Heliyon

Role of Corporate Social Responsibility in Sustaining Earning Value: Insights from an

Emerging Country
--Manuscript Draft--

Manuscript Number: HELIYON-D-20-02122R1

Article Type: Original Research Article

Section/Category: Business and Economics

Keywords: Corporate Social Responsibility; Corporate Financial Performance; High-Level CSR,

Moderate Level; Worst Level CSR; Earning Value; Autoregressive Distributed Lag
(ARDL)

Abstract: The corporate sector in Pakistan, being more focused on profits, has displayed implicit
reluctance in taking care of employees or other stakeholders, particularly the society.
As such, this study evaluated the relationship between Corporate Social Responsibility
(CSR) and Earning Value (EV). In this modern industrial era, the industry has failed to
pace with the needs of stakeholders, thus adversely affecting business and sustainable
socio-economic development in the society. This study identified CSR offenses, like
contemporary studies have found that the policies of the best award-winning CSR
organizations exert a positive contribution to CSR activities. The study outcomes are
mixed which revealed that in Pakistan, firms with better CSR process displayed a
positive association with earning value (EV). However, its influence was insignificant on
firms’ financial condition who have worst/lowest CSR participation. This study
highlights that rigorous but well-designed CSR processes can trigger earning value in
long run, while improperly crafted CSR processes may harm financial condition in the
short run. Firms with mid-level CSR indicate a positive relationship with EV, apart from
enhancing their business structure, project/firm performance, and corporate prospects.
These mid-level CSR practices appeared to have a positive connection with EV.

Powered by Editorial Manager® and ProduXion Manager® from Aries Systems Corporation
Response to Reviewers

Ref: Decision on Manuscript ID HELIYON-D-20-02122

"Role of Corporate Social Responsibility in Sustaining Earning Value: Insights from an Emerging Country"
Journal: Heliyon
We are very grateful to the Editor-in-Chief for giving us the opportunity to revise and resubmit our manuscript for consideration for publication in
Heliyon. We are also extremely grateful for the valuable suggestions and constructive comments from the reviewers, which have been instrumental
in strengthening the manuscript significantly. We have carefully addressed reviewer’s comments and do so sincerely hope that the manuscript meets
the expectations. Please find here under the required improvements in red texts. Thank you very much for your precious time.
Journal Revision Instructions: When you revise your manuscript, please highlight the changes you make in the manuscript by inserting all changes
in by highlighting color Blue (please or track changes mode in MS Word, but it makes manuscript messy and uncomfortable for reviewers to read).
You must also prepare a detailed rejoinder that clearly explains below each reviewer comment on how you have addressed it and wherein the paper
(identify page number(s) these changes appear. In case these instructions are not followed, there will be delays in processing your paper.
Reviewers all comments addition and improvement changes are in Blue in the main manuscript. Also, overall manuscript language and
proofediting/checking have completed.

Sr. # Comment Response

Reviewer: 1
We are incredibly thankful for the valuable suggestions and constructive comments to improve the manuscript. We have carefully try to address
each comment and do so sincerely hope that the manuscript meets your expectations.
1 Methods do not explore sufficient techniques to Method sufficient explanation in support of technique and relationship added and
explain the relationship between the topic of CSR improved in section 3 and 4.
and the problem needed to solve it.
2 Hypotheses do not clearly state since the differences Hypotheses are improved as suggested and placed at page number 9. High practice,
among the three conditions of the high, middle, and middle-level, and low-level Hypotheses and discussion relationship differentiated
low level were not differentiated yet in advance in section 4.
concerning financial performance. What are the
differences between high practice, middle-level, and
low-level if your expected relationship is similar to
be significant and positive? Please state clearly the
hypotheses, especially from H2 to H4, when they
are being discussed in light of past studies rather
than showing all hypotheses under one list. The

Page 1 of 21
 hypothesis might be separated into more than 5 to
explain more problems to be solved.
3 The findings look complicated but they are not Results in section 4 are improved and as suggested difference between 1) long term
showing a clear relationship among them. I suggest and short term effects; 2) high, middle, and low-level effects. Response placed at
that the result should show the difference between page 12 to half 17.
1) long term and short term effects; 2) high, middle,
and low-level effects.
4 The conclusions are too shallow and not collaborate Improved the conclusion and collaborate hypothesis findings and answer of the
on each hypothesis. Generally, a conclusion should problem. Please see page 17, 18 and 19.
answer the problem of the research which is
represented by their hypotheses.

Reviewer: 2
We are incredibly thankful for the valuable suggestions and constructive comments to improve the manuscript. We have carefully try to address
each comment and do so sincerely hope that the manuscript meets your expectations.
1 The authors may modify the leading statement in Modify the leading statement as suggested in section 5.
conclusion, which is unleaded to the finding of data The role of CSR is important in every society, especially in developing countries.
analysis and not a new addition to the theoretical This study endeavors to unearth the viewpoint of the impact of CSR on the earning
framework of the paper. value of firms in Pakistan. In both model and co-integrating as well long run results
showed mixed statistics. Summary of model one ARDL results report that EV had
significant lag effects, and its lag was substantial while at lag 2, it turned
significantly positive at a significance level which support first hypothesis. The
CSP Reputation Index was insignificant, with the negative sign, but at lag 1 it is
significantly positive with a 1.3% significance level, which endorse our alternate
hypothesis. Operating sales has positive impact on earning value in both co-
integrating and long run and accept the proposed alternate hypothesis. In model one
Co-integrating and long run results of corporate social performance index exhibited
a considerable role with EV, and in the long run, the CSP index showed a 1.5%
slightly insignificant level therefore in both co-integrating and long run accept null
hypothesis.
Model two of second equation using the CSR levels summary showed that EV had
significant lag effects. At lag 2, it was significantly positive with a 13.5%

Page 2 of 21
significance level. Therefore the CSR had a significant effect on EV when we
measure at levels decomposed model and accept the alternate hypothesis (H1) of
CSR activities has significant impact on earning value. The mid-level CSR results
showed a significantly positive impact on the significance level of 7% at zero lag
and this is in line with existing literature which claimed that CSR has a positive
impact on EV, thus supporting the study hypothesis (H5). The highest CSR was
significantly positive at 32.09% at initial which endorse the alternate hypotheses
that, but it became insignificant with a negative sign at lags 1 and 2 with 12.9% and
9% insignificance level which indicate that too much high investment on CSR
activities cause the loss in the financials statement. The lowest CSR level turned
insignificant with a negative sign and worst/lowest CSR firms explained that the
CSR affected EV with decreased CSR activities which endorse the proposed
alternate hypothesis. The Operating income Asset had an insignificantly negative
impact at lag 1 and 2 which reject the alternate hypothesis (H2) while the operating
income sale demonstrated a significantly positive impact at a 34% significance
level which endorse alternate hypothesis (H3), in model two.
For second model the result of co-integrating displayed a negative value for EV at
lag 1 while in the long run, CSR had a considerably significant role in EV and
accept the hypothesis H1 of CSR levels has significant role in earning value.
Highest level CSR had been significantly positive at significance level in both Co-
integrating and long run, which support the alternate hypothesis (H4a and H4b). The
lowest CSR was substantially positive in co-integrating form while significant
positive in long run which endorse the null hypothesis of H5a and H5b. Mid-level
CSR results in a co-integrating form and long run were positive and support the
alternate hypothesis H5a and H5b. In Model two operating income to assets showed
a positive significance level therefore accept the alternate hypothesis (H2, H2a, H2b)
in both co-integrating and long run while the operating income had mixed results
of positive impact on co-integrating form and via-a-via in long run. The findings
highlighted the resource-based view that due to cultural, social, and political issues,
CSR practices relied on firm resources. Hence, developing countries need
awareness so that their stakeholders can promote and invest in CSR activities.

Page 3 of 21
The study outcomes conclude that firms with CSR at the initial stage would hurt
their earning value, as increased CSR activities reduced the revenue of the firms.
The findings established that in Pakistan, firms with best CSR had a positive
relationship with EV. There was an adverse effect on the firm’s earning value upon
worst CSR activities due to the imperfect CSR process in first model. Firms with
mid-level CSR displayed a positive relationship with EV in both co-integrating
form and long run, besides enhancing their financial structure, project/ firm
performance, and firm prospects. The findings suggest stakeholder’s theory
recommendations that CSR performance and moderate/balanced activities
maximizes stakeholder value. In literature observe fact that in most developing
countries such as Pakistan, firms hesitate to enthusiastically participate in CSR
activities due to fear that such practices may not stimulate the performance or
profitability of the firm.
By incorporating social responsibility in project management, managers can ensure
the credibility, integrity, and the reputation of the organization. To achieve the
higher standard of CSR and for the successful implementation of CSR practices,
managers have to line up the project requirements or commitments with their
business operations, along with personal objectives. If the CSR is incorporated in
project management, then the project manager has to perform particular tasks, such
as following only those processes to avoid harming the environment and society,
selection of industrial venues that are far from residential areas, protect the natural
resources, participate in social work, as well as adhere to legal rules and regulations.
Government can promote and take advantage of companies CSR activates by
introducing specialized programs to participate in poverty reduction and for a better
society development. It is imperative that the government should legislate on CSR
procedures to offer relief and protection to welfare-oriented organizations.
Policymakers and the government should improve information transparency and
the regulatory framework by considering weak social performance by firms from
less developed regions. A comprehensive-ranging set of guidelines and plans
concerning the corporation’s duties and accountability to the society exist in
socially-responsible companies, whereby they should combine with their corporate
working and decision-making procedures. However, socially-irresponsible

Page 4 of 21
 corporations need to act as responsible affiliates of the community, in which CSR
practices should be included in the planning, controlling, and decision-making
process of the firm. Firms should accomplish their social responsibilities in a way
as one who is obliged to his family. The globalization can support a business to
enhance corporate social performance, although this is not expected to occur. It is
hoped that the upcoming review will investigate the probable repetitive nature of
the connection between CSR and internationalization in the future study. Future
studies also may look into cross-country comparison involving both developed and
developing states, apart from expanding data from multiple countries with
globalization effect.
Page 17 to 19

Reviewer: 3
We are incredibly thankful for the valuable suggestions and constructive comments to improve the manuscript. We have carefully try to address
each comment and do so sincerely hope that the manuscript meets your expectations.
1 The discussion of the results is very poor. It should Discussion of results and policy implication as been improved in secton 4 and
be improved to better show the real contribution of impactions in section 5.
the current study. Discussion of results improved in section 4.1. and 4.2. Improved the contribution
part to add more value of results in section are improved. Please see draft changes
in blue highlighted. Section 4 and 5.

4.1. Model 1 Results CSR and Earning Value (Equation 1)

Table 1 presents the effects of the ARDL model to determine the relationships of
variable EV with CSP Reputation Index, Operating Income Asset, and Operating
income Sale. The adjusted sample variables included 368, while after adjustments,
366 observations were applied. The automatic selection of four maximum
dependent lags and an appropriate model were chosen based on AIC. The results
of the ARDL model showed that the R-squared and the adjusted R-squared results
were 0.447 and 0.437, respectively. The value of F-statistic showed that the model
was significant at 48.39, while the Durbin-Watson results showed no
autocorrelation in the data. The results showed that EV had significant lag effects,
and its lag was substantial. At lag 2, it turned significantly positive at a 55.7%

Page 5 of 21
significance level. The results support Hypothesis 1 and the study by Rana and
Asad (2018). Table 1 tabulates the detailed results.
The CSP Reputation Index was insignificant, with the negative sign, but at lag 1 it
is significantly positive with a 1.3% significance level, which endorse our alternate
hypothesis (H1) and is in line with the study by Arshad, Anees, and Ullah (2016).
Operating income Asset had an insignificantly negative impact which accept the
null hypothesis and reject alternate (H2), while Operating Income Sale demonstrate
14.7% significance level and accept the third hypotheses (H3) based on operating
sales has significant impact on firm performance.
[Table 1 near here]
Table 2 presents the results of the ARDL bound test, which indicated that the null
hypothesis had no long term relationship. Based on the variables with I (0)
or I (1) assumption, the ARDL bounds test was conducted. Harris and Sollis (2003)
claimed that by applying the ARDL technique of bound test, one could attain long-
run model unbiased estimates. The bounds test is mainly based on the joint F-
statistic under the null hypothesis of no co-integration. Since its asymptotic
distribution was non- standard and based on the assumption that all the variables in
the ARDL model were integrated of order zero, the first level was calculated. While
the variables were integrated of order, one was based on assumption, while the
second was calculated; thus, no co-integration of null hypothesis is rejected. Our
test statistics exceeded the upper critical bounds value, and it has been accepted
with the view that F-statistic is less than the lower bounds value. Table 2 shows the
outcomes.
[Table 2 near here]
Table 3 tabulates the results of co-integrating and long-run effects of dependent and
independent variables. The cointegration ARDL model was used in this study to
determine that the correlations of variable EV with CSP Reputation Index,
Operating Income Assets, and Operating Income to Sale ratio variables. The results
of co-integration showed a negative value of earning at lag 2 therefore reject
alternate hypothesis (H1a) and accept null hypothesis (H0a) of co-integrating form.
The value of the corporate social performance index was 1.5%, with a significantly
negative impact. The results are similar to those reported by Arshad, Anees, and

Page 6 of 21
Ullah (2016) using panel regression, which revealed that in short-term scenarios,
there was no significant impact of CSR on financial performance at 5% confidence
level. The ratio of operating earnings displayed a 9% negative significance level.
Operating earnings to sale ratio exerted a significantly positive impact with a 14.2%
significance level and had a significantly positive impact on EV and corporate
social activities in the long term. Therefore, endorse the acceptance of alternate
hypothesis (H3a) in both co-integrating and long run that operating sales has positive
impact on firm performance. Corporate social performance index exhibited a
considerable role with EV, and in the long run, the CSP index showed a 1.5%
slightly insignificant level therefore in both co-integrating and long run accept null
hypothesis. Both EV and operating income to assets ratio had a decisive role, and
the study results showed negative value therefore we reject the alternate hypothesis
and accept null hypothesis. The constant value reflected a significantly positive
impact on model variables.
[Table 3 near here]
For diagnostics and to avoid both multicollinearity and autocorrelation, multiple
tests, such as Breusch-Godfrey Serial Correlation and LM Test, were performed to
analyze the correlations between the variables. The results of the Breusch–Godfrey
LM test displayed no evidence of autocorrelation in residuals. Besides, the model
had no heteroscedasticity, as indicated by the results of the Breusch-Pagan-Godfrey
test with an F-statistic value of 2.1342 and Probability at 0.048. The result of the
Ramsey RESET test t-value is 1.310 and F-statistics 1.718 fitted values. The
complete robustness test results of equation 1 are demonstrated in appendix A.
4.2. Model 2 Results of CSR Levels
Table 4 shows the complete results of the second equation using the ARDL model
for CSR levels and other variables. The best model was selected based on the AIC
by using the automatic selection of maximum dependent lags 4. The results showed
that the R-squared and adjusted R-squared indicated the goodness of fit at 0.512
and 0.495. The results of F-statistics demonstrate that the model was significant at
30.580, while the Durbin-Watson results showed no auto-correlation in the data.
The results showed that EV had significant lag effects. At lag 2, it was significantly
positive with a 13.5% significance level. Therefore, the CSR had a significant effect

Page 7 of 21
on EV when we measure at levels decomposed model and accept the alternate
hypothesis (H1) of CSR activities has significant impact on firm performance. The
outcomes are consistent with the findings reported by Kim, Kim, and Qain (2018),
which led to a competitive advance of the firms as a result of the stability of
stakeholders. Therefore, the theoretical literature suggests that an increase in CSR
led to betterment in the long run.
The highest CSR was significantly positive at 32.09% at initial which endorse the
alternate hypotheses that, but it became insignificant with a negative sign at lags 1
and 2 with 12.9% and 9% insignificance level which indicate that too much high
investment on CSR activities cause the loss in the financials statement. The lowest
CSR level was significant with a positive sign at a 27.9% significance level in
initial, but at lag 1, it turned insignificant with a negative sign. The worst/lowest
CSR firms explained that the CSR affected EV with decreased CSR activities. The
results of the ARDL method confirmed that lowest CSR activities had a negative
effect on worst-CSR firms and endorse the proposed alternate hypothesis (H6) that
worst-level of CSR practices has insignificant and negative impact on firm earning
value in terms of financial performance.
The mid-level CSR results showed a significantly positive impact on the
significance level of 7% at zero lag. Corporates with mid-level CSR had a
significant role in financial performance with EV. This is in line with existing
literature which claimed that CSR has a positive impact on EV, thus supporting the
study hypothesis (H5) that middle-level CSR has a statistically significant and
positive relationship with firm earning value. The results are consistent with Kim,
Kim, and Qain (2018), who claimed that all theoretical results in favor of CSR
suggested an increase in activities of CSR to bring positive change in productivity
firm earning value. The Operating income Asset had an insignificantly negative
impact at lag 1 and 2 which reject the alternate hypothesis (H2) of CSR levels and
accept null hypothesis that operating income do not have positive effect on earnings
in sample firms of Pakistan. The Operating Income Sale demonstrated a
significantly positive impact at a 34% significance level which endorse alternate
hypothesis (H3), but it turned negative at lag 1. Table 4 above shows the results in
detail.

