Received: 20 February 2020 Revised: 14 January 2021 Accepted: 18 February 2021

DOI: 10.1111/irfi.12348

LETTERS SECTION

Firm's tax aggressiveness under mandatory CSR

regime: Evidence after mandatory CSR regulation
of India

Mehul Raithatha1 | Tara Shankar Shaw2

1
Indian Institute of Management Indore,
Indore, Madhya Pradesh, India
Abstract
2
Department of Humanities and Social In this study, we use the mandatory CSR spending regulation
Sciences, Indian Institute of Technology implemented by India in 2015 to examine whether firms that
Bombay, Mumbai, Maharashtra, India
comply with the regulation change their tax aggressiveness.
Correspondence We document that firms that comply with CSR regulation
Mehul Raithatha, Indian Institute of end up having less tax aggression which supports the argu-
Management Indore, Prabandh Shikhar, Rau-
Pithampur Road, Indore 453556, Madhya
ment that enhanced visibility and firm-level reputational con-
Pradesh, India. cerns play a vital role in shaping up the relationship between
Email: mehulr@iimidr.ac.in; mehular83@
CSR and taxation policy. Our results are consistent with the
gmail.com
number of robustness checks.

KEYWORDS
CSR, tax aggressiveness, visibility

1 | I N T RO DU CT I O N

Increasingly, as different economies and various multilateral institutions are becoming concerned about an organization's
corporate social responsibility (CSR), a topical question that arises is how firms would pan out their taxation policy in the
wake of such increased demand for CSR. For a firm, both taxation and CSR are the usage of its resources for societal
development, which has both organizational and societal implications. Corporate taxes are direct payments by the firms to
the government that is ultimately used for the procurement of public goods for the society. Whereas, CSR is a comprehen-
sive set of policies on business ethics, community, and environment, etc. that are integrated into the firm's activities
(Carroll, 1991). On the one hand, firm's taxation policy can be considered as part of its CSR activities and therefore firms
with proactive CSR policy are less likely to engage in aggressive tax practices (Sikka, 2010). On the other hand, since both
CSR and tax expenditure are not related to its core business activities, firms with higher CSR expenditure are likely to have
an aggressive taxation policy to economize on cost. The literature that tries to disentangle the causal relationship between
CSR and taxation is scant and mostly concentrated in developed economies where CSR expenditure is voluntary. However,
in a mandatory CSR regime, the relationship between CSR and taxation would be more complex. On the one hand, compliance

© 2021 International Review of Finance Ltd.

