and M. Miller (1958), ‘The cost of capital, corporation finance and the theory of
investment’, The American Economic Review, Vol. 48, No. 3, pp. 261-297.
Morck, R., B. Yeung, and W. Yu (2000), ‘The information content of stock markets: why do
33

emerging markets have synchronous stock price movements?’, Journal of Financial
Economics, Vol. 58, No. 1-2, pp. 215–260.
Palepu, K., V. Bernard, P. Healy, and E. Peek (2007), ‘Business Analysis and Valuation: Text
and Cases. Cengage Learning EMEA.
Petrovits, C. M, (2006), ‘Corporate-sponsored foundations and earnings management’, Journal
of Accounting and Economics, Vol. 41, No. 3, pp. 335-362.
Piotroski, J. D., and D. T. Roulstone (2004), ‘The influence of analysts, institutional investors,
and insiders on the incorporation of market, industry, and firm-specific information into
stock prices’, The Accounting Review, Vol. 79, No. 4, pp. 1119-1151.
Pondeville, S., V. Swaen, and Y. De Rongé (2013), ‘Environmental management control
systems: The role of contextual and strategic factors’, Management Accounting Research,
Vol. 24, No. 4, pp. 317-332.
Porter, and M.R. Kramer (2011), ‘Creating Shared Value. Harvard Business Review. Available
at: https://hbr.org/2011/01/the-big-idea-creating-shared-value.
Prior, D., J. Surroca, and J. A. Tribo (2008), ‘Are socially responsible managers really ethical?
Exploring the relationship between earnings management and corporate social
responsibility’, Corporate Governance, Vol. 16, No. 3, pp. 160–177.
Rodrigue, M., M. Magnan, and E. Boulianne (2013), ‘Stakeholders’ influence on environmental
strategy and performance indicators: A managerial perspective’, Management Accounting
Research, Vol. 24, No. 4, pp. 301-316.
Servaes, H., and A. Tamayo (2013), ‘The impact of corporate social responsibility on firm value:
The role of customer awareness’, Management Science, Vol. 59, No. 5, pp. 1045–1061.
Siegel, D. S., and D. F. Vitaliano (2007), ‘An empirical analysis of the strategic use of corporate
34

social responsibility’, Journal of Economics and Management Strategy, Vol. 16, No. 3, pp.
773-792.
Sloan, R. G. (1996), ‘Do stock prices fully reflect information in accruals and cash flows about
future earnings?’, Accounting Review, Vol. 71, No. 3, pp. 289-315.
US SIF (20190, US|SIF 2019 Annual Report. Available at:
https://www.ussif.org/files/Publications/2019USSIFAnnualReport_online.pdf.
Xu, H., J. Wu, and M. Dao (2020), ‘Corporate social responsibility and trade credit’, Review of
Quantitative Finance and Accounting, Vol. 54, pp. 1389-1416.
35

Appendix - Variable Definitions
Variable Definition
CSR_METRIC = Takes the value of one of the CSR measures defined below
CSR_NARROW = k
# of firm − year strengths within each category
�(
maximum possible # of strengths in each category year
1
# of firm − year concerns within each category

− )
maximum possible # of concerns in each category year
where k represents community, diversity, employee relations,
environment, human rights, and product quality and safety
categories
CSR_BROAD = k
�(
1

− )
sum of controversial business involvement concerns

−( )
maximum possible of controversial business
involvement concerns
where k represents the six rating categories above and controversial
business involvement
CSR_N − Minimum CSR_N