Page 8 of 21
 [Table 4 near here]
Following the equation 2, the bound test results are demonstrated in Table 5. This
test is relatively more efficient for small sample sizes like in Pakistan, the CSR
concept is not fully implemented in organizations. Therefore, the sample was
collected on the basis of available data of 46 firms, irrespective of the regressors in
the model. Table 5 shows that the null hypothesis has no long term relationship
because the F-statistics is lower than the lower bounds values. Because, when the
independent variables have a lower value, one can assume that the regressors are I
(0), while higher value leads to the assumption of I (1). The null hypothesis of no
long-run relationship can be rejected irrespective of the orders of integration if the
F- statistics are higher than the upper critical value. Results depicted in table 5
below.
[Table 5 near here]
Table 6 demonstrates the results of the co-integrating and long-run effect for the
second model of the dependent and independent variables. In the second equation,
the relationships of dependent variable EV with CSR levels, Operating income
Asset, and Operating income to Sale ratio variables were considered. The result of
co-integrating displayed a negative value for EV at lag 1. Highest level corporate
social responsibility at zero and first lags had been significantly positive at 32%
and 9% significance level, respectively which indicate that results accept the
alternate hypothesis (H4a) that high level CSR practices has significant and positive
impact on earning value in co-integrating form. The lowest CSR was substantially
positive at a 28% significance level at lag 1 in co-integrating form. Mid-level CSR
results in a co-integrating form were positive, with 9% zero lag. The ratio of
operating income to assets showed a 25% positive significance level therefore
accept the alternate hypothesis (H2) that operating income assets has significant
impact on firm performance, while the operating income to sale ratio had a negative
impact on co-integrating form.
In the long run, CSR had a considerably significant role in EV and accept the
hypothesis H1 of CSR levels has significant role in financial performance. The
results demonstrated that in the long run, the highest CSR exhibited a 29%
positively significance level, and the results support Hypothesis (H4b). Meanwhile,

Page 9 of 21
the lowest and mid-level CSR had 39% and 21% significance level. Therefore,
accept the alternate hypothesis of Mid-level CSR while reject the proposed
alternate hypothesis in model 2 of long run. Operating income to sales positively
and significantly contributed to EV and corporate social activities in the long run
and the hypothesis (H3b) of operating sales has positive impact on firm earning
value in long run. Both EV and operating income to assets ratio had a decisive role,
and this study showed a positive value for the long run in equation 2 and endorse
the hypothesis (H2b) that operating assets has negative impact on firm performance
in long run. The results refer to the theory in which a corporation acts as an
instrument for wealth creation, and the objective of their social activities is to
generate good economic outcomes only (Seele & Lock, 2015). The overall results
of equation 2 showed a significantly positive association with firm CSR levels and
earning value but with negative coefficient value which is in line with a recent
study by Golrida et al. (2018) based on stakeholder and legitimacy theories because
discretionary CSR environment enhances earning value in term of financial
performance and vice versa. However, mixed results with the proposed CSR levels
hypothesis as accept alternate of High and Mid-level CSR but worst CSR level and
firm earning value therefore accept H6a, H6b null hypothesis. The findings
highlighted the resource-based view that due to cultural, social, and political issues,
CSR practices relied on firm resources. Hence, developing countries need
awareness so that their stakeholders can promote and invest in CSR activities. Table
6 shows the complete results.
[Table 6 near here]
In equation two, multiple tests were performed for diagnostics and autocorrelation,
as well as to avoid multicollinearity. The LM test was conducted to analyze
correlation, and for this, the Breusch-Godfrey LM test was employed for serial
correlation analysis between the variables, while for equation two, the results of
Breusch–Godfrey LM test showed no evidence of auto-correlation (see Appendix
B). The robustness test of heteroscedasticity results showed no heteroscedasticity
in the model, as demonstrated in the results of the Breusch-Pagan- Godfrey test
with F-statistics value at 1.439 and probability at 0.145(see Appendix B). The level
of collinearity between the regressors in the equation can be measured efficiently

Page 10 of 21

2 What does your result imply for banks, investors,
managers and regulators?
using variance inflation factors (VIFs) test because this test showed that due to
collinearity with the other regressors, the variance of a coefficient estimate of a
regressor had inflated. Another test Ramsey RESET Test is conducted for the
robustness to determine the untitled relationship of variables. The complete
robustness test results are given in appendix B.
Page 12 to 16
Improved the contribution part to add more value as well as implications of
results in section are improved. Please see draft changes in blue highlighted.
Section 4 and 5.
The study outcomes conclude that firms with CSR at the initial stage would hurt
their earning value, as increased CSR activities reduced the revenue of the firms.
The findings established that in Pakistan, firms with best CSR had a positive
relationship with EV. There was an adverse effect on the firm’s earning value upon
worst CSR activities due to the imperfect CSR process in first model. Firms with
mid-level CSR displayed a positive relationship with EV in both co-integrating
form and long run, besides enhancing their financial structure, project/ firm
performance, and firm prospects. The findings suggest stakeholder’s theory
recommendations that CSR performance and moderate/balanced activities
maximizes stakeholder value. In literature observe fact that in most developing
countries such as Pakistan, firms hesitate to enthusiastically participate in CSR
activities due to fear that such practices may not stimulate the performance or
profitability of the firm. This effect has more worth in service oriented sector
especially in banks where regulations are strict by involving large investors and
more managerial skills.
By incorporating social responsibility in project management, managers can ensure
the credibility, integrity, and the reputation of the organization. To achieve the
higher standard of CSR and for the successful implementation of CSR practices,
managers have to line up the project requirements or commitments with their
business operations, along with personal objectives. If the CSR is incorporated in
project management, then the project manager has to perform particular tasks, such
as following only those processes to avoid harming the environment and society,
Page 11 of 21
 selection of industrial venues that are far from residential areas, protect the natural
resources, participate in social work, as well as adhere to legal rules and regulations.
Government can promote and take advantage of companies CSR activates by
introducing specialized programs to participate in poverty reduction and for a better
society development. It is imperative that the government should legislate on CSR
procedures to offer relief and protection to welfare-oriented organizations.
Policymakers and the government should improve information transparency and
the regulatory framework by considering weak social performance by firms from
less developed regions.
Page 18 and 19.
3 The results presented in session 4 are not This issue has been improved and can see in question no 2 here while in draft
interpreted. They only refer if the relationship is page 12 to 16 highlighed text significant improvement has been done.
significantly (positive/negative) or insignificantly
(positive/negative). This issue should be improved.

4 There is a poor theoretical discussion between the
key issues: CSR and corporate financial
performance. Which is the underlying theory (or
theories) behind hypotheses tested? This issue
should be improved.
This issue has been imrpved. Collaborate hypothesis, and answer/ discussion of
theoretical key problems added.
Based on the past literature, theory of the social responsibility theory and stake
holder theory concept following hypotheses are formulated and need to explore in
Pakistan;

Reviewer: 4
We are incredibly thankful for the valuable suggestions and constructive comments to improve the manuscript. We have carefully try to address
each comment and do so sincerely hope that the manuscript meets your expectations. Please see draft changes in blue.
1 In the conclusion, the study says that (1) CSR ``can Improved and correct the incorrect leading statement.
return sustainability'', and ``lead to better
revenue''. But this is incorrect.
2 What is the theory that motivates this model and this Social corporate responsibility and stake holder theory motives to study the impact
choice of variables? on firms earning value. Motivation of the study in section 1 and variables choice
Improved in section 3. In sections 3 approach improved.

Page 12 of 21
3 The results of the analysis do not show this or Results summary added in conclusion section and support with concluding remakrs
support these conclusions.
also improved in section 5.
Summary of model one ARDL results report that EV had significant lag effects,
and its lag was substantial while at lag 2, it turned significantly positive at a
significance level which support first hypothesis. The CSP Reputation Index was
insignificant, with the negative sign, but at lag 1 it is significantly positive with a
1.3% significance level, which endorse our alternate hypothesis. Operating sales
has positive impact on earning value in both co-integrating and long run and accept
the proposed alternate hypothesis. In model one Co-integrating and long run results
of corporate social performance index exhibited a considerable role with EV, and
in the long run, the CSP index showed a 1.5% slightly insignificant level therefore
in both co-integrating and long run accept null hypothesis.
Model two of second equation using the CSR levels summary showed that EV had
significant lag effects. At lag 2, it was significantly positive with a 13.5%
significance level. Therefore the CSR had a significant effect on EV when we
measure at levels decomposed model and accept the alternate hypothesis (H1) of
CSR activities has significant impact on earning value. The mid-level CSR results
showed a significantly positive impact on the significance level of 7% at zero lag
and this is in line with existing literature which claimed that CSR has a positive
impact on EV, thus supporting the study hypothesis (H5). The highest CSR was
significantly positive at 32.09% at initial which endorse the alternate hypotheses
that, but it became insignificant with a negative sign at lags 1 and 2 with 12.9% and
9% insignificance level which indicate that too much high investment on CSR
activities cause the loss in the financials statement. The lowest CSR level turned
insignificant with a negative sign and worst/lowest CSR firms explained that the
CSR affected EV with decreased CSR activities which endorse the proposed
alternate hypothesis. The Operating income Asset had an insignificantly negative
impact at lag 1 and 2 which reject the alternate hypothesis (H2) while the operating
income sale demonstrated a significantly positive impact at a 34% significance
level which endorse alternate hypothesis (H3), in model two.
For second model the result of co-integrating displayed a negative value for EV at
lag 1 while in the long run, CSR had a considerably significant role in EV and
Page 13 of 21
 accept the hypothesis H1 of CSR levels has significant role in earning value.
Highest level CSR had been significantly positive at significance level in both Co-
integrating and long run, which support the alternate hypothesis (H4a and H4b). The
lowest CSR was substantially positive in co-integrating form while significant
positive in long run which endorse the null hypothesis of H5a and H5b. Mid-level
CSR results in a co-integrating form and long run were positive and support the
alternate hypothesis H5a and H5b. In Model two operating income to assets showed
a positive significance level therefore accept the alternate hypothesis (H2, H2a, H2b)
in both co-integrating and long run while the operating income had mixed results
of positive impact on co-integrating form and via-a-via in long run. The findings
highlighted the resource-based view that due to cultural, social, and political issues,
CSR practices relied on firm resources. Hence, developing countries need
awareness so that their stakeholders can promote and invest in CSR activities.
Page 17 and 18.
4 All the authors show is a link, but the link can take Relationship of variables has been improved. More hypothesis are developed as
many forms. Better CSR can increase sales and suggested by reviwers for better link.
customers and thus financial performance. But it
could also be that more profitable firms spend more
to boost CSR because they can afford to. There is a
link, but the results shown do not provide clear
answers or support for the hypotheses.
5 It is important to make a strong case as to why The concept of CSR in the corporate sector within the context of Pakistan differs
studying firms in Pakistan is interesting. Are the from western perception. In Pakistan, CSR is taken for granted as most of the
data better or different in interesting ways? business communities are focused on minting profit for firms, thus neglecting social
welfare, e.g., environment-friendly policies, employees, stakeholders, and social
welfare. Lack of exposure and giving the least importance to procedures mentioned
above with the argument of declaring it as a helpful tool to attract donation (Qazi
et al., 2015; Yunus et al., 2017). However, the Securities Act 2005 and 2007 had
been enacted by the SECP to control deficiencies for the protection of stakeholders
(Yunus et al., 2017). In Pakistan, both awareness and implication of CSR are at its
embryonic stage (Rana & Asad, 2018). However, some firms of Pakistan have
published their annual sustainability reports, i.e., telecom and banking firms, which

Page 14 of 21
genuinely depict their contribution to environmental protection, society, and
welfare of employees at sites.
Most organizations, with a few exceptions, profess the entire claim on the profit
earned. This feature demands extensive debates on NFEH (National Forum for
Environment and Health, 2016). Regarding this essence, this issue cannot be swept
under the carpet as maximum foreign companies feel reluctant to penetrate the
Pakistani market. Developed countries, such as Canada, the UK, the USA, and
Germany even developing country like China, have proper systems for CSR.
Contribution to society is the responsibility of every organization. Thus, it does not
impose much impact on the profit of an organization. However, in the context of
developing countries like Pakistan, only several organizations contribute to society,
hence viewed as a disadvantage when compared to other organizations (NFEH,
2016).
The role of CSR in CFP (ROA, ROE) has been explored by some studies, but not
at all clear understanding has been established due to the contradicting outcomes
(Sacconi & Degli-Antoni 2011; Ahmed, Butt, & Majeed, 2018; Zulfiqar et al.,
2019). However, the relationships of CSR with CFP (Earning Value) is relatively
untapped or with confines in developing countries as compare to developed
countries especially in Pakistan through the use of the advanced statistical
technique on longitudinal data. According to the Khattak, Ronchi, Castelli, and
Sparks (2014), phenomenon of CSR is relatively new in developing countries, thus
gaining attention from researchers since the past decade.
Prior study have examined the role of CSR on employee’s behavior and investor’s
intentions (see Zulfiqar, Sadaf, Popp, Veinhardt, & Máté, 2019). In Pakistan few
studies such as Javeed, and Lefen (2019) have inspected CSR and profitability
(ROA and ROE) by taking moderating effect of corporate governance factors CEO
power and ownership structure. Rana and Asad, (2018) study limitation was only
restricted to manufacturing pharmaceutical sector 9 firms financial performance
and the four-year observation period in Pakistan. Kiran (2015) study only limited
to oil and gas sector 10 firm’s profitability and conducted simple regression in excel
by considering ROA and ROE. Iqbal, Ahmad, Hamad, Bashir, and Sattar (2014)
reviewed CSR possible impact on profitability narrow to banking sector (sample