International Review of Finance. 2021;1–9. wileyonlinelibrary.com/journal/irfi 1

2 RAITHATHA AND SHAW

with the mandatory CSR policy can be considered as additional tax but on the other hand, it enhances firms' visibility leading to
higher scrutiny that results in the loss of reputation under aggressive tax policy. In 2013, India implemented the mandatory
CSR regulation depending on the firm's profitability and size under its new Companies Act 2013 (CA2013). It required that
companies whose net worth exceeds ₹5 billion, or whose annual turnover exceeds ₹ 10 billion, or whose profit exceeds ₹
50 million in any financial year, have to spend 2 % of their past three-years average profit on CSR. We use the mandatory
CSR regulation in India, to study the causal effect of a firm's mandatory CSR spending on its taxation behavior.
The extant empirical literature has largely focused on the effect of voluntary CSR on a firm's taxation policy that
is explained based on two competing theories (Col & Patel, 2019). First, the corporate culture theory, which argues
that the firm's decisions are based on its corporate belief and firms that believe in the right corporate behavior would
be more active on CSR but would also follow less tax aggressive policies resulting in a negative relationship between
the both. Studies like Watson (2015) and Lanis and Richardson (2012) have documented such a negative relation-
ship. Second, is the risk management theory, which argues that firms would expend more on CSR activities to cover
its reputational damage arising from tax aggressive behavior like tax lobbying activities (Davis, Guenther, Krull, &
Williams, 2015) and opening up tax haven affiliates (Col & Patel, 2019). Theoretical literature has documented that
managers have incentives to engage in tax aggressiveness if the benefits like reduced tax liabilities, and maintenance
of favorable credit ratings, etc. exceed the costs, from litigation, penalties, and reputation damage (Scholes, Wolfson,
Erickson, Maydew, & Shevlin, 2014). We posit that firms that comply with the mandatory CSR regulation enhance
their domain-specific visibility to the tax authority since it signals about its higher present and future profitability
(Lys, Naughton, & Wang, 2015) resulting in a higher likelihood of scrutiny and sanctions in case of lower tax pay-
ment. Therefore, they will be willing to forgo the short-term profit achieved through an aggressive taxation policy to
contain the reputation risk that follows from the tax authorities' indictment of non-compliance to taxation policies.
We estimated the causal effect of CSR compliance on a firm's taxation policy on a sample of 1,577 non-financial firms
(6,082 firm-year observations) listed on the Bombay Stock Exchange (BSE) that were required to comply with the CSR reg-
ulation in any year between 2015 to 2018. Using a two-stage approach to account for potential self-selection bias arising
from the “comply or explain” clause (which does not prescribe any penalties under non-compliance) we document that firms
that comply with CSR regulation end up showing less tax aggressiveness measured by the effective tax rate (ETR) and
book-tax difference measure of Desai and Dharmapala (2006) (BTD_DD). Further, we found that the decline in the tax
aggressiveness was more prevalent among the firms that did not spend on any CSR activities before 2015 but complied
with CSR regulation of CA2013 after 2015. Our results are consistent with the number of robustness checks.
Our paper makes three important contributions to the literature relating to CSR and tax aggression. First, to the best
of our knowledge, this is the first work that relates CSR compliance to tax aggression under the mandatory CSR regime.
We show that the firms that comply with the mandatory CSR regulation (that is based on profit and size) decrease their
tax aggression most likely due to increased visibility with the regulatory authorities. Second, we provide evidence to the
claim of Mukherjee, Bird, and Duppati (2018) the mandatory CSR regulation in India was implemented to enforce socially
responsible behavior among firms that had resources to undertake CSR expenditure but did not do so voluntarily. We find
firms that did not spend on CSR before 2015 but complied with the regulation after 2015 had the highest decline in tax
aggressiveness after 2015 compared to years before it. Finally, we also show how reforms like mandatory CSR regulations
had real effects on strategic variables of the firms like tax aggression for the firm's that complied with the regulation.

2 | L I T E R A T U R E A N D H Y P O TH E S I S

The relationship between CSR and tax aggression is quite an enigma since different researchers have documented
mixed and contradictory evidence. According to Col and Patel (2019), prior studies draw their inferences motivated
by corporate culture theory and risk management theory. Corporate culture theory advocates that all the firm-level
decisions are based on their shared beliefs. Firms believing in “right” corporate behavior would not only have more
CSR expenditure but would also have lower tax aggression. Watson (2015), using the Kinder, Lydenberg, and Domini
RAITHATHA AND SHAW 3