CSR_IDX = Maximum CSR_N − Minimum CSR_N
where CSR_N is a CSR performance score at the firm-year level
(=total strength – total concerns in the six rating categories);
Minimum CSR_N is the minimum CSR performance score for the
36

firm’s industry in that year; and Maximum CSR_N is the maximum
CSR performance score for the firm’s industry in that year
CSR_STRENGTHS = k
�( )
1
where k = 1-6 represent community, diversity, employee relations,
environment, human rights, and product quality and safety KLD
categories
CSR_CONCERNS = k
�( )
1
where k = 1-6 represent community, diversity, employee relations,
environment, human rights, and product quality and safety KLD
categories
FirmE = Difference between a firm’s operating income after depreciation
divided by average assets and the industry average of operating
income after depreciation divided by average assets
FirmCF = Difference between a firms’ operational cash flow divided by
average assets and the industry average of operating cash flows
divided by average assets
FirmAcc = Difference between FirmE and FirmCF
NORMAcc = Firm-specific normal accruals predicted from the performance-
adjusted discretionary accrual model (Kothari, et al. 2005)
37

DISCAcc = Firm-specific abnormal accruals, estimated as the residuals from
the performance-adjusted discretionary accrual model (Kothari, et
al. 2005)
SIZE = Natural log of the market value of common equity
MktShare = Firm sales divided by the total sales of the two-digit SIC code in
which the firm operates
StdROA = Standard deviation of return on assets over five years. Return on
asset equals income before extraordinary and special items divided
by the beginning of year total assets
MB = Ratio of market value of common equity to book value of common
equity
DIVERSE = Revenue-based Herfindahl Index for firm diversification using the
reported business segments of the firm
HERF = Revenue-based Herfindahl Index for industry-level concentration
LEVERAGE = Ratio of long-term debt to the sum of long-term debt and book
value of equity
Nind = Number of firms in the industry
KLD = An indicator equal to 1 for firms with MSCI KLD ratings during
the sample period, and 0 for control firms that never received a
MSCI KLD rating
POST = An indicator variable equal to 1 for the post-MSCI KLD coverage
period and 0 for the pre-coverage years
Note: All continuous variables are winsorized at the 1st and 99th percentiles.
38

Table 1 Sample Distribution by Year
Year Frequency Percent

1991 323 1.19
1992 336 1.24
1993 353 1.30
1994 367 1.36
1995 380 1.40
1996 383 1.41
1997 388 1.43
1998 383 1.41
1999 382 1.41
2000 385 1.42
2001 575 2.12
2002 572 2.11
2003 1,288 4.76
2004 1,611 5.95
2005 1,539 5.69
2006 1,489 5.50
2007 1,435 5.30
2008 1,482 5.47
2009 1,544 5.70
2010 1,599 5.91
2011 1,509 5.57
2012 1,400 5.17
2013 1,384 5.11
2014 1,279 4.72
2015 1,142 4.22
2016 1,154 4.26
2017 1,130 4.17
2018 1,258 4.65
Total 27,070 100.00
39

Table 2 – Descriptive Statistics
Variable Mean Standard 25th Median 75th

(N = 27,070) Deviation Percentile Percentile
FirmE 0.694 0.914 0.122 0.385 0.901
FirmCF 0.287 0.287 0.080 0.216 0.423
NORMAcc 0.398 0.727 0.139 0.025 0.456
DISCAcc 0.015 0.099 0.018 -0.029 0.064
CSR_NARROW 0.022 0.640 -0.333 0.000 0.242
CSR_BROAD 0.004 0.642 -0.333 0.000 0.206
CSR_IDX 0.412 0.257 0.220 0.382 0.576
SIZE 7.225 1.526 6.114 7.160 8.270
MB 3.323 3.319 1.579 2.441 3.935
MktShare 0.016 0.033 0.001 0.003 0.015
StdROA 0.074 0.116 0.019 0.038 0.080
DIVERSE 0.734 0.274 0.490 0.804 1.000
HERF 0.063 0.056 0.033 0.041 0.069
LEVERAGE 0.205 0.182 0.033 0.188 0.316
NIND 5.336 1.175 4.419 5.670 6.394
Table 2 reports the descriptive statistics for all variables included in our primary research models. All variables are
defined in the Appendix.
40