Page 15 of 21

6 Under Hypotheses, it is not clear what is meant by
high-, medium- and low-level CSR. Moreover, there
is no hypothesis development. By this, I mean that
there is no clear theory to explain the channels or
mechanisms by which CSR is linked to better
financial performance.
7 In section 3, the authors should provide more details
about the Business Recorder Data Stream. It is
important for future readers (and referees) to
understand very well the source of the data, to
only UBL bank) by considering donation as CSR and financial performance as net
profit margin and EPS. There is a lack of understating on the need for allocating
budget for CSR activities at the corporate level. The role of CSR in EV within the
context of Pakistan is rather a novel phenomenon and not fully explored with
different dimensions (Ahmed, Butt, & Majeed, 2018; Rana & Asad, 2018; Javeed,
& Lefen, 2019). Part of the corporate segment has been much conventional and has
remained captive to financial contemplations setting group intrigues in Pakistan.
Due to interest disparity between the organizational part and society for sustainable
development, CSR has been disregarded in Pakistan, mainly social wellbeing and
training that have been given less attention. In the stringent modern and competitive
business arena, the environmental sustainability of businesses mostly depends on
the efficiency and the effectiveness of the financial function of its management.
However, social and economic indicators have been ranked very low in the comity
of the nation. The role of CSR in terms of corporate social performance (financial
aspect) requires further comprehension and exploration, particularly in a
developing country overall sectors, such as Pakistan. Consequently, sustainable
growth is envisioned (Rana & Asad, 2018). Prado et al., (2019) explored future
endeavors by identifying the existing gaps in literature, used statistical analysis,
employed variables or methods is existing studies on current state of the corporate
social performance and corporate financial performance. They conclude that there
is not an overwhelming evidence that would lead to a relationship of truth
relationship so the direction could be pointed out with latest methods.
Hypothesis and relationship has been improved in detail in section 2 and 3.
Based on the past literature, theory of the social responsibility theory and stake
holder theory concept following hypotheses are formulated and need to explore in
Pakistan;
Data description has been provided in detail and in secion 3.
This study used secondary data for 2009-2016 and data were gathered from annual
financial reports of respective organizations. The sample size was limited to 46
firms (see, Appendix C) with 368 total observations, mainly because most firms
Page 16 of 21
 facilitate future replication and comparison. had incomplete and inaccurate CSR data. Given the requirements of data, only
Currently the description of the data sources is not those firms that report at least two consecutive years of data are included by
satisfactory. following (Roberts & Dowling, 2002). In Pakistan, financial reporting of CSR
activities is still at the infancy stage. Sample include manufacturing; fertilizer;
cement building material; pharmaceuticals; automobile; consumer goods and
foods; petroleum, energy and power; textile; information and technological
communication companies listed in Stock Exchange of Pakistan and had at least
hundred million rupees in outstanding shares, because larger companies generally
publish sustainability reports on a regular basis as well regularly participate more
in CSR practices as compare to small firms.
Page 9
8 thee sample of firms with CSR data is a sub-sample Improved.
of all firms. Importantly, it is not a random sub-
sample. Thus, inevitably some biases are
introduced. If we cannot determine the nature and
effect of such biases, we cannot be confident in the
results and their interpretation.
9 The measure of CSR used is CSR expenses. But Improved.
note that this variable scales up with the size of the
firm. Thus, we could spuriously conclude that
bigger firms are better at CSR when in reality they
are simply larger and can afford to spend more on
CSR.

10 There is also a measure of CSR which is a reputation Improved

index, but there is not much detail about how this is Followed Robert and Dowling, (2002) reputation index methodology where each
computed or whether it is robust or reliable. firm’s overall reputation score from firms. The reputation index measurement of
CSR had been based on environment (EVN), employee relations (ER), product
quality and its relationship with customers and providers (PRD), shareholder
relations (SHA), and community (COM). The annual disbursement on CSR
activities served as a proxy of CSR practices in firms based on annual reports of
the firm.

Page 17 of 21
 Data of three proxies were used to measure financial condition as the accounting
data discards data distortion. The variables MVE represent market value of equity,
TA indicate total assets, and BVD represent book value of debt. Therefore, EV
measured from (MVE +BVD) minus TA and then divided by total sales. The
earning access value was used to evaluate the future aspect of firms. Increment in
EV indicated increased financial performance and enhanced CSR. Operating
income to assets was measured from the proxy of operating income divided by the
total assets. Operating income to sales proxy of operating income divided by the
total sales was measured. The ratio of operating income to sale considered a proxy
in empirical research based on leverage variances. Finally, operating income from
assets was used to examine the revenue performance of firms.
Pre-conditions are necessary, such as running the ARDL and ensuring that the
model “Dependent” is non-stationary (Hussain & Mazhar, 2018). In order to
capture the data generating process, the ARDL model takes a sufficient number of
lags in general to a specific modeling framework because this model can capture
the correlation of the co-integrated variables for both short- and long-run (Shrestha
& Bhatta, 2018)
11 In general, the estimation of the models should use The sample size was limited to 46 firms (see, Appendix C) with 368 total
variables that are scaled, to avoid having very observations, mainly because most firms had incomplete and inaccurate CSR data.
different size variables for firms that are big or Given the requirements of data, only those firms that report at least two consecutive
smaller. years of data are included. Sample include manufacturing; fertilizer; cement
building material; pharmaceuticals; automobile; consumer goods and foods;
petroleum, energy and power; textile; information and technological
communication companies listed in Stock Exchange of Pakistan and had at least
hundred million rupees in outstanding shares, because larger companies generally
publish sustainability reports on a regular basis as well regularly participate more
in CSR practices as compare to small firms.
12 The authors use the ARDL econometric approach,
but this is not strongly motivated. There are many Improve and better motivation for the technique has been add
econometric approaches for dynamic panel data. There are multiple regression techniques, such as simple regression, fixed or
Why is this approach better than others in this random effect. But this study applied the Auto Regressive Distributed Lag (ARDL)
problem? approach proposed by Pasaran and Shin (1999) with cointegration as this approach

Page 18 of 21
applied and suggested by recent studies. Such as Ghouse, Khan, and Rehman
(2018) have expressed that one of the most general dynamic unrestricted models in
the econometric literature is the ARDL model, primarily because the ARDL co-
integration approach offers specific tests to determine the presence of a single co-
integrating vector, instead of assuming uniqueness. Pre-conditions are necessary,
such as running the ARDL and ensuring that the model “Dependent” is non-
stationary (Hussain & Mazhar, 2018). In order to capture the data generating
process, the ARDL model takes a sufficient number of lags in general to a specific
modeling framework because this model can capture the correlation of the co-
integrated variables for both short- and long-run (Shrestha & Bhatta, 2018). ARDL
approach is the latest and has been widely accepted due to its numerous significant
advantages, such as its applicability in all cases, regardless of series I (0), I(1), or
mutually co-integrated variables (Oskooee, & Oyolola, 2007). The second
advantage of this contemporary approach is that it can obtain a more comprehensive
array of lags. Notably, in a more common-to-particular method of the ARDL
technique, a wider variety of lag numbers are apprehended in the data to create the
procedure. Based on the lag method, the ARDL estimation assesses the co-
integration of data. For a long-run relationship tested via the ARDL model, two
steps are embedded. In the first step, the presence of the long-run is determined.
The second step then estimates both short- and long-run coefficients of the same
equation (McCann, Baylis, & Williams, 2010).
Sulaiman, Baharin, and AlHadi (2019) reported the concept of a single reduced
equation utilized in the ARDL model. This generates outcome in a more
comprehensive, detailed, reliable, and effective method than any other method of
traditional integration to determine the relationship of both long- and short-run. It
can test any fixed relationship between the variables without taking into
contemplation of its capability to underlying regression values, which are purely I
(0) and I (1) correspondingly or both. In order to discover different and similar logs
from several variables in the dataset, the ARDL approach makes it possible. Thus,
the ARDL is an appropriate method for this study that looked into CSR and firm
Earning Value (financial performance). The ARDL approach was suggested and

Page 19 of 21
 employed by Mamun, Sohog, and Akhter (2013) in Bangladesh to confirm the long-
run relationship of CSR expenditure with firm financial determinants.
This study also applied the bound test co-integration for both long- and short-run
relations. Hussain and Mazhar (2018) reported that the use of the co-integration
technique for estimation avoids spurious regression and endogeneity problems. The
first difference (I) is the number of lag and (t-1) is the level lag of variables. The
Akaike’s Information Criterion (AIC) and some other diagnostic tests were
employed to determine the lag length for each variable, as well as to identify the
presence of a co-integrating relationship. Next, the bound test was performed after
identifying an appropriate lag length. Davidescu (2015) applied the ARDL co-
integration approach to investigate the long-run causal linkages and the short-run
dynamics by stating that the ARDL error correction representation becomes
relatively more efficient with a single long-run relationship and small sample data
size.
This study employed robustness tests, including Breusch- Godfrey Serial
Correlation LM Test, Heteroscedasticity Test Breusch-Pagan-Godfrey, Ramsey
RESET Test, and Variance Inflation Factors for diagnostics. Shrestha and Bhatta
(2018) expressed that for the residual diagnostic LM test (Lagrange multiplier),
heteroscedasticity and correlogram tests are the primary testing approaches for
ARDL.
In determining the long-run relationship between CSR, and financial performance
(EV) activities on varied variables, the ARDL approach was adopted for this study.
It is composed of two steps. The first model presents the conceptual model verified
by the ARDL method to identify the co-integrating form. Second, the long-run
coefficients relationship was determined by adding lags of dependent and
independent variables. On the basis of the above explanation and data, the ARDL
econometric has been used.
13 There is no clear definition of Earning Value or Improved and reported
where it is measured from financial statements. In a nutshell, “Earning Value is an approach where you monitor the project plan,
actual work, and work completed value to see if a project is on track. Earned Value
shows how much of the budget and time should have been spent, considering the
amount of work done so far”. The Earned Value method has been developed as a

Page 20 of 21
 tool facilitating project progress control. Many organizations worldwide adopted
Earning Value as a standard management tool (Czernigowska, 2008). EV analysis
is an important tool to measure the performance of a project. The importance of EV
is because it not only focuses on planned expenditure and actual cost but also
identifies future opportunities and analyses actual performance (Suresh, &
Ganapathy-Ramasamy, 2015).
Page 5 and 6 section 2.
Data of three proxies were used to measure financial condition as the accounting
data discards data distortion. The variables MVE represent market value of equity,
TA indicate total assets, and BVD represent book value of debt. Therefore, EV
measured from (MVE +BVD) minus TA and then divided by total sales. The
earning access value was used to evaluate the future aspect of firms. Increment in
EV indicated increased financial performance and enhanced CSR. Operating
income to assets was measured from the proxy of operating income divided by the
total assets. Operating income to sales proxy of operating income divided by the
total sales was measured. The ratio of operating income to sale considered a proxy
in empirical research based on leverage variances. Finally, operating income from
assets was used to examine the revenue performance of firms.
Page 10 and 11 section 3.
14 There should be enough information for readers or Improved overall paper as best possible.
referees to understand what the data looks like, and
potentially to replicate or compare future analysis.
15 Why is figure 2 so different from figures 1 and 3? Improved. Because figure two besed on mid level. Levels are based on rating
Fig 1 and 3 show curves for highest and lowest, agency provided and in fiure dot in midle incidiate the mid level activies only.
respectively, CSR relative to earnings. But figure 2 Mid-level CSR demonstrate moderate effect therefore the show the balanced
shows only dots. positive point in the mid of figure.

Page 21 of 21
Revised manuscript file - highlighting revisions made Click here to view linked References