(KLD) index as a measure of CSR performance in the US context showed that it reduced unrecognized tax benefits
and concluded that the socially responsible firms are less tax aggressive. Similarly, Lanis and Richardson (2012) found
a significant negative association between CSR performance and tax aggressiveness. Several other studies, like Hus-
eynov and Klamm (2012), Lanis and Richardson (2015), and others using the KLD database, confirm the previous
results. On the other hand, the risk management theory argues that firms might use CSR activities as a potential
hedging tool against reputation damaging tax aggressive activities like tax lobbying (Davis et al., 2015) and investing
in entities in a tax haven (Col & Patel, 2019). These authors concluded that, on average, managers and other stake-
holders do not view the payment of corporate taxes as an important part of a firm's CSR activities.
In the Indian economy, the direct tax contributes more than 50% of the total tax revenue, of which 50% is
accounted for by the corporate tax. Indian corporate profit tax rate of 30% is ranked as one of the highest corporate
tax rates in the world which is higher than other developed countries like Japan (29.79%), and USA (25.8%). How-
ever, Indian firms have been using tax aggressive practices resulting in the effective tax rate being lower than 25%.
The Income Tax Department has been actively involved in regularly scrutinizing the corporate tax payment therefore,
the corporate taxes as a percentage of gross tax revenue has increased from 21% in 2003 to 32% in 2018. The
increased security has resulted in the companies using complex accounting procedures to save taxes by increasing
the information asymmetry with the tax authority. Consequently, the regulators are more vigilant about the firm's
resources and their potential to pay taxes. Therefore, firms that are more visible due to their high profit and size
would have more incentive to be less tax aggressive and avoid any indictment from the tax authorities.
Organizational visibility is the extent to which an organization can be noticed by other stakeholders and it is considered
as an important attribute of an organization. The extant literature has discussed that the visibility of the organization helps
to reduce information asymmetry (Brammer, Millington, & Pavelin, 2006), acts as a disciplinary mechanism, improves com-
pliances (King, 2008), and incentivizes the firms to adopt transparent accounting practices (Watts & Zimmerman, 1978). It
has been found that visible firms are subjected to higher scrutiny and therefore they are more compliant to the regulatory
pressure (Marquis, Toffel, & Zhou, 2016). According to Marquis et al. (2016), organization visibility can be classified into two
types: generic visibility and domain-specific visibility. Firms possessing generic visibility are reputed and are well known
around the society, whereas firms having domain-specific visibility have some specific characteristics that expose the firms
to a greater degree of scrutiny by stakeholders specific to that domain. Previous research has used firm size and profitability
as a proxy for domain-specific visibility (Marquis et al., 2016). Since in India, the mandatory CSR regulation is based on a
firm's size and profitability, the firms that comply with the regulation enhance their domain-specific visibility to the tax
authorities since it signals about their higher future profitability (Lys et al., 2015). This results in a higher likelihood of scru-
tiny and sanctions in case of lower tax payment that adversely affects their reputation and has an unfavorable effect on
their performance (Sánchez & Sotorrío, 2007). Thus, firms having higher visibility of their profit would be willing to forgo
short-term benefits achieved through aggressive taxation policy to contain the reputation risk that follows from the tax
authorities' indictment of noncompliance to tax payment. From the firm's point of view, being tax aggressive is beneficial to
the shareholders since it increases the firm's profit, but it is costly to the firm's insiders if they receive partial benefit but face
the entire cost of scrutiny (Kanagaretnam, Lee, Lim, & Lobo, 2018). Thus, in face of higher scrutiny by firms' stakeholders
especially by the government, being tax aggressive is costlier for the firms and so we hypothesize that:
Hypothesis: Firms that comply with the mandatory CSR norms are less likely to engage in tax aggressive activities.

3 | DATA, METHODOLOGY, AND VARIABLES

Our data consists of the listed firms in the BSE that were mandated by the CA2013 to incur 2% of their average three-year
profit before tax as CSR expenditure in the year 2015–2018. The data was sourced from the Prowess database maintained
by the Centre for Monitoring Indian Economy. In total, there were 1,577 listed firms that were mandated to undertake
CSR expenditure accounting to 6,082 firm-year data. Among our sample firms, the annual compliance with the CSR statute
was 26%, 42%, 48%, and 52% for the year 2015, 2016, 2017, and 2018, respectively. It is likely that the sample of firms
4 RAITHATHA AND SHAW

that chose to comply with the regulation have a different incentive, in terms of cost and benefits of compliance, than those
firms who did not comply creating a potential for self-selection bias.

3.1 | Empirical model and variables

The self-selection bias can result either from firm-specific unobserved time-invariant or variant factors. Firm-level fixed-
effect controls for time-invariant factors like a firm's beliefs and culture. Unobserved firm-level, time-variant factors like
political connectedness,1 both with local and national political leaders can increase the firm's tax aggressive policy. How-
ever, it can also persuade a firm to comply with the CSR regulation to maintain political capital. This creates a positive
selection bias that is addressed by estimating a two-stage Heckman model (Heckman, 1979). In the first stage, we calcu-
late the Inverse Mills Ratio (IMR) from a Probit estimation of the binary variable Comply (which takes the value 1 if the
firm complies with the mandated CSR) on other control variables and two additional exogenous variables: the propor-
tion of firms in the industry that comply with the mandatory CSR regulation (Pro_Ind_Comply) and average sales in an
industry for a given year (Ind_Avg_Sales) to meet the exclusion restriction (Lennox, Francis, & Wang, 2011). We argue
that firms belonging to the industry which have high Pro_Ind_Comply are more likely to comply with the regulation.
However, firms belonging to industries with high Ind_Avg_Sales are more likely to comply with the regulation since they
have high resources but simultaneously, they are more resistant to stakeholder's pressure making them less likely to
comply. Thus, the direction is not clear for Ind_Avg_Sales. Importantly, these variables are unlikely to be directly associ-
ated with the firm's tax aggressiveness and thus satisfy the exclusion restriction. Our first-stage model is:

Complyit = π 0 + π 1 Pro_Ind_Complyit + π 2 Ind_Avg_Salesit + πXit + νit ð1Þ