Table 3 – Pearson Correlation Matrix
Panel A – Variables 1 through 8
(1) (2) (3) (4) (5) (6) (7) (8)

(1) FirmE 1.00
(2) FirmCF 0.73 1.00
(3) NORMAcc 0.95 0.51 0.99
(4) DISCAcc 0.14 0.08 0.01 1.00
(5) CSR_NARROW 0.10 0.16 0.07 0.00 1.00
(6) CSR_BROAD 0.11 0.16 0.07 0.00 1.00 1.00
(7) CSR_IDX -0.08 -0.12 -0.06 -0.03 0.53 0.53 1.00
(8) SIZE -0.04 -0.02 -0.04 0.02 0.27 0.25 0.15 1.00
(9) MB 0.11 0.18 0.07 -0.01 0.12 0.12 0.03 -0.02
(10) MktShare -0.19 -0.24 -0.15 -0.04 0.07 0.06 0.14 0.43
(11) StdROA 0.11 0.07 0.12 -0.10 -0.07 -0.06 -0.05 -0.31
(12) DIVERSE 0.02 0.01 0.02 -0.04 -0.05 -0.03 -0.01 -0.33
(13) HERF -0.26 -0.33 -0.20 -0.07 -0.04 -0.05 0.07 0.04
(14) LEVERAGE -0.03 -0.09 -0.01 -0.02 0.05 0.04 0.03 0.35
(15) NIND 0.41 0.51 0.32 0.13 0.05 0.06 -0.11 -0.11
Panel B – Variables 9 through 15
(9) (10) (11) (12) (13) (14) (15)

(9) MB 1.00
(10) MktShare -0.01 1.00
(11) StdROA 0.10 -0.15 1.00
(12) DIVERSE 0.09 -0.14 0.19 1.00
(13) HERF -0.09 0.32 -0.12 -0.03 1.00
(14) LEVERAGE -0.01 0.13 -0.10 -0.16 0.03 1.00
(15) NIND 0.14 -0.45 0.22 0.09 -0.68 -0.14 1.00
Table 3 reports the Pearson correlation coefficient matrix for all variables included in our primary research models.
Variables are defined in the Appendix. N = 27,501. Bold represent significance at the 0.05 level.
41

Table 4 – Association of CSR with Firm-specific Earnings
Pred. (1) (2) (3)

CSR_NARROW H1: + 0.043***
(3.049)
CSR_BROAD H1: + 0.044***
(3.018)
CSR_IDX H1: + 0.154*
(1.993)
SIZE + 0.067** 0.067** 0.067**
(2.290) (2.290) (2.273)
MB + 0.008* 0.008* 0.008*
(1.999) (2.002) (1.935)
MktShare ? 0.409 0.416 0.408
(0.708) (0.720) (0.696)
StdROA + 0.226 0.227 0.230
(1.259) (1.261) (1.274)
DIVERSE + 0.047 0.047 0.044
(1.081) (1.071) (1.004)
HERF - -0.006 -0.004 -0.006
(-0.009) (-0.006) (-0.009)
LEVERAGE - -0.115 -0.115 -0.113
(-1.601) (-1.604) (-1.558)
NIND + 0.445*** 0.445*** 0.452***
(4.592) (4.592) (4.603)
Constant -2.232*** -2.230*** -2.327***
(-3.525) (-3.525) (-3.571)
Adjusted R2 0.351 0.351 0.352
Year fixed effects Yes Yes Yes
Firm fixed effects Yes Yes Yes
Observations 27,070 27,070 27,070
Table 4 reports estimations of Equation (4) where the dependent variable is FirmE and the CSR_METRIC is
varied as indicated in the table:
FirmE = β0 + β1 CSR_METRIC + β2 SIZE+ β3 MB + β4 MktShare + β5 StdROA + β6 DIVERSE + β7 HERF +
β8 LEVERAGE + β9 NIND + Year Fixed Effects + Firm Fixed Effects + ɛ
All variables are defined in the Appendix. ***, **, * indicate significance at 1%, 5%, and 10%, respectively, using
t-statistics based on heteroscedasticity robust standard errors adjusting for clustering at the firm and year levels. t-
statistics are reported under the coefficient estimates.
42