1 Role of Corporate Social Responsibility in Sustaining Earning Value:

2 Insights from an Emerging Country
3
4
5
6 Atta Ullaha*, Chen Pinglua*, Saif Ullahb, Asif Ali Safeera and Hafiz Syed Mohsin Abbasa
7
a
8 Huazhong University of Science and Technology, China
9
b
10 SZABIST, Pakistan ORCID iD https://orcid.org/0000-0001-9799-2293
11
12 Corresponding author: Email: attaullah142@gmail.com & chen.pinglu@hust.edu.cn
13
14 Atta Ullah, PhD Scholar, attaullah142@gmail.com ORCID iD https://orcid.org/0000-0002 0590-563X
15 Chen Pinglu, Professor, Head of Department of financial management, chen.pinglu@hust.edu.cn
16
Saif Ullah, PhD Scholar: saifullah_142@yahoo.com (); saifullah_142@yahoo.com
17
18 Asif Ali Safeer, PhD Scholar: aasafeerphd@outlook.com, ORCID iD https://orcid.org/0000-0001-
19 5877-1201
20 Hafiz Syed Mohsin Abbas, PhD Scholar: mohsin01@163.com (ORCID iD https://orcid.org/0000-
21
22 0002-8554-5485
23
24
25
26
Abstract
27
28 The corporate sector in Pakistan, being more focused on profits, has displayed implicit reluctance
29
30 in taking care of employees or other stakeholders, particularly the society. As such, this study
31
32 evaluated the relationship between Corporate Social Responsibility (CSR) and Earning Value
33
(EV). In this modern industrial era, the industry has failed to pace with the needs of stakeholders,
34
35 thus adversely affecting business and sustainable socio-economic development in the society.
36
37 This study identified CSR offenses, like contemporary studies have found that the policies of the
38
39 best award-winning CSR organizations exert a positive contribution to CSR activities. The study
40
41 outcomes are mixed which revealed that in Pakistan, firms with better CSR process displayed a
42
43 positive association with earning value (EV). However, its influence was insignificant on firms’
44 financial condition who have worst/lowest CSR participation. This study highlights that rigorous
45
46 but well-designed CSR processes can trigger earning value in long run, while improperly crafted
47
48 CSR processes may harm financial condition in the short run. Firms with mid-level CSR indicate
49
50 a positive relationship with EV, apart from enhancing their business structure, project/firm
51
52 performance, and corporate prospects. These mid-level CSR practices appeared to have a
53
54 positive connection with EV.
55
56 Keywords: Corporate Social Responsibility; Corporate Financial Performance; High-Level
57
58 CSR, Moderate Level; Worst Level CSR; Earning Value; Autoregressive Distributed Lag
59
60 (ARDL)
61
62
63 Page 1 of 29
64
65
 1. Introduction
At present, the brilliant minds of the entire world are discussing a vital question of how to
1
2 generate more socially-responsible corporate organizations amidst this global competitive
3
4 market. Some companies tend to present themselves as socially-responsible firms. In fact, more
5
6 studies are determined to assess if organizations should focus more on socially-responsible
7 behavior. Managers in outmoded cost-effective discourse are compelled to make decisions to
8
9 aim at maximizing firm equity. In order to meet this goal, strategy planners have resorted to
10
11 every kind of abuses for the present value maximization of an organization and for the
12
13 generation of maximum profits. Financial transparency hinders managers from practicing
14
15 socially-responsible activities as these actions are not aligned with the aforementioned economic
16
17 goals. Hence, the management may ignore essential stakeholders, such as employees,
18 community, and consumers, thus decreasing the present value of net cash flow and affecting the
19
20 interest of firm equity (Yunus, Wazid, Hairi, Choo & Hairi, 2017).
21
22
23 The introduction of the globalization and industrial revolution has resulted in enhanced CSR
24 concepts significantly among practitioners and policymakers (Pimpa & Moore, 2018). Corporate
25
26 social responsibility (CSR) not only accounts for maximizing the stakeholder values, but it is
27
28 also a novel concept to incorporate social duties into the core operation of the organization for
29
30 managing the uprising environmental and social issues (Baumann-Pauly, Wickert, Spence, &
31
32 Scherer, 2013). It does not merely mean that only executives can initiate and incorporate CSR
33
34
in business activities, as project managers can integrate social responsibility practices in their
35 work and facilitate social goods within the organization (Zeng, Ma, Lin, Zeng & Tam, 2015).
36
37 Firms play an essential role in the economic growth of countries, but one cannot dismiss the fact
38
39 that the economic boom of organizations can also be a cause of many environmental and social
40
41 issues (Jiang & Wong, 2016).
42
43 Corporate Social Responsibility (CSR) is defined as "companies encouraging drive aimed at
44
45 creating awareness about their business among people, environment, and economy, including
46
47 their stakeholders" (Melissen, Mzembe, Idemudia & Novakovic, 2017). Meanwhile, Jamal and
48
49 Karam (2018) defined CSR as a procedure by which a corporation regularly contributes toward
50
51
governance improvisation, ecological conditions, and ethical standards. The concept of CSR was
52 coined by Cadbury (owner of the organization) in the 18th century when he decided to invest
53
54 money in Bernville Cadbury farms for the cultivation of plants. However, it was in the late 19 th
55
56 century when the concept gained popularity, especially in the USA and the European nation.
57
58 Currently, this concept has been renowned worldwide but not fully untapped in developing
59
60 countries (Qazi, Kashif & Qureshi, 2015).
61
62
63 Page 2 of 29
64
65
 The concept of CSR in the corporate sector within the context of Pakistan differs from western
perception. In Pakistan, CSR is taken for granted as most of the business communities are
1
2 focused on minting profit for firms, thus neglecting social welfare, e.g., environment-friendly
3
4 policies, employees, stakeholders, and social welfare. Lack of exposure and giving the least
5
6 importance to procedures mentioned above with the argument of declaring it as a helpful tool to
7
8
attract donation (Qazi et al., 2015; Yunus et al., 2017). However, the Securities Act 2005 and
9 2007 had been enacted by the SECP to control deficiencies for the protection of stakeholders
10
11 (Yunus et al., 2017). In Pakistan, both awareness and implication of CSR are at its embryonic
12
13 stage (Rana & Asad, 2018). However, some firms of Pakistan have published their annual
14
15 sustainability reports, i.e., telecom and banking firms, which genuinely depict their contribution
16
17 to environmental protection, society, and welfare of employees at sites.
18
19 Most organizations, with a few exceptions, profess the entire claim on the profit earned. This
20
21 feature demands extensive debates on NFEH (National Forum for Environment and Health,
22
23 2016). Regarding this essence, this issue cannot be swept under the carpet as maximum foreign
24
25
companies feel reluctant to penetrate the Pakistani market. Developed countries, such as Canada,
26 the UK, the USA, and Germany even developing country like China, have proper systems for
27
28 CSR. Contribution to society is the responsibility of every organization. Thus, it does not impose
29
30 much impact on the profit of an organization. However, in the context of developing countries
31
32 like Pakistan, only several organizations contribute to society, hence viewed as a disadvantage
33
34 when compared to other organizations (NFEH, 2016).
35
36 The role of CSR in CFP (ROA, ROE) has been explored by some studies, but not at all clear
37
38 understanding has been established due to the contradicting outcomes (Sacconi & Degli-Antoni
39
40 2011; Ahmed, Butt, & Majeed, 2018; Zulfiqar et al., 2019). However, the relationships of CSR
41
with CFP (Earning Value) is relatively untapped or with confines in developing countries as
42
43 compare to developed countries especially in Pakistan through the use of the advanced statistical
44
45 technique on longitudinal data. According to the Khattak, Ronchi, Castelli, and Sparks (2014),
46
47 phenomenon of CSR is relatively new in developing countries, thus gaining attention from
48
49 researchers since the past decade.
50
51 Prior study have examined the role of CSR on employee’s behavior and investor’s intentions
52
53 (see Zulfiqar, Sadaf, Popp, Veinhardt, & Máté, 2019). In Pakistan few studies such as Javeed,
54
55 and Lefen (2019) have inspected CSR and profitability (ROA and ROE) by taking moderating
56
57 effect of corporate governance factors CEO power and ownership structure. Rana and Asad,
58 (2018) study limitation was only restricted to manufacturing pharmaceutical sector 9 firms
59
60 financial performance and the four-year observation period in Pakistan. Kiran (2015) study only
61
62
63 Page 3 of 29
64
65
 limited to oil and gas sector 10 firm’s profitability and conducted simple regression in excel by
considering ROA and ROE. Iqbal, Ahmad, Hamad, Bashir, and Sattar (2014) reviewed CSR
1
2 possible impact on profitability narrow to banking sector (sample only UBL bank) by
3
4 considering donation as CSR and financial performance as net profit margin and EPS. There is
5
6 a lack of understating on the need for allocating budget for CSR activities at the corporate level.
7
8
The role of CSR in EV within the context of Pakistan is rather a novel phenomenon and not fully
9 explored with different dimensions (Ahmed, Butt, & Majeed, 2018; Rana & Asad, 2018; Javeed,
10
11 & Lefen, 2019). Part of the corporate segment has been much conventional and has remained
12
13 captive to financial contemplations setting group intrigues in Pakistan.
14
15 Due to interest disparity between the organizational part and society for sustainable
16
17 development, CSR has been disregarded in Pakistan, mainly social wellbeing and training that
18
19 have been given less attention. In the stringent modern and competitive business arena, the
20
21 environmental sustainability of businesses mostly depends on the efficiency and the
22
23 effectiveness of the financial function of its management. However, social and economic
24
25
indicators have been ranked very low in the comity of the nation. The role of CSR in terms of
26 corporate social performance (financial aspect) requires further comprehension and exploration,
27
28 particularly in a developing country overall sectors, such as Pakistan. Consequently, sustainable
29
30 growth is envisioned (Rana & Asad, 2018). Prado et al., (2019) explored future endeavors by
31
32 identifying the existing gaps in literature, used statistical analysis, employed variables or
33
34 methods is existing studies on current state of the corporate social performance and corporate
35
financial performance. They conclude that there is not an overwhelming evidence that would
36
37 lead to a relationship of truth relationship so the direction could be pointed out with latest
38
39 methods. This study is an attempt to complement the existing literature by providing valuable
40
41 insights into the role of CSR in Earning Value (EV). This study, based on CSR activities of firms
42
43 across Pakistan, investigated the impact of CSR on EV in Pakistan.
44
45 The CSR is a complex concept mostly distinct as a vigorous and sometimes voluntary
46
47 contribution of enterprise resources to activities that are aimed at achieving sustainable
48
49 development. This study contributes to extend the economic effect of non-financial disclosure
50
51 in several ways. First, it demonstrates that CSR generates distinct consequences on financial
52 performance by various perspectives, such as CSP reputation index, as well as firms with best,
53
54 mid, and low CSR practices. Thus, this study augments the literature of CSR for developing
55
56 countries and enhances the research of vigorous effects of CSR on EV. Second, it is predicted
57
58 that the impact of CSR on financial limitations is not constant across the flows typically assumed
59
60 in prior studies. As per our knowledge of literature, noteworthy study to report such findings in
61
62
63 Page 4 of 29
64
65
 Pakistan by applying the ADRL method of “Co-integrating” and “Long Run Form,” as well as
the assessment of CSR impact at a different high, moderate and low dimensions level.
1
2
3
Findings contribution have depicted that firms with better CSR have a positive relationship with
4 earning value (EV) in the long run, while too much increased CSR activities of the firm have an
5
6 imperfect CSR process that may harm EV in the short run. Hence, the best CSR company policies
7
8 have a positive contribution to CSR activities. Besides, the middle-level CSR also indicates a
9
10 positive relationship with EV, apart from enhancing financial structure, firm performance, and
11
12 firm prospects. However, poor low level CSR activities has insignificant effect. For robustness
13
14
and diagnostics, as well as to avoid multicollinearity and autocorrelation, multiple tests were
15 performed in this study to validate the outcomes.
16
17
18 The findings revealed that CSR practices are not only concerned with stakeholders’ demands,
19
20
but also augmented the firm’s financial position by enhancing the corporate image in the eyes of
21 society, government, and media. Additionally, CSR practices enhance corporate reputation,
22
23 innovation, and business performance, along with better human resources. Finally, the study
24
25 outcomes have significant inferences, not only for those who can improve investing decisions
26
27 such as investors, but also for regulators and listed firms because maximum CSR actions demand
28
29 financial support, and CSR activities cannot increase business performance within a short period
30
31
with weak strategic framework. In developing countries especially in Pakistan, the literature on
32 CSR and firm performance seems to have evolved slower than in practice.
33
34
35 Section 2 depicts a short review of the literature pertaining to CSR and CFP, as well as their role.
36
37
Section 3 establishes the methodology adopted in this study. Section 4 presents the results, the
38 analysis, and the discussion. Lastly, Section 5 concludes the research study after listing its
39
40 drawbacks and future research direction.
41
42
43
44
45 2. Literature Review
46
47
48 Social responsibility refers to responsibilities that businesses perform to maximize stakeholder
49 values and to offer benefits to society (Carroll, 1999; Okoye, 2009). Social responsibility is
50
51 essential to both consumers and investors, as they have to seek those investments that are not
52
53 only profitable for them, but also contribute to social welfare (Robin & Reidenbach, 1987). The
54
55 main perspective of the social responsibility theory is to make such policies that enable
56
57 businesses to maintain the balance between two mandates, ensuring high profitability, and
58
59
providing benefits to the society (Carroll & Shabana, 2010).
60
61
62
63 Page 5 of 29
64
65
 In a nutshell, “Earning Value is an approach where you monitor the project plan, actual work,
and work completed value to see if a project is on track. Earned Value shows how much of the
1
2 budget and time should have been spent, considering the amount of work done so far”. The
3
4 Earned Value method has been developed as a tool facilitating project progress control. Many
5
6 organizations worldwide adopted Earning Value as a standard management tool (Czernigowska,
7
8
2008). EV analysis is an important tool to measure the performance of a project. The importance
9 of EV is because it not only focuses on planned expenditure and actual cost but also identifies
10
11 future opportunities and analyses actual performance (Suresh, & Ganapathy-Ramasamy, 2015).
12
13
14
Corporate social responsibility (CSR) refers to business operations in favor of society needs; to
15 provide such environment that is safe and healthy for employees and their families by creating
16
17 worthy relationships among employees, management, and stakeholders, so that the organization
18
19 can also achieve its objectives and gain financial competitiveness (Aga, Khan, Wasim & Shah,
20
21 2012). The aspect of CSR is an essential issue in the literature pertaining to management and
22
23 accounting since the 1950s (Akhtar & Awan, 2014). However, organizations and societies have
24
25
significantly enhanced their focus on CSR over the past few years. Companies, with the aim of
26 business expansion, have developed approaches that tend to shift the focus of business operations
27
28 from firms to society. Hence, such market-oriented tactics have been regarded as CSR activities
29
30 by eminent scholars, such as (Melissen, Mzembe, Idemudia, & Novakovic (2017).
31
32 CSR has four theories. The first is an instrumental theory that refers that the corporation acts as
33
34 an instrument for wealth creation, and their social activities objective is to only generate good
35
36 economic outcomes. Second, the political theory is concerned about the power of corporations
37
38 in both the political arena and society. The third theory, integrative theory, concerns the leading
39
40 objective of the corporation, which is to fulfill the societal demands. Lastly, ethical theories refer
41
to the ethical and social responsibilities of the corporation for the society (Maon, Lindgreen, &
42
43 Swaen, 2009; Seele & Lock, 2015). All of the social responsibility theories represent the four
44
45 dimensions; profit, political performance, societal demands, and ethical demands (Okoye, 2009).
46
47 According to Golrida, Subroto, Sutrisno, and Saraswati (2018), the link between financial
48
49 performance (EV) and CSR is significantly positive, as suggested by the stakeholder and
50
51 legitimacy theories, primarily due to discretionary CSR environment that enhances financial
52 performance (EV) and vice versa. Meanwhile, the resource-based view reports that due to
53
54 cultural, social, and political issues in developing countries, firms lack awareness and
55
56 communications of CSR to stakeholders because CSR practices rely on firm resources.
57
58 Many prominent scholars, including Wright and Ferris (1997), found the execution of CSR
59
60 associated with agency problems or signified the interest with shareholders and management.
61
62
63 Page 6 of 29
64
65
 They professed that inappropriate CSR results in a conflict of interest between managers and
stakeholders if a firm is more oriented towards profit maximization when compared to adverse
1
2 environmental impacts. The dispute between managers and stakeholders can be resolved if
3
4 managers adopt the stakeholder’s oriented approach. If managers have a profit-oriented plan,
5
6 conflict is bound to occur with stakeholders. Kim, Kim, and Qain (2018) found a significant
7
8
impact on financial returns of U.S 113 publically listed firms, which oriented its business
9 activities toward the protection of stakeholder’s interests. This leads to a competitive advance of
10
11 the firms as a result of the stability of stakeholders. The theories suggest the positive role of CSR
12
13 in firm performance and firm value.
14
15 The significance of CSR is commended worldwide. Business practices and theories have been
16
17 improved in light of globalization (Yunis et al., 2017). The Institute of Economic Cooperation,
18
19 established in the 19th century to encourage CSR, prescribed a separate budget. Several
20
21 conferences have been organized under the United Nations Environment Program to protect the
22
23 social environment with the aid of social and environmental protection agencies to adopt CSR
24
25
policies for progression to take place in a country. A sharp increase has noted due to globalization
26 in social initiatives undertaken by financial organizations (essay UK, 2013). These emerging
27
28 trends offer flexibility to the corporate sector at both domestic and multinational levels. These
29
30 emerging trends help in shaping the CSR concept in this present world (Kell, 2016). The CSR
31
32 concept is originated from the USA and the UK (Orudzheva & Gaffney, 2018). Particularly in
33
34 the USA, this concept has been much debated in the 1970s and 1980s. In 2009, the Canadian
35
government designated CSR for socio-environmental development. According to Sacconi and
36
37 Degli Antoni (2011), CSR is a good proxy for social capital of firm activities; the prevalent view
38
39 between corporations and practitioners so that firm operations of CSR can generate and more
40
41 social capital. Second, a significant challenge during the financial crisis of 2008-2009 is a public
42
43 trust, which declined unexpectedly in corporations, capital markets, and institutions.
44
45 Muruviwa, Nekhwevha, and Akpan (2018) claimed that CSR improved the economic
46
47 development of society’s living standards in the Zimbabwe for the long term. Pirsch, Gupta, and
48
49 Grau (2007) identified the materialization of CSR activities for stakeholders, which reckoned
50
51 firm achievements by realizing non-financial and financial targets in the interest of stakeholders.
52 In a similar study conducted in the UK, Singh (2014) studied the impact of CSR disclosure on
53
54 firm financial performance. For the designated effects of linear regression, the model unraveled
55
56 the significant impact of CSR on the short- and long-run economic performance of firms. In
57
58 another study that linked CSR with financial performance in Bangladesh banks, Sarwar, Zahid,
59
60 and Ikram (2012) found that return on asset (ROA) was higher for banks that adopted CSR
61
62 policies when compared to those neglecting it.
63 Page 7 of 29
64
65
 Kanwal, Khanam, and Hameed (2013) empirically assessed data retrieved from 15 listed
companies of Karachi Stock Exchange (KSE) found that the corporations were socially
1
2 responsible. A strong financial base is essential for companies that carry out CSR activities. The
3
4 CSR is still at the perception stage in Pakistan (Khattak et al., 2014). Preferably, it should be
5
6 introduced at the academic level to adopt in business organizations across the country. A case
7
8
study in Kasur (a city in Pakistan) showed that an insufficient amount was invested by business
9 organizations in the social sector. Consequently, underground water and soil pollution were
10
11 harmful to the health of society.
12
13
14
Shaheed et al. (2015) found that CSR activities had a positive impact on promoting the firm’s
15 financial performance. Arshad, Anees, and Ullah (2016) studied CSR and CFP in Pakistan using
16
17 panel regression and analyzed the financial performance between 2009 and 2013 of 125 KSE
18
19 listed companies in Pakistan by considering financial performance through ROA and Tobin’s Q
20
21 variables. They reported that CSR was measured as a summation of donation and ecological
22
23 cost, concluding that CSR could increase the aspect of goodwill in businesses. They revealed
24
25
that for short-term scenarios, there was no significant effect of CSR on CFP at a 5% confidence
26 level. At a 10% confidence level, a positive effect was noted for the short-run, while CSR had
27
28 no impact on Tobin’s Q for the selected companies in long- term position. Meanwhile, CSR
29
30 estimated using spending on an element by measuring the reputation index indicated that CSR
31
32 brought a positive change in the financial condition of the firm. The correlation between CSR
33
34 and CFP was studied by Rana and Asad (2018) for 2013-2016 using the panel fixed effect
35
method by considering earning per share, ROA, and ROE proxies that portrayed the significantly
36
37 positive impact of CSR on firm performance. They concluded that a developing country like
38
39 Pakistan, CSR had a positive effect on CFP.
40
41
Mamun, Sohog, and Akhter (2013) applied the ARDL estimation technique to determine the
42
43 relationship between CSR and financial determinants in Bangladesh by involving 30 private
44
45 commercial banks from 2002 to 2011. They confirmed the significant relationship of several
46
47 financial determinants, such as total investments, the number of employees, and the number of
48
49 branches, with CSR expenditure. They concluded that conglomerates are more CSR-oriented
50
51 than small firms, while increment in CSR is a firm commitment to social betterment. Al-Hussaini
52 (2019) assessed the relationship between financial management and environmental Kuznets
53
54 curve by using the ARDL bound test approach in Kuwait from 1981 to 2017 and discovered a
55
56 positive link between financial management and environmental degradation in the long run.
57
58 Economic growth and energy consumption were determined as active contributors.
59
60
61
62
63 Page 8 of 29
64
65
 Ahmed, Butt, and Majeed (2018) studied the level of CSR impact (best, mid, and low CSR firms)
by determining the association between CSR and CFP. It was concluded that firms with poor
1
2 CSR activities displayed an insignificant impact. The best CSR practices by firms had a positive
3
4 role in earning valuue, but this was not evident. However, the effect was integral for CSP
5
6 reputation or intellectual capital and social value. The mid-level CSR firms with excess value
7
8
exhibited a positive link to improve the prospect and stature in terms of earning value.
9
10 Many firms lacked awareness about the concept and implementation of this field because most
11
12 of the firm head or owners considered CSR as merely spending money on donations stemming
13
14
from this unimplemented concept in Pakistan. The CSR is the instrument for commercial
15 accomplishments by means of ethical values, such as working for the wellbeing of communities,
16
17 respecting employees at the workplace, and introducing environment- friendly regulations (Qazi
18
19 et al., 2015). Based on the past literature, theory of the social responsibility theory and stake
20
21 holder theory concept following hypotheses are formulated and need to explore in Pakistan;
22
23 H1: CSR has a significant pivotal influence on firm earning value. Roberts and Dowling, (2002)
24
25 confirmed and suggested the proposed hypothesis relationship.
26
27 H1a: CSR reputation index has a decisive role in earning value in co-integrating form.
28
29 H1b: CSR reputation index has a decisive role earning value in co-integrating form.
30
31
H2: Operating earning assets has significant impact on firm earning value.
32 H2a: Operating assets has significant influence on firm earning value in co-integrating form.
33
34 H2b: Operating assets has negative impact on firm earning value in long run.
35
36 H3: Operating sales has significant impact on firm earning value.
37
38 H3a: Operating sales has positive impact on firm earning value in co-integrating form.
39
40 H3b: Operating sales has positive impact on firm earning value in long run.
41
H4: CSR high practices have a statistically significant affiliation with earning value.
42
43 H4a: High-level CSR practices has significant impact on firm earning value in co-integrating.
44
45 H4b: High-level CSR practices has significant impact on financial earning value in long- run.
46
47 H5: Middle-level CSR has a statistically positive relationship with earning value earning value.
48
49 H5a: Mid-level CSR practices has positive impact on firm earning value in co-integrating form.
50
51 H5b: Mid-level CSR practices has positive impact on firm earning value in long run
52 H6: Worst-level CSR practices has insignificant and negative impact on earning value.
53
54 H6a: Lower CSR practices has insignificant impact on firm earning value in co-integrating form.
55
56 H6b: Lower level CSR practices has insignificant impact on firm earning value in long run.
57
58
59
60
61 3. Research Methodology
62
63 Page 9 of 29
64
65
 Study aim is to explore the role of CSR in firm Earning Value in Pakistan. This study used
secondary data for 2009-2016 and data were gathered from annual financial reports of respective
1
2 organizations. The sample size was limited to 46 firms (see, Appendix C) with 368 total
3
4 observations, mainly because most firms had incomplete and inaccurate CSR data. Given the
5
6 requirements of data, only those firms that report at least two consecutive years of data are
7
8
included by following (Roberts & Dowling, 2002). In Pakistan, financial reporting of CSR
9 activities is still at the infancy stage. Sample include manufacturing; fertilizer; cement building
10
11 material; pharmaceuticals; automobile; consumer goods and foods; petroleum, energy and
12
13 power; textile; information and technological communication companies listed in Stock
14
15 Exchange of Pakistan and had at least hundred million rupees in outstanding shares, because
16
17 larger companies generally publish sustainability reports on a regular basis as well regularly
18
19
participate more in CSR practices as compare to small firms. In Pakistan the significant
20 contribution and implications of these companies CSR activities are health and education
21
22 improvement, reduction in poverty, local society’s infrastructure development, technical and
23
24 vocational trainings, assistance in natural disasters emergency and human resource development
25
26 through social welfare programs.
27
28 The reputation index method is the first, whereby firms are rated by multiple dimensions of
29
30 social performance. The index has the advantage of equal treatment of each firm and easy to
31
32 compare. Another benefit is that it makes no pretense of implementing explicit purpose that
33
34 quantifies a measurement that can be skewed. In the 1960s, the Council of Economic Priorities
35
(CEP) used the first reputation index. The initial reputation index method was employed in many
36
37 studies, including Marlin (1972), Folger and Nutt (1975), and Spicer (1978). In 1972, Moskowitz
38
39 established another reputation index that was further modified by Beresford (1976). The CSR
40
41 reputation index was developed based on the firm’s ratings for multiple dimensions of social
42
43 performance. Carroll (1999) categorized a 3D model by incorporating philosophies of social
44
45 issues. Followed Robert and Dowling, (2002) reputation index methodology where each firm’s
46 overall reputation score from firms. The reputation index measurement of CSR had been based
47
48 on environment (EVN), employee relations (ER), product quality and its relationship with
49
50 customers and providers (PRD), shareholder relations (SHA), and community (COM). The
51
52 annual disbursement on CSR activities served as a proxy of CSR practices in firms based on
53
54 annual reports of the firm. These proxies determined the CFP (Cochran et al., 1984; Roberts &
55
56
Dowling, 2002).
57
58 This study further consider the ranking issued by ESG Ranking of Pakistan sample organizations
59
60 into the lowest, best, and middle CSR firms based on the ratings by using dummy zero and one.
61
62 Data of three proxies were used to measure financial condition as the accounting data discards
63 Page 10 of 29
64
65
 data distortion. The variables MVE represent market value of equity, TA indicate total assets,
and BVD represent book value of debt. Therefore, EV measured from (MVE +BVD) minus TA
1
2 and then divided by total sales. The earning access value was used to evaluate the future aspect
3
4 of firms. Increment in EV indicated increased financial performance and enhanced CSR.
5
6 Operating income to assets was measured from the proxy of operating income divided by the
7
8
total assets. Operating income to sales proxy of operating income divided by the total sales was
9 measured. The ratio of operating income to sale considered a proxy in empirical research based
10
11 on leverage variances. Finally, operating income from assets was used to examine the revenue
12
13 performance of firms.
14
15 There are multiple regression techniques, such as simple regression, fixed or random effect. But
16
17 this study applied the Auto Regressive Distributed Lag (ARDL) approach proposed by Pasaran
18
19 and Shin (1999) with cointegration as this approach applied and suggested by recent studies.
20
21 Such as Ghouse, Khan, and Rehman (2018) have expressed that one of the most general dynamic
22
23 unrestricted models in the econometric literature is the ARDL model, primarily because the
24
25
ARDL co-integration approach offers specific tests to determine the presence of a single co-
26 integrating vector, instead of assuming uniqueness. Pre-conditions are necessary, such as
27
28 running the ARDL and ensuring that the model “Dependent” is non-stationary (Hussain &
29
30 Mazhar, 2018). In order to capture the data generating process, the ARDL model takes a
31
32 sufficient number of lags in general to a specific modeling framework because this model can
33
34 capture the correlation of the co-integrated variables for both short- and long-run (Shrestha &
35
Bhatta, 2018). ARDL approach is the latest and has been widely accepted due to its numerous
36
37 significant advantages, such as its applicability in all cases, regardless of series I (0), I(1), or
38
39 mutually co-integrated variables (Oskooee, & Oyolola, 2007). The second advantage of this
40
41 contemporary approach is that it can obtain a more comprehensive array of lags. Notably, in a
42
43 more common-to-particular method of the ARDL technique, a wider variety of lag numbers are
44
45 apprehended in the data to create the procedure. Based on the lag method, the ARDL estimation
46 assesses the co-integration of data. For a long-run relationship tested via the ARDL model, two
47
48 steps are embedded. In the first step, the presence of the long-run is determined. The second step
49
50 then estimates both short- and long-run coefficients of the same equation (McCann, Baylis, &
51
52 Williams, 2010).
53
54 Sulaiman, Baharin, and AlHadi (2019) reported the concept of a single reduced equation utilized
55
56 in the ARDL model. This generates outcome in a more comprehensive, detailed, reliable, and
57
58 effective method than any other method of traditional integration to determine the relationship
59
60 of both long- and short-run. It can test any fixed relationship between the variables without taking
61
62 into contemplation of its capability to underlying regression values, which are purely I (0) and I
63 Page 11 of 29
64
65
 (1) correspondingly or both. In order to discover different and similar logs from several variables
in the dataset, the ARDL approach makes it possible. Thus, the ARDL is an appropriate method
1
2 for this study that looked into CSR and firm Earning Value (financial performance). The ARDL
3
4 approach was suggested and employed by Mamun, Sohog, and Akhter (2013) in Bangladesh to
5
6 confirm the long-run relationship of CSR expenditure with firm financial determinants.
7
8 This study also applied the bound test co-integration for both long- and short-run relations.
9
10 Hussain and Mazhar (2018) reported that the use of the co-integration technique for estimation
11
12 avoids spurious regression and endogeneity problems. The first difference (I) is the number of
13
14
lag and (t-1) is the level lag of variables. The Akaike’s Information Criterion (AIC) and some
15 other diagnostic tests were employed to determine the lag length for each variable, as well as to
16
17 identify the presence of a co-integrating relationship. Next, the bound test was performed after
18
19 identifying an appropriate lag length. Davidescu (2015) applied the ARDL co-integration
20
21 approach to investigate the long-run causal linkages and the short-run dynamics by stating that
22
23 the ARDL error correction representation becomes relatively more efficient with a single long-
24
25
run relationship and small sample data size.
26
27 This study employed robustness tests, including Breusch- Godfrey Serial Correlation LM Test,
28
29 Heteroscedasticity Test Breusch-Pagan-Godfrey, Ramsey RESET Test, and Variance Inflation
30
31
Factors for diagnostics. Shrestha and Bhatta (2018) expressed that for the residual diagnostic
32 LM test (Lagrange multiplier), heteroscedasticity and correlogram tests are the primary testing
33
34 approaches for ARDL.
35
36
37
In determining the long-run relationship between CSR, and financial performance (EV) activities
38 on varied variables, the ARDL approach was adopted for this study. It is composed of two steps.
39
40 The first model presents the conceptual model verified by the ARDL method to identify the co-
41
42 integrating form. Second, the long-run coefficients relationship was determined by adding lags
43
44 of dependent and independent variables. On the basis of the above explanation and data, the
45
46 econometric model is as follows;
47
48 Equation 1 is as follows;
49
50 𝐸𝑉𝑖,𝑡 = β0 + ∑∞ ∞ ∞ ∞
𝑛=1 β1 (EVt−i ) + ∑𝑛=1 β2 (CSPRIt−i ) + ∑𝑛=1 β3 (OIAt−i ) + ∑𝑛=1 β4 (OESt−i ) + μi,t (1)
51
52
53 Decomposed CSR level Equation 2 is given below;
54
55 𝐸𝑉𝑖,𝑡 = β0 + ∑∞ ∞
𝑛=1 β1 (EVt−i ) + ∑𝑛=1 β2 (HCSR t−i ) +
56
57 ∑∞ ∞ ∞ ∞
𝑛=1 β3 (LCSR t−i ) + ∑𝑛=1 β4 (MCSR t−i ) + ∑𝑛=1 β5 (OIAt−i ) + ∑𝑛=1 β6 (OISt−i ) + μi,t (2)
58
59 Where, explained as EV represents earning value, CSPRI is corporate social performance
60
61 reputation index, OIA denotes operating income assets, OIS signifies operating income sales,
62
63 Page 12 of 29
64
65
 while HCSR, LCSR, and MCSR indicate the highest, the lowest, and the mid-CSR levels,
respectively. B0 is the constant value, whereas µ serves as an error term.
1
2
3 4. Results and Discussion
4
5 4.1. Model 1 Results CSR and Earning Value (Equation 1)
6
7 Table 1 presents the effects of the ARDL model to determine the relationships of variable EV
8
9 with CSP Reputation Index, Operating Income Asset, and Operating income Sale. The adjusted
10