Table 5 – Association of CSR with Firm-specific Cash Flow
Pred. (1) (2) (3)

CSR_NARROW H2: + 0.013***
(3.689)
CSR_BROAD H2: + 0.014***
(3.794)
CSR_IDX H2: + 0.033*
(1.806)
SIZE + 0.033*** 0.033*** 0.033***
(4.620) (4.621) (4.646)
MB + 0.004*** 0.004*** 0.004***
(4.011) (4.016) (3.941)
MktShare ? -0.036 -0.034 -0.034
(-0.182) (-0.172) (-0.168)
StdROA + 0.071* 0.071* 0.071*
(1.919) (1.924) (1.913)
DIVERSE + 0.017 0.017 0.016
(1.361) (1.348) (1.311)
HERF - -0.072 -0.072 -0.072
(-0.598) (-0.594) (-0.593)
LEVERAGE - -0.136*** -0.136*** -0.135***
(-7.516) (-7.522) (-7.450)
NIND + 0.134*** 0.134*** 0.136***
(10.888) (10.879) (11.039)
Constant -0.659*** -0.658*** -0.683***
(-8.428) (-8.419) (-8.577)
Adjusted R2 0.564 0.564 0.564
Observations 27,070 27,070 27,070
Table 5 reports estimations of Equation (4) where the dependent variable is FirmCF and the CSR_METRIC is varied
as indicated in the table:
FirmCF = β0 + β1 CSR_METRIC + β2 SIZE+ β3 MB + β4 MktShare + β5 StdROA + β6 DIVERSE + β7 HERF +
the t-statistics based on heteroscedasticity robust standard errors adjusting for clustering at the firm and year levels.
t-statistics are reported under the coefficient estimates.
43

Table 6 – Association of CSR with Firm-specific Accruals
Panel A: Firm-specific Normal Accruals
Pred. (1) (2) (3)

CSR_NARROW H3a: + 0.021***
(3.458)
CSR_BROAD H3a: + 0.021***
(3.370)
CSR_IDX H3a: + 0.067***
(4.434)
SIZE + 0.035*** 0.035*** 0.035***
(3.597) (3.598) (3.572)
MB + 0.003** 0.003** 0.002**
(2.163) (2.168) (2.069)
MktShare ? -0.253 -0.249 -0.243
(-0.966) (-0.953) (-0.909)
StdROA + 0.161*** 0.161*** 0.164***
(4.074) (4.079) (4.116)
DIVERSE + 0.028 0.028 0.027
(1.579) (1.565) (1.504)
HERF - -0.045 -0.043 -0.041
(-0.271) (-0.263) (-0.246)
LEVERAGE - 0.022 0.022 0.023
(1.027) (1.029) (1.081)
NIND + 0.146*** 0.147*** 0.150***
(4.276) (4.279) (4.412)
Constant -0.828*** -0.828*** -0.873***
(-3.730) (-3.730) (-3.954)
Adjusted R2 0.449 0.449 0.449
Observations 27,070 27,070 27,070
Table 6 Panel A reports estimations of Equation (4) where the dependent variable is NORMAcc, and the
CSR_METRIC is varied as indicated in the table:
NORMAcc = β0 + β1 CSR_METRIC + β2 SIZE+ β3 MB + β4 MktShare + β5 StdROA + β6 DIVERSE + β7 HERF +
44

Table 6 (Continued)
Panel B: Firm-specific Discretionary Accruals
Pred. (1) (2) (3)