11 sample variables included 368, while after adjustments, 366 observations were applied. The
12
13 automatic selection of four maximum dependent lags and an appropriate model were chosen
14
15 based on AIC. The results of the ARDL model showed that the R-squared and the adjusted R-
16
17 squared results were 0.447 and 0.437, respectively. The value of F-statistic showed that the
18 model was significant at 48.39, while the Durbin-Watson results showed no autocorrelation in
19
20 the data. The results showed that EV had significant lag effects, and its lag was substantial. At
21
22 lag 2, it turned significantly positive at a 55.7% significance level. The results support
23
24 Hypothesis 1 and the study by Rana and Asad (2018). Table 1 tabulates the detailed results.
25
26 The CSP Reputation Index was insignificant, with the negative sign, but at lag 1 it is significantly
27
28 positive with a 1.3% significance level, which endorse our alternate hypothesis (H1) and is in
29
30 line with the study by Arshad, Anees, and Ullah (2016). Operating income Asset had an
31
32 insignificantly negative impact which accept the null hypothesis and reject alternate (H2), while
33
34 Operating Income Sale demonstrate 14.7% significance level and accept the third hypotheses
35 (H3) based on operating sales has significant impact on firm performance.
36
37
38 [Table 1 near here]
39
40 Table 2 presents the results of the ARDL bound test, which indicated that the null hypothesis
41 had no long term relationship. Based on the variables with I (0) or I (1) assumption, the
42
43 ARDL bounds test was conducted. Harris and Sollis (2003) claimed that by applying the ARDL
44
45 technique of bound test, one could attain long-run model unbiased estimates. The bounds test is
46
47 mainly based on the joint F-statistic under the null hypothesis of no co-integration. Since its
48
49 asymptotic distribution was non- standard and based on the assumption that all the variables in
50
51
the ARDL model were integrated of order zero, the first level was calculated. While the variables
52 were integrated of order, one was based on assumption, while the second was calculated; thus,
53
54 no co-integration of null hypothesis is rejected. Our test statistics exceeded the upper critical
55
56 bounds value, and it has been accepted with the view that F-statistic is less than the lower bounds
57
58 value. Table 2 shows the outcomes.
59
60 [Table 2 near here]
61
62
63 Page 13 of 29
64
65
 Table 3 tabulates the results of co-integrating and long-run effects of dependent and independent
variables. The cointegration ARDL model was used in this study to determine that the
1
2 correlations of variable EV with CSP Reputation Index, Operating Income Assets, and Operating
3
4 Income to Sale ratio variables. The results of co-integration showed a negative value of earning
5
6 at lag 2 therefore reject alternate hypothesis (H1a) and accept null hypothesis (H0a) of co-
7
8
integrating form. The value of the corporate social performance index was 1.5%, with a
9 significantly negative impact. The results are similar to those reported by Arshad, Anees, and
10
11 Ullah (2016) using panel regression, which revealed that in short-term scenarios, there was no
12
13 significant impact of CSR on financial performance at 5% confidence level. The ratio of
14
15 operating earnings displayed a 9% negative significance level. Operating earnings to sale ratio
16
17 exerted a significantly positive impact with a 14.2% significance level and had a significantly
18
19
positive impact on EV and corporate social activities in the long term. Therefore, endorse the
20 acceptance of alternate hypothesis (H3a) in both co-integrating and long run that operating sales
21
22 has positive impact on firm performance. Corporate social performance index exhibited a
23
24 considerable role with EV, and in the long run, the CSP index showed a 1.5% slightly
25
26 insignificant level therefore in both co-integrating and long run accept null hypothesis. Both EV
27
28 and operating income to assets ratio had a decisive role, and the study results showed negative
29 value therefore we reject the alternate hypothesis and accept null hypothesis. The constant value
30
31 reflected a significantly positive impact on model variables.
32
33
34 [Table 3 near here]
35
36 For diagnostics and to avoid both multicollinearity and autocorrelation, multiple tests, such as
37
38 Breusch-Godfrey Serial Correlation and LM Test, were performed to analyze the correlations
39
40 between the variables. The results of the Breusch–Godfrey LM test displayed no evidence of
41
autocorrelation in residuals. Besides, the model had no heteroscedasticity, as indicated by the
42
43 results of the Breusch-Pagan-Godfrey test with an F-statistic value of 2.1342 and Probability at
44
45 0.048. The result of the Ramsey RESET test t-value is 1.310 and F-statistics 1.718 fitted values.
46
47 The complete robustness test results of equation 1 are demonstrated in appendix A.
48
49
50 4.2. Model 2 Results of CSR Levels
51
52
53 Table 4 shows the complete results of the second equation using the ARDL model for CSR levels
54
55 and other variables. The best model was selected based on the AIC by using the automatic
56
57 selection of maximum dependent lags 4. The results showed that the R-squared and adjusted R-
58
59
squared indicated the goodness of fit at 0.512 and 0.495. The results of F-statistics demonstrate
60 that the model was significant at 30.580, while the Durbin-Watson results showed no auto-
61
62
63 Page 14 of 29
64
65
 correlation in the data. The results showed that EV had significant lag effects. At lag 2, it was
significantly positive with a 13.5% significance level. Therefore, the CSR had a significant effect
1
2 on EV when we measure at levels decomposed model and accept the alternate hypothesis (H1)
3
4 of CSR activities has significant impact on firm performance. The outcomes are consistent with
5
6 the findings reported by Kim, Kim, and Qain (2018), which led to a competitive advance of the
7
8
firms as a result of the stability of stakeholders. Therefore, the theoretical literature suggests that
9 an increase in CSR led to betterment in the long run.
10
11
12 The highest CSR was significantly positive at 32.09% at initial which endorse the alternate
13
14
hypotheses that, but it became insignificant with a negative sign at lags 1 and 2 with 12.9% and
15 9% insignificance level which indicate that too much high investment on CSR activities cause
16
17 the loss in the financials statement. The lowest CSR level was significant with a positive sign at
18
19 a 27.9% significance level in initial, but at lag 1, it turned insignificant with a negative sign. The
20
21 worst/lowest CSR firms explained that the CSR affected EV with decreased CSR activities. The
22
23 results of the ARDL method confirmed that lowest CSR activities had a negative effect on worst-
24
25
CSR firms and endorse the proposed alternate hypothesis (H6) that worst-level of CSR practices
26 has insignificant and negative impact on firm earning value in terms of financial performance.
27
28
29 The mid-level CSR results showed a significantly positive impact on the significance level of
30
31
7% at zero lag. Corporates with mid-level CSR had a significant role in financial performance
32 with EV. This is in line with existing literature which claimed that CSR has a positive impact on
33
34 EV, thus supporting the study hypothesis (H5) that middle-level CSR has a statistically
35
36 significant and positive relationship with firm earning value. The results are consistent with Kim,
37
38 Kim, and Qain (2018), who claimed that all theoretical results in favor of CSR suggested an
39
40 increase in activities of CSR to bring positive change in productivity firm earning value. The
41
Operating income Asset had an insignificantly negative impact at lag 1 and 2 which reject the
42
43 alternate hypothesis (H2) of CSR levels and accept null hypothesis that operating income do not
44
45 have positive effect on earnings in sample firms of Pakistan. The Operating Income Sale
46
47 demonstrated a significantly positive impact at a 34% significance level which endorse alternate
48
49 hypothesis (H3), but it turned negative at lag 1. Table 4 above shows the results in detail.
50
51
52 [Table 4 near here]
53
54
55 Following the equation 2, the bound test results are demonstrated in Table 5. This test is
56
57 relatively more efficient for small sample sizes like in Pakistan, the CSR concept is not fully
58
59
implemented in organizations. Therefore, the sample was collected on the basis of available data
60 of 46 firms, irrespective of the regressors in the model. Table 5 shows that the null hypothesis
61
62
63 Page 15 of 29
64
65
 has no long term relationship because the F-statistics is lower than the lower bounds values.
Because, when the independent variables have a lower value, one can assume that the regressors
1
2 are I (0), while higher value leads to the assumption of I (1). The null hypothesis of no long-run
3
4 relationship can be rejected irrespective of the orders of integration if the F- statistics are higher
5
6 than the upper critical value. Results depicted in table 5 below.
7
8
9 [Table 5 near here]
10
11
12 Table 6 demonstrates the results of the co-integrating and long-run effect for the second model
13
14 of the dependent and independent variables. In the second equation, the relationships of
15
dependent variable EV with CSR levels, Operating income Asset, and Operating income to Sale
16
17 ratio variables were considered. The result of co-integrating displayed a negative value for EV
18
19 at lag 1. Highest level corporate social responsibility at zero and first lags had been significantly
20
21 positive at 32% and 9% significance level, respectively which indicate that results accept the
22
23 alternate hypothesis (H4a) that high level CSR practices has significant and positive impact on
24
25 earning value in co-integrating form. The lowest CSR was substantially positive at a 28%
26 significance level at lag 1 in co-integrating form. Mid-level CSR results in a co-integrating form
27
28 were positive, with 9% zero lag. The ratio of operating income to assets showed a 25% positive
29
30 significance level therefore accept the alternate hypothesis (H2) that operating income assets has
31
32 significant impact on firm performance, while the operating income to sale ratio had a negative
33
34 impact on co-integrating form.
35
36 In the long run, CSR had a considerably significant role in EV and accept the hypothesis H1 of
37 CSR levels has significant role in financial performance. The results demonstrated that in the
38
39 long run, the highest CSR exhibited a 29% positively significance level, and the results support
40
41 Hypothesis (H4b). Meanwhile, the lowest and mid-level CSR had 39% and 21% significance
42
43 level. Therefore, accept the alternate hypothesis of Mid-level CSR while reject the proposed
44
45 alternate hypothesis in model 2 of long run. Operating income to sales positively and
46
47
significantly contributed to EV and corporate social activities in the long run and the hypothesis
48 (H3b) of operating sales has positive impact on firm earning value in long run. Both EV and
49
50 operating income to assets ratio had a decisive role, and this study showed a positive value for
51
52 the long run in equation 2 and endorse the hypothesis (H2b) that operating assets has negative
53
54 impact on firm performance in long run. The results refer to the theory in which a corporation
55
56 acts as an instrument for wealth creation, and the objective of their social activities is to generate
57
58
good economic outcomes only (Seele & Lock, 2015). The overall results of equation 2 showed
59 a significantly positive association with firm CSR levels and earning value but with negative
60
61 coefficient value which is in line with a recent study by Golrida et al. (2018) based on
62
63 Page 16 of 29
64
65
 stakeholder and legitimacy theories because discretionary CSR environment enhances earning
value in term of financial performance and vice versa. However, mixed results with the proposed
1
2 CSR levels hypothesis as accept alternate of High and Mid-level CSR but worst CSR level and
3
4 firm earning value therefore accept H6a, H6b null hypothesis. The findings highlighted the
5
6 resource-based view that due to cultural, social, and political issues, CSR practices relied on firm
7
8
resources. Hence, developing countries need awareness so that their stakeholders can promote
9 and invest in CSR activities. Table 6 shows the complete results.
10
11
12 [Table 6 near here]
13
14 In equation two, multiple tests were performed for diagnostics and autocorrelation, as well as to
15
16 avoid multicollinearity. The LM test was conducted to analyze correlation, and for this, the
17
18 Breusch-Godfrey LM test was employed for serial correlation analysis between the variables,
19
20
while for equation two, the results of Breusch–Godfrey LM test showed no evidence of auto-
21 correlation (see Appendix B). The robustness test of heteroscedasticity results showed no
22
23 heteroscedasticity in the model, as demonstrated in the results of the Breusch-Pagan- Godfrey
24
25 test with F-statistics value at 1.439 and probability at 0.145(see Appendix B). The level of
26
27 collinearity between the regressors in the equation can be measured efficiently using variance
28
29 inflation factors (VIFs) test because this test showed that due to collinearity with the other
30
31
regressors, the variance of a coefficient estimate of a regressor had inflated. Another test Ramsey
32 RESET Test is conducted for the robustness to determine the untitled relationship of variables.
33
34 The complete robustness test results are given in appendix B.
35
36
37
4.3. Graphical Presentation of CSR Levels
38
39 The graph shows that the EV had significant relationships with CSR, whereas CSR had a positive
40
41 role in the financial performance of firms. Firms with the best CSR had a positive trend with EV
42
43
while lower level CSR demonstrate worst trend. Mid-level CSR demonstrate moderate effect
44 therefore the show the balanced positive point in the mid of figure. In this section, graphical
45
46 analysis is used to describe the financial indicators of the firm.
47
48
49
[Figure 1-3 near here]
50
51
52 5. Conclusion and Recommendations
53
54 The role of CSR is important in every society, especially in developing countries. This study
55
56 endeavors to unearth the viewpoint of the impact of CSR on the earning value of firms in
57
58 Pakistan. In both model and co-integrating as well long run results showed mixed statistics.
59 Summary of model one ARDL results report that EV had significant lag effects, and its lag was
60
61 substantial while at lag 2, it turned significantly positive at a significance level which support
62
63 Page 17 of 29
64
65
 first hypothesis. The CSP Reputation Index was insignificant, with the negative sign, but at lag
1 it is significantly positive with a 1.3% significance level, which endorse our alternate
1
2 hypothesis. Operating sales has positive impact on earning value in both co-integrating and long
3
4 run and accept the proposed alternate hypothesis. In model one Co-integrating and long run
5
6 results of corporate social performance index exhibited a considerable role with EV, and in the
7
8
long run, the CSP index showed a 1.5% slightly insignificant level therefore in both co-
9 integrating and long run accept null hypothesis.
10
11
12 Model two of second equation using the CSR levels summary showed that EV had significant
13
14
lag effects. At lag 2, it was significantly positive with a 13.5% significance level. Therefore the
15 CSR had a significant effect on EV when we measure at levels decomposed model and accept
16
17 the alternate hypothesis (H1) of CSR activities has significant impact on earning value. The mid-
18
19 level CSR results showed a significantly positive impact on the significance level of 7% at zero
20
21 lag and this is in line with existing literature which claimed that CSR has a positive impact on
22
23 EV, thus supporting the study hypothesis (H5). The highest CSR was significantly positive at
24
25
32.09% at initial which endorse the alternate hypotheses that, but it became insignificant with a
26 negative sign at lags 1 and 2 with 12.9% and 9% insignificance level which indicate that too
27
28 much high investment on CSR activities cause the loss in the financials statement. The lowest
29
30 CSR level turned insignificant with a negative sign and worst/lowest CSR firms explained that
31
32 the CSR affected EV with decreased CSR activities which endorse the proposed alternate
33
34 hypothesis. The Operating income Asset had an insignificantly negative impact at lag 1 and 2
35
which reject the alternate hypothesis (H2) while the operating income sale demonstrated a
36
37 significantly positive impact at a 34% significance level which endorse alternate hypothesis (H3),
38
39 in model two.
40
41
For second model the result of co-integrating displayed a negative value for EV at lag 1 while in
42
43 the long run, CSR had a considerably significant role in EV and accept the hypothesis H1 of CSR
44
45 levels has significant role in earning value. Highest level CSR had been significantly positive at
46
47 significance level in both Co-integrating and long run, which support the alternate hypothesis
48
49 (H4a and H4b). The lowest CSR was substantially positive in co-integrating form while significant
50
51 positive in long run which endorse the null hypothesis of H5a and H5b. Mid-level CSR results in
52 a co-integrating form and long run were positive and support the alternate hypothesis H 5a and
53
54 H5b. In Model two operating income to assets showed a positive significance level therefore
55
56 accept the alternate hypothesis (H2, H2a, H2b) in both co-integrating and long run while the
57
58 operating income had mixed results of positive impact on co-integrating form and via-a-via in
59
60 long run. The findings highlighted the resource-based view that due to cultural, social, and
61
62
63 Page 18 of 29
64
65
 political issues, CSR practices relied on firm resources. Hence, developing countries need
awareness so that their stakeholders can promote and invest in CSR activities.
1
2
3
The study outcomes conclude that firms with CSR at the initial stage would hurt their earning
4 value, as increased CSR activities reduced the revenue of the firms. The findings established that
5
6 in Pakistan, firms with best CSR had a positive relationship with EV. There was an adverse effect
7
8 on the firm’s earning value upon worst CSR activities due to the imperfect CSR process in first
9
10 model. Firms with mid-level CSR displayed a positive relationship with EV in both co-
11
12 integrating form and long run, besides enhancing their financial structure, project/ firm
13
14
performance, and firm prospects. The findings suggest stakeholder’s theory recommendations
15 that CSR performance and moderate/balanced activities maximizes stakeholder value. In
16
17 literature observe fact that in most developing countries such as Pakistan, firms hesitate to
18
19 enthusiastically participate in CSR activities due to fear that such practices may not stimulate
20
21 the performance or profitability of the firm. This effect has more worth in service oriented sector
22
23 especially in banks where regulations are strict by involving large investors and more managerial
24
25
skills.
26
27 By incorporating social responsibility in project management, managers can ensure the
28
29 credibility, integrity, and the reputation of the organization. To achieve the higher standard of
30
31
CSR and for the successful implementation of CSR practices, managers have to line up the
32 project requirements or commitments with their business operations, along with personal
33
34 objectives. If the CSR is incorporated in project management, then the project manager has to
35
36 perform particular tasks, such as following only those processes to avoid harming the
37
38 environment and society, selection of industrial venues that are far from residential areas, protect
39
40 the natural resources, participate in social work, as well as adhere to legal rules and regulations.
41
42 Government can promote and take advantage of companies CSR activates by introducing
43
44 specialized programs to participate in poverty reduction and for a better society development. It
45
46 is imperative that the government should legislate on CSR procedures to offer relief and
47
protection to welfare-oriented organizations. Policymakers and the government should improve
48
49 information transparency and the regulatory framework by considering weak social performance
50
51 by firms from less developed regions. A comprehensive-ranging set of guidelines and plans
52
53 concerning the corporation’s duties and accountability to the society exist in socially-responsible
54
55 companies, whereby they should combine with their corporate working and decision-making
56
57 procedures. However, socially-irresponsible corporations need to act as responsible affiliates of
58 the community, in which CSR practices should be included in the planning, controlling, and
59
60 decision-making process of the firm. Firms should accomplish their social responsibilities in a
61
62
63 Page 19 of 29
64
65
 way as one who is obliged to his family. The globalization can support a business to enhance
corporate social performance, although this is not expected to occur. It is hoped that the
1
2 upcoming review will investigate the probable repetitive nature of the connection between CSR
3
4 and internationalization in the future study. Future studies also may look into cross-country
5
6 comparison involving both developed and developing states, apart from expanding data from
7
8
multiple countries with globalization effect.
9
10
11
12 Reference
13
14 Aga, G., Khan, S., Wasim, D. & Shah, A. (2012). The Impact of Corporate Social Responsibility
15 on The Company’s Financial Performance: A Study of Pharmaceuticals Firms of Peshawar
16
17
Pakistan. City University Research Journal, Volume 03.
18
19 Ahmed, T., Butt, B. Z. & Majeed, W. K. (2018). The Impact of Corporate Social Responsibility
20 on Firm’s Financial Performance. Information Management and Business Review (IMBR),
21 10 (2) ISSN 2220-3796.
22
23 Akhtar. N. & Awan.G.A. (2014). The impact of corporate social responsibility (CSR) on
24
25
profitability firms: A case study of fertilizer & Cement industry in Southern Punjab,
26 Pakistan. International Journal of Development and Economic Sustainability, Vol: 2 (4),
27 pp.70-79.
28
29 Al-Hussaini, A. N. (2019). The Role of Financial Management in Testing Environmental
30
31
Kuznets Curve in Kuwait: Evidence from ARDL Bound Testing Approach. International
32 Journal of Energy Economics and Policy, 9(3), 353-359.
33
34 Arshad, G. M, Anees, F. & Ullah, R. M. (2015). The Impact of Corporate Social Responsibility
35 on Firm’s Financial Performance. International Journal of Linguistics, Social and Natural
36 Sciences. Inter J Ling Soc Nat Sci, Vol: 1(1): pp. 33-39.
37
38
39
Baumann-Pauly, D., Wickert, C., Spence, L. J., & Scherer, A. G. (2013). Organizing corporate
40 social responsibility in small and large firms: Size matters. Journal of Business Ethics,
41 115(4), 693-705.
42
43 Beresford, D. R. (1976). Social responsibility disclosure—1970s survey of Fortune 500 annual
44 reports. Ernst and Ernst.
45
46
Carroll, A. B. (1999). Corporate social responsibility: Evolution of a definitional construct.
47
48 Business & society, 38(3), 268-295.
49
50 Carroll, A. B., & Shabana, K. M. (2010). The business case for corporate social responsibility:
51 A review of concepts, research, and practice. International journal of management reviews,
52 12(1), 85-105.
53
54
Cochran. L. P. & Wood.A.R. (1984). Corporate Social Responsibility and Financial
55
56 Performance. Academy of Management Journal, Vol: 27 (1), pp.42-56. 43
57
58 Czernigowska, A. (2008). Earned value method as a tool for project control. Budownictwo i
59 Architektura, 3, 15-32.
60
61
62
63 Page 20 of 29
64
65
 Davidescu, A. A. (2015). The Relationship between Shadow Economy and Unemployment Rate.
An ARDL Causality Analysis for the Case of Romania. Romanian Statistical Review, pp.
1 46 – 62.
2
3 Degli Antoni, G., & Sacconi, L. (2011). Modeling cognitive social capital and corporate social
4 responsibility as preconditions for sustainable networks of relations. In Social Capital,
5
6 Corporate Social Responsibility, Economic Behaviour and Performance (pp. 161-239).
7 Palgrave Macmillan, London.
8
9 Folger, H. R., & Nutt, F. (1975). A note on social responsibility and stock valuation. Academy
10 of Management Journal, Vol: 18, 155-159.
11
12 Ghouse, G., Khan, S. A. & Rehman, A. U. (2018). ARDL model as a remedy for spurious
13
14 regression: problems, performance, and prospectus. Munich Personal RePEc Archive,
15 Paper No. 83973, https://mpra.ub.uni-muenchen.de/83973/
16
17 Golrida K. P., Subroto, B., Sutrisno, Saraswati, E. (2018). The Complexity of Relationship
18 between Corporate Social Responsibility (CSR) and Financial Performance. Emerging
19
20 Markets Journal. 8(2). ISSN 2158-8708 (online), DOI 10.5195/emaj.2018.155.
21
22 Harris, R. & Sollis, R. (2003). Applied Time Series Modelling and Forecasting. Wiley, West
23 Sussex.
24
25 Hussain, S. I. & Mazhar, U. (2018). Role of Domestic Demand in Development of Export Supply
26 from Pakistan: An ARDL Approach. Pakistan Journal of Applied Economics: Special
27
28 Issue, pp. 151-169.
29
30 Iqbal, N., Ahmad, N., Hamad, N., Bashir, S. & Sattar, W., (2014). Corporate Social
31 Responsibility and Its Possible Impact on Firm's Financial Performance in Banking Sector
32 of Pakistan. Arabian Journal of Business and Management Review (OMAN Chapter),
33
34
3(12), p.150.
35
36 Jamali, D., & Karam, C. (2018). Corporate social responsibility in developing countries as an
37 emerging field of study. International Journal of Management Reviews, 20(1), 32-61. 7
38
39 Javeed, S.A. & Lefen, L. (2019). An analysis of corporate social responsibility and firm
40 performance with moderating effects of CEO power and ownership structure: A case study
41
42
of the manufacturing sector of Pakistan. Sustainability, 11(1), p.248.
43
44 Jiang, W., & Wong, J. K. (2016). Key activity areas of corporate social responsibility (CSR) in
45 the construction industry: a study of China. Journal of cleaner production, 113, 850-860.
46
47 Kanwal, M., Khanam, F., Nasreen, S., & Hameed, S. (2013).The impact of corporate social
48 responsibility for the firm's financial performance. 35
49
50 Kell, G. (2016). Everyone benefits from a better world: The role of business in society (1st ed.,
51
52
pp. 4–5). UN: United Nations Global Compact International Yearbook.
53
54 Kiran, S., (2015). Corporate social responsibility and firm profitability: A case of oil and gas
55 sector of Pakistan. City University Research Journal, 5(1).
56
57 Khattak, F., Ronchi, A., Castelli, P. &Sparks, N. (2014). Effects of a natural blend of essential
58 oil on growth performance, blood biochemistry, cecal morphology, and carcass quality of
59
60
broiler chickens. doi.org/ 10.3382/ps.2013-03387.
61
62
63 Page 21 of 29
64
65
 Kim, K. H., Kim, M., & Qian, C. (2018). Effects of corporate social responsibility on corporate
financial performance: A competitive-action perspective. Journal of Management, 44(3),
1 1097-1118.
2
3 Mamun, M. A., Sohog, K. & Akhter, A. (2013). A Dynamic Panel Analysis of the Financial
4 Determinants of CSR in the Bangladeshi Banking Industry. Asian Economic and Financial
5
6 Review, Asian Economic and Social Society, vol. 3(5), pages 560-578.
7
8 Maon, F., Lindgreen, A., & Swaen, V. (2009). Designing and implementing corporate social
9 responsibility: An integrative framework grounded in theory and practice. Journal of
10 Business Ethics, 87(1), 71-89.
11
12 Melissen, F., Mzembe, A. N., Idemudia, U., & Novakovic, Y. (2017). Institutional Antecedents
13
14 of the Corporate Social Responsibility Narrative in the Developing World Context:
15 Implications for Sustainable Development. Business strategy and the environment.
16
17 Moskowitz, M. (1972). Choosing socially responsible stocks. Business and Society Review, 1,
18 71-75.
19
20 Muruviwa, A. T., Nekhwevha, F. H., & Akpan, W. (2018). Corporate social responsibility as a
21
22 drive to community development and poverty reduction: A stakeholder approach to
23 development in Zimbabwe. TD: The Journal for Transdisciplinary Research in Southern
24 Africa, Vol: 14(1), pp. 1-9.
25
26 NFEH. (2016). CSR in Pakistan the good the bad and the ugly. 8th international Summit Award,
27
28 Vol 1.
29
30 Okoye, A. (2009). Theorizing corporate social responsibility as an essentially contested concept:
31 is a definition necessary? Journal of Business Ethics, 89(4), 613-627.
32
33 Orudzheva, L., & Gaffney, N. (2018). Country-of-origin and CSR initiatives: a social dominance
34 perspective. Social Responsibility Journal, Vol: 14(3), 501-515.
35
36 Bahmani-Oskooee, M., & Oyolola, M. (2007). Export growth and output growth: an application of bounds
37 testing approach. Journal of Economics and Finance, 31(1), 1-11.
38
39 Pesaran, M.H. & Shin, Y. (1999). An Autoregressive Distributed Lag Modelling Approach to
40
41
Cointegration Analysis. Econometrics and Economic Theory in the 20th Century: The
42 Ragnar Frisch Centennial Symposium, Strom, S. (ed.) Cambridge University Press.
43
44 Pimpa, N., & Moore, T. (Eds.). (2018). Corporate Social Responsibility and the Inclusivity of
45 Women in the Mining Industry: Emerging Research and Opportunities, Emerging
46
Research, and Opportunities. IGI Global.
47
48
49
Pirsch, J., Gupta, S., & Grau. S.L. (2007). A framework for understanding corporate social
50 responsibility programs as a continuum: An exploratory study. J Busin Ethic, Vol: 70, pp.
51 125-140.
52
53 Prado, G. F. D., Piekarski, C. M., Luz, L. M. D., Souza, J. T. D., Salvador, R., & Francisco, A.
54 C. D. (2019). Sustainable development and economic performance: Gaps andtrends for
55
56 future research. Sustainable Development. 28, pp. 368–
57 384. https://doi.org/10.1002/sd.1982
58
59 Qazi. W.S, Ahmed. M, Kashif. S & Ali Qureshi. A.Z. (2015). Company's financial performance
60 & CSR: Pakistan Context. Global Advanced Research Journal of Management and
61
62 Business Studies (ISSN: 2315-5086), Vol. 4(5) pp. 196-202.
63 Page 22 of 29
64
65
 Rana, I. & Asad, F. (2018). Impact of Corporate Social Responsibility on Financial Performance
Evidence from Pharmaceutical Sector Listed Companies of Pakistan. European Business
1 & Management, 4(1): 1-8, DOI: 10.11648/j.ebm.20180401.11
2
3 Robin, D. P., & Reidenbach, R. E. (1987). Social responsibility, ethics, and marketing strategy:
4 Closing the gap between concept and application. Journal of Marketing, 51(1), 44-58.
5
6 Roberts P.W., & Dowling G.R. (2002). Corporate reputation and sustained superior financial
7
8 performance. Strategic Management Journal, 23:1077–1093. 44
9
10 Sarwar, A., Zahid, I., & Ikram, H. (2012). Corporate Social Responsibility and Financial
11 Performance linkage evidence from the Banking Sector of Bangladesh: Journal of
12 Organizational Management, pp.14-21.
13
14 Seele, P., & Lock, I. (2015). Instrumental and/or deliberative? A typology of CSR
15
16 communication tools. Journal of Business Ethics, 131(2), 401-414.
17
18 Singh, S. (2014). Impact of corporate social responsibility disclosure on the financial
19 performance of firms in UK. Master Thesis, Business Administration - Financial
20 Management, School of Management and Governance, University of Twente
21
22 Shrestha, M. B. & Bhatta, G. R. (2018). Selecting an appropriate methodological framework for
23
24 time series data analysis. The Journal of Finance and Data Science, 4, pp. 71-89,
25 doi.org/10.1016/j.jfds.2017.11.001.
26
27 Spicer, B. H. (1978). Investors, corporate social performance and information disclosure: An
28 empirical study. Accounting Review, 94-111.
29
30 Sulaiman, A., Baharin, R. & AlHadi, A. A. (2019). Impact of Import and Export on GDP of
31
32 Egypt: Application of ARDL Model. International Journal of Asian Social Science, Vol.
33 9 (1), pp. 1-10, ISSN (e): 2224-4441 ISSN (p): 2226-5139, DOI:
34 10.18488/journal.1.2019.91.1.10.
35
36 Suresh, S., & Ganapathy Ramasamy, N. (2015). Analysis of Project Performance Using Earned
37
38 Value Analysis. International Journal of Science, Engineering and Technology
39 Research, 4(4).
40
41 Wright, P., & Ferris, S. (1997). Agency conflict and corporate strategy: The effect of divestment
42 on corporate value. Strategic management journal, 18: 77-83.
43
44 Yunus R., Wazid, S., Hairi, N., Choo, WY. &Hairi, F. (2017). Correction: Association between
45
46 elder abuse and poor sleep: A cross-sectional study among rural older Malaysians. PLOS
47 ONE, Vol: 12(11): e0187782. https://doi.org/10.1371/journal.pone.
48
49 Zeng, S., Ma, H., Lin, H., Zeng, R., & Tam, V. W. (2015). Social responsibility for major
50 infrastructure projects in China. International Journal of Project Management, 33(3), 537-
51
52
548.
53
54 Zulfiqar, S., Sadaf, R., Popp, J., Vveinhardt, J., & Máté, D. (2019). An Examination of Corporate
55 Social Responsibility and Employee Behavior: The Case of
56 Pakistan. Sustainability, 11(13), 3515.
57
58
59
60
61
62
63 Page 23 of 29
64
65
 1
2
3
4
5
6 Model 1 Results CSR and Earning Value (Equation 1)
7
8
9 Table 1: Results of ARDL Model (Equation 1)
10
11 Variable Coefficient Std. Error t-Statistic Prob.*
12 Constant 0.0045 0.0702 0.0643 0.9487
13 Earning Value EV(-1) 0.5571 0.0822 6.7768 0.0000
14 Earning Value EV(-2) 0.1009 0.0657 1.5356 0.1255
15
16 CSP Reputation Index -0.0158 0.0081 -1.9480 0.0522
17 CSP Reputation Index(-1) 0.0131 0.0078 1.6789 0.0941
18 Operating Earnings Asset -0.1032 0.0823 -1.2550 0.2103
19
20 Operating Earnings Sale 0.1472 0.0718 2.0504 0.0410
21 R-squared 0.4472
22 Adjusted R-squared 0.4379
23
F-statistic 48.3972
24
25 Prob(F-statistic) 0.0000
26 Durbin-Watson stat 1.9894
27 Dependent Variable: EV, Method: ARDL, Selected Model: ARDL(2, 1, 0, 0), *Note: p-values
28 and any subsequent tests do not account for model
29
30
31 Table 2. ARDL Bounds Test
32
33 Test Statistic Value k
34 F-statistic 10.49117 3
35 Critical Value Bounds
36 Significance I0 Bound I1 Bound
37 10% 2.37 3.2
38
5% 2.79 3.67
39
40 2.50% 3.15 4.08
41 1% 3.65 4.66
42
43
44 Table 3. Results ARDL Co-integrating and Long Run Form
45
46 Variable Coefficient Std. Error t-Statistic Prob.
47 Co-integrating Form
48 D(EV(-1)) -0.1005 0.0517 -1.9442 0.0526
49
50 D(CSP Reputation Index) -0.0157 0.0075 -2.0795 0.0383
51
52 D(Operating Earnings Asset) -0.0964 0.0754 -1.2792 0.2017
53
54 D(Operating Earnings Sale) 0.1402 0.0731 1.9167 0.0561
55
56 Co-intEq(-1) -0.3428 0.0454 -7.5554 0.0000
57 Long Run Coefficients
58 CSP Reputation Index -0.0076 0.0123 -0.6190 0.5363
59 Operating Earnings Asset -0.3019 0.2333 -1.2937 0.1966
60
Operating Earnings Sale 0.4305 0.1840 2.3394 0.0199
61
62 C 0.0132 0.2051 0.0643 0.9487
63 Page 24 of 29
64
65
 Dependent Variable: D(EV(-1)), Method: ARDL, Selected Model: ARDL(2, 1, 0, 0).
*Note: for model do not account p-values and any subsequent tests
1
2
3
4
5
6
7 Results of Model 2 CSR Levels (Equation 2)
8
9 Table 4. Results of ARDL Model 2 CSR Levels (Equation 2)
10 Variable Coefficient Std. Error t-Statistic Prob.*
11
12
Earning Value EV(-1) 0.5270 0.0869 6.0615 0.0000
13 Earning Value EV(-2) 0.1358 0.0678 2.0044 0.0458
14 Highest CSR 3.2092 0.4656 6.8931 0.0000
15 Highest CSR (-1) -1.2945 0.5892 -2.1972 0.0287
16 Highest CSR (-2) -0.9064 0.4222 -2.1471 0.0325
17 Lowest CSR 0.2791 0.0974 2.8668 0.0044
18 Lowest CSR (-1) -0.1453 0.0924 -1.5719 0.1169
19 Mid CSR 0.0714 0.0521 1.3717 0.1710
20
21
Operating Earnings Asset 0.2506 0.1388 1.8054 0.0719
22 Operating Earnings Asset(-1) -0.2152 0.1156 -1.8620 0.0634
23 Operating Earnings Sale -0.3269 0.1373 -2.3807 0.0178
24 Operating Earnings Sale(-1) 0.3427 0.1248 2.7457 0.0063
25 C -0.4563 0.2006 -2.2740 0.0236
26 R-squared 0.5120
27
Adjusted R-squared 0.4954
28
29 F-statistic 30.8592
30 Prob(F-statistic) 0.0000
31 Durbin-Watson stat 2.0177
32 Dependent Variable: EV, Method: ARDL, Selected Model: ARDL(2, 2, 1, 0, 1, 1), *Note: p-
33 values and any subsequent tests do not account for model
34
35
36 Table 5. ARDL Bounds Test (Equation Model 2)
37 Test Statistic Value k
38 F-statistic 7.946074 5
39 Critical Value Bounds
40 Significance I0 Bound I1 Bound
41
10% 2.08 3
42
43 5% 2.39 3.38
44 2.50% 2.7 3.73
45 1% 3.06 4.15
46
47
48
49
Table 6. ARDL Cointegrating and Long Run Form of CSR Levels (Equation Model 2)
50 Variable Coefficient Std. Error t-Statistic Prob.
51
52 Cointegrating Form
53
54 D(Ev(-1)) -0.1365 0.0506 -2.6992 0.0073
55 D(Highest CSR) 3.2126 0.5174 6.2091 0.0000
56 D(Highest CSR (-1)) 0.9137 0.3712 2.4614 0.0143
57 D(Lowest CSR) 0.2870 0.0921 3.1152 0.0020
58 D(Mid CSR) 0.0935 0.0897 1.0433 0.2975
59
D(Operating Earnings Asset) 0.2500 0.0877 2.8510 0.0046
60
61 D(Operating Earnings Sale) -0.3275 0.0997 -3.2856 0.0011
62 Cointeq(-1) -0.3360 0.0447 -7.5095 0.0000
63 Page 25 of 29
64
65
 Long Run Coefficients
Highest CSR 2.9902 1.0786 2.7722 0.0059
1
Lowest CSR 0.3969 0.1496 2.6532 0.0083
2 Mid CSR 0.2118 0.1449 1.4622 0.1446
3 Operating Earnings Asset 0.1051 0.1825 0.5760 0.5650
4 Operating Earnings Sale 0.0469 0.2070 0.2265 0.8209
5 C -1.3531 0.5018 -2.6965 0.0073
6 Dependent Variable: D(EV(-1)), Method: Cointegrating And Long Run Form,
7 Selected Model: ARDL(2, 2, 1, 0, 1, 1)
8
9
10 Graphical Presentation of CSR Levels
11
12 .5
13
14
.4
15
16
17 .3
18 HIGHEST_CSR1H_EV
19 HIGHEST_CSR2H_EV
HIGHEST_CSR3H_EV
20 HIGHEST_CSR4H_EV .2
21
22
23 .1
24
25
26 .0
27 .0 .1 .2 .3 .4 .5 .6