CSR_NARROW H3b: - -0.003***
(-3.701)
CSR_BROAD H3b: - -0.003***
(-3.682)
CSR_IDX H3b: - -0.005*
(-1.707)
SIZE + -0.008*** -0.008*** -0.008***
(-4.423) (-4.423) (-4.436)
MB + -0.002*** -0.002*** -0.002***
(-7.465) (-7.466) (-7.351)
MktShare ? 0.068 0.067 0.065
(1.164) (1.151) (1.116)
StdROA + -0.035** -0.035** -0.034**
(-2.548) (-2.550) (-2.542)
DIVERSE + -0.003 -0.003 -0.003
(-0.703) (-0.694) (-0.695)
HERF - -0.009 -0.009 -0.010
(-0.410) (-0.420) (-0.438)
LEVERAGE - 0.046*** 0.046*** 0.046***
(7.278) (7.277) (7.222)
NIND + -0.013*** -0.013*** -0.014***
(-4.452) (-4.457) (-4.657)
Constant 0.108*** 0.108*** 0.113***
(5.532) (5.532) (5.810)
Adjusted R2 0.284 0.284 0.283
Observations 27,070 27,070 27,070
Table 6 Panel B reports estimations of Equation (4) where the dependent variable is DISCAcc, and the
CSR_METRIC is varied as indicated in the table:
DISCAcc = β0 + β1 CSR_METRIC + β2 SIZE+ β3 MB + β4 MktShare + β5 StdROA + β6 DIVERSE + β7 HERF +
45

Table 7 - Difference-in-Differences Analysis
(1) (2) (3) (4)

FirmE FirmCF NORMAcc DISCAcc
KLD -0.016 -0.005 -0.004 -0.002
(-0.766) (-0.890) (-0.496) (-0.819)
POST 0.282*** 0.054*** 0.109*** 0.006**
(4.583) (3.198) (4.351) (2.639)
KLD X POST 0.061** 0.030*** 0.027*** -0.008***
(2.284) (4.154) (2.860) (-3.472)
SIZE 0.039*** 0.028*** 0.011*** -0.010***
(4.691) (10.189) (4.558) (-15.105)
MB 0.002 0.003** -0.000 -0.002***
(0.621) (2.610) (-0.072) (-7.776)
MktShare -1.922*** -1.113*** -0.737*** 0.277***
(-5.163) (-9.288) (-4.664) (10.065)
StdROA 0.141 -0.145*** 0.214*** 0.055***
(1.395) (-6.409) (7.628) (6.155)
DIVERSE -0.053* -0.010 -0.009 -0.007**
(-1.913) (-1.276) (-0.739) (-2.683)
HERF -0.149 -0.247** -0.142 -0.011
(-0.264) (-2.316) (-0.630) (-0.630)
LEVERAGE -0.109*** -0.099*** 0.054*** 0.031***
(-3.026) (-8.279) (3.598) (8.812)
NIND 0.216 0.012 0.010 -0.016***
(1.620) (0.449) (0.176) (-4.815)
Constant -0.965 0.008 -0.004 0.135***
(-1.293) (0.053) (-0.012) (7.389)
Adjusted R2 0.246 0.355 0.296 0.054
Year fixed effects Yes Yes Yes Yes
Firm fixed effects Yes Yes Yes Yes
Observations 66,758 66,758 66,758 66,758
Table 7 reports estimations of Equation (5) where the dependent variable is varied as indicated in the table:
Dep.Var. = β0 + β1 KLDt-1 + β2 Post + β3 (KLDt-1 × Post) + β4 SIZE+ β5 MB + β6 MktShare + β7 StdROA +
β8 DIVERSE + β9 HERF + β10 LEVERAGE + β11 NIND + Year Fixed Effects + Firm Fixed Effects + ɛ
46

Table 8 –CSR Components and Firm-Specific Earnings, Cash Flows, and Accruals
(1) (2) (3) (4)

ENV_NET 0.199*** 0.031* 0.136** 0.006
(3.209) (1.734) (2.510) (1.141)
COM_NET 0.047 -0.003 0.023 0.000
(0.700) (-0.217) (0.418) (0.065)
DIV_NET 0.010 0.010 0.013 -0.006**
(0.244) (1.033) (0.444) (-2.182)
EMP_NET -0.033 0.006 0.002 -0.004
(-0.877) (0.503) (0.048) (-0.810)
HUM_NET 0.100 0.019 0.089 -0.013**
(0.762) (0.992) (0.741) (-2.287)
PRO_NET 0.011 0.024* 0.006 -0.001
(0.184) (1.754) (0.131) (-0.141)
IND_CONCERN 0.149 0.128** 0.050 0.040**
(0.777) (2.115) (0.320) (2.151)
SIZE 0.067** 0.033*** 0.037 -0.002
(2.338) (4.713) (1.526) (-0.576)
MB 0.008** 0.004*** 0.003 0.000
(2.234) (4.102) (1.061) (0.231)
MktShare 0.787 0.053 0.582 0.151***
(1.353) (0.278) (1.218) (2.814)
StdROA 0.220 0.071* 0.316* -0.085***
(1.205) (1.939) (1.972) (-3.703)
DIVERSE 0.048 0.016 0.017 0.012*
(0.956) (1.241) (0.470) (2.029)
HERF 0.415 -0.060 0.080 0.051*
(0.567) (-0.512) (0.122) (1.930)
LEVERAGE -0.115 -0.135*** 0.021 -0.027***
(-1.501) (-7.462) (0.331) (-2.784)
NIND 0.454*** 0.133*** 0.299*** 0.025***
(4.609) (11.011) (3.308) (7.168)
Constant -2.288*** -0.652*** -1.551** -0.111***
(-3.497) (-8.361) (-2.687) (-3.678)
Adjusted R2 0.346 0.564 0.228 0.162
Observations 27,070 27,070 27,070 27,070
47

Table 8 (Continued)
Table 8 reports estimations of a modification of Equation (4) where the dependent variable is varied as indicated in
the table and the CSR_METRIC is the array of CSR performance scores for the six MSCI KLD categories
(ENV_NET, COM_NET, DIV_NET, EMP_NET, HUM_NET, PRO_NET, IND_CONCERN):
Dep. Var. = β0 + β1 CSR_METRIC + β2 SIZE+ β3 MB + β4 MktShare + β5 StdROA + β6 DIVERSE + β7 HERF +
48

Table 9 – Analysis of the Effects of CSR Strengths and Concerns
(1) (2) (3) (4)

CSR_N_STRENGTHS 0.060*** 0.015*** 0.027*** -0.002**
(3.515) (2.881) (3.315) (-2.127)
CSR_N_CONCERNS 0.009 0.005 0.008 -0.006***
(0.328) (0.562) (0.859) (-2.880)
SIZE 0.063** 0.036*** 0.034*** -0.009***
(2.147) (5.011) (3.513) (-4.537)
MB 0.008* 0.004*** 0.003** -0.002***
(2.029) (3.719) (2.240) (-7.440)
MktShare 0.597 -0.074 -0.167 0.064
(1.023) (-0.365) (-0.613) (1.150)
StdROA 0.222 0.022 0.159*** -0.035**
(1.238) (0.757) (4.025) (-2.587)
DIVERSE 0.047 0.020 0.027 -0.003
(1.094) (1.684) (1.533) (-0.730)
HERF 0.004 -0.171 -0.039 -0.013
(0.006) (-1.334) (-0.238) (-0.554)
LEVERAGE -0.112 -0.128*** 0.023 0.046***
(-1.565) (-6.906) (1.085) (7.286)
NIND 0.442*** 0.122*** 0.145*** -0.013***
(4.605) (7.883) (4.236) (-4.455)
Constant -2.202*** -0.617*** -0.818*** 0.110***
(-3.528) (-7.031) (-3.696) (5.587)
Adjusted R2 0.351 0.568 0.449 0.284
Observations 27,070 27,070 27,070 27,070
Table 9 reports estimations of Equation (4) where the dependent variable is varied as indicated in the table and the
CSR_METRIC is CSR_N_STRENTHS and CSR_N_CONCERNS:
Dep. Var. = β0 + β1 CSR_METRIC + β2 SIZE+ β3 MB + β4 MktShare + β5 StdROA + β6 DIVERSE + β7 HERF +
49

SSRN Id4452395

Uploaded by

Copyright:

Available Formats

SSRN Id4452395

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

SSRN Id4452395

Uploaded by

Copyright:

Available Formats

Effect of CSR on the Overall and Components of Firm-Specific Earnings

(Running title: CSR and Firm-Specific Earnings)

March 20, 2023

Electronic copy available at: https://ssrn.com/abstract=4452395

We investigate whether firm corporate social responsibility (CSR) performance is associated

unique resources, such as CSR performance, to develop competitive advantages. Competitive

CSR activities and drivers of effective CSR applications.

Keywords: CSR performance, firm-specific accruals, firm-specific cash flows, firm-specific

Electronic copy available at: https://ssrn.com/abstract=4452395

al. 2012; Di Giuli and Kostovetsky, 2014).

Electronic copy available at: https://ssrn.com/abstract=4452395

We focus on firm-specific earnings to provide evidence for this question. Firm-specific

As an idiosyncratic component of firm-level earnings, firm-specific earnings are linked to a

develop exclusive CSR resources and capabilities compared to its peers.

competitive advantage. CSR performance is therefore expected to be positively associated with

conventional and a broader category of CSR activities as well as an industry-adjusted CSR

Electronic copy available at: https://ssrn.com/abstract=4452395

firm’s disclosed earnings and industry-wide earnings.

discretionary accruals to identify relevant components of firm-specific earnings related to CSR

(controversial industry involvement) related CSR activities being value-enhancing (value-

Electronic copy available at: https://ssrn.com/abstract=4452395

and provides evidence consistent with CSR being value-enhancing.

value-enhancing areas of CSR activities.

LITERATURE REVIEW AND HYPOTHESES DEVELOPMENT

Electronic copy available at: https://ssrn.com/abstract=4452395

within the firm.

Internal managerial motivations originate from managers’ focus on efficiency (Burnett

are positively associated with firm financial performance.

The growing number of CSR-related investments implies that CSR is believed to be

associated with firm financial performance, stimulating increasing amounts of sustainable

Electronic copy available at: https://ssrn.com/abstract=4452395

Consistent with these theoretical arguments, some empirical studies document a

Electronic copy available at: https://ssrn.com/abstract=4452395

earnings is a promising avenue by which to begin to answer the CSR-financial performance

Electronic copy available at: https://ssrn.com/abstract=4452395

RBV theory for the following reasons.

CSR is an integral part of firm-level product differentiation strategies. For example,

Harley-Davidson introduced an all-electric motorcycle, LiveWire, in 2019. Harley-Davidson

be part of a firm’s differentiation strategy.

(Lattanzio and Litov, 2020).

Electronic copy available at: https://ssrn.com/abstract=4452395

In sum, CSR is a competitive tool that is part of product differentiation strategies,

performance and firm-specific earnings.

We state the following hypothesis:

to be positively associated with cash flows than accruals.

Electronic copy available at: https://ssrn.com/abstract=4452395

specific cash flows.

Electronic copy available at: https://ssrn.com/abstract=4452395

with normal accruals.

increased abnormal accruals.

3 Total accruals can be separated into normal and abnormal accruals.

Electronic copy available at: https://ssrn.com/abstract=4452395

abnormal accruals. Our discussion leads to the following hypotheses:

specific normal accruals.

specific abnormal accruals.

Measurement of CSR Performance

Electronic copy available at: https://ssrn.com/abstract=4452395

controversial business involvement category includes alcohol, gaming, firearms, military,