28 highest_csr0H
29
30 Figure 1: Graphs of EV for Highest/Best CSR relation with Excess Earning
31
32
33
34
35 1.0
36
37
38 0.9
39
40
0.8
41 MID_CSR1H_EV
42 MID_CSR2H_EV
MID_CSR3H_EV
43 MID_CSR4H_EV 0.7
44
45
46 0.6
47
48
49 0.5
.32 .33 .34 .35 .36
50
51 MID_CSR0H_EV
52
53 Figure 2: Graphs of EV for Mid/Fair CSR relation with Excess Earning
54
55
56
57
58
59
60
61
62
63 Page 26 of 29
64
65
 1.0

0.9
1
2
3 0.8
4
5 LOWEST_CSR1H_EV
LOWEST_CSR2H_EV 0.7
6 LOWEST_CSR3H_EV
7 LOWEST_CSR4H_EV
8 0.6
9
10
11 0.5
12
13 0.4
14 .2 .3 .4 .5 .6 .7
15
16 LOWEST_CSR0H_EV
17
18 Figure 3: Graphs of EV for Lowest/Worst CSR relation with Excess Earning
19
20
21
22
23
24 Appendix A. Results of Robustness Tests (Equation 1)
25
26 Breusch-Godfrey Serial Correlation LM Test
27 F-statistic 0.146 Prob. F(2,357) 0.864
28 Obs*R-squared 0.299 Prob. Chi-Square(2) 0.86
29
30 Heteroskedasticity Test: Breusch-Pagan-Godfrey
31 F-statistic 2.1342 Prob. F(6,359) 0.048
32
Obs*R-squared 12.605 Prob. Chi-Square(6) 0.049
33
34 Scaled explained SS 74.096 Prob. Chi-Square(6) 0
35 Ramsey RESET Test
36
37 Value Df Probability
38 t-statistic 1.310 358 0.1907
39 F-statistic 1.718 (1, 358) 0.1907
40
41 F-test summary:
42 Sum of Sq. Df Mean Squares
43 Test SSR 0.26910 1 0.26910
44
45 Restricted SSR 56.3310 359 0.15691
46 Unrestricted SSR 56.0619 358 0.15659
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63 Page 27 of 29
64
65
 1
2
3
4
5
6
7 Appendix B. Results of Robustness Tests of CSR Levels (Equation 2)
8 Breusch-Godfrey Serial Correlation LM Test
9 F-statistic 0.6258 Prob. F(2,351) 0.5354
10
11 Obs*R-squared 1.3004 Prob. Chi-Square(2) 0.5219
12 Heteroskedasticity Test: Breusch-Pagan-Godfrey
13 F-statistic 1.4394 Prob. F(12,353) 0.1459
14 Obs*R-squared 17.0734 Prob. Chi-Square(12) 0.1469
15
16
Scaled explained SS 98.9127 Prob. Chi-Square(12) 0
17 Variance Inflation Factors
18 Coefficient Uncentered Centered
19 Variable Variance VIF VIF
20 EV(-1) 0.0075 5.6365 5.6362
21
EV(-2) 0.0046 4.4922 4.3847
22
23 Highest CSR 0.2167 148.1731 4.0986
24 Highest CSR (-1) 0.3471 234.2183 6.4001
25 Highest CSR (-2) 0.1782 114.1685 4.0902
26
27
Lowest CSR 0.0095 15.3797 8.7056
28 Lowest CSR (-1) 0.0085 13.7816 7.7364
29 Mid CSR 0.0027 3.7739 2.1694
30 Opreating Earnings Asset 0.0193 67.4888 14.7175
31
32
33 Opreating Earnings Asset(-1) 0.0134 61.1153 17.7922
34 Opreating Earnings Sale 0.0188 54.3113 13.6337
35 Opreating Earnings Sale(-1) 0.0156 61.3594 21.7468
36 C 0.0403 173.4091 NA
37 Ramsey RESET Test
38
39
Value df Probability
40 t-statistic 1.417701 352 0.1572
41 F-statistic 2.009877 (1, 352) 0.1572
42 F-test summary:
43 Sum of Sq. df Mean Squares
44 Test SSR 0.282331 1 0.282331
45
46 Restricted SSR 49.72839 353 0.140874
47 Unrestricted SSR 49.44606 352 0.140472
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
Page 28 of 29
64
65
 1
2
3
4
5 Appendix C. List of Sample Companies
6 Sr. # Company Name Sr. # Company Name
7 1 Abbott Laboratories (Pakistan) Ltd. 24 Indus Motor Company Ltd.
8
9 2 AL-Ghazi Tractors 25 Kohinoor mills
10 3 Atlas Honda Ltd. 26 Kot Addu Power Co ltd.
11
12 4 Attock Cement 27 Javedan Co ltd.
13 5 Attock Petroleum 28 Lotte pakistan PTA Ltd.
14
15
6 Bata Pakistan Ltd. 29 Lucky Cement Ltd.
16 7 Bestway Cement Ltd. 30 Mari Gas Company Ltd.
17 8 Byco Petroleum 31 Millat
18
19 9 Clover Co. 32 Murree Brewery Company Ltd.
20 10 Colgate - Polmolive (Pakistan) Ltd. 33 media Co ltd.
21
22 11 D. G. Khan Cement Company Ltd. 34 National Refinery
23 12 Dawood Hercules Chemicals Ltd. 35 Nestle Pakistan
24 13 Engro Corporation Ltd. 36 Netsol Technologies Ltd.
25
26 14 Engro Polymer & Chemical Ltd. 37 Nishat Mills Ltd.
27 15 Fatima Fertilizer Co.Ltd. 38 Oil & Gas Dev. Corp. (OGDC) (Pub.)
28
29 16 Fauji Fertilizer Bin Qasim Ltd. 39 Pakistan National Shipping Corp. Ltd. (Pub.)
30 17 Fauji Fertilizer Company Ltd. 40 Pakistan Telecommunication Co. Ltd. (PTCL)
31
32
18 Ghani Glass Ltd. 41 Pakistan State Oil Company Ltd. (Pub.)
33 19 Glaxo Smithkline (Pakistan) Ltd. 42 Pace (Pakistan) Ltd.
34 20 The Hub Power Company Ltd. 43 Packages Limited
35
36 21 Ibrahim Fibres Ltd. 44 Pak Elektron Ltd
37 22 ICI Pakistan Ltd. 45 Pak Suzuki Motor Company Ltd
38
39 23 Indus Dyeing & Manu. Company Ltd. 46 Pakistan Telephone Cables Ltd.
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
Page 29 of 29
64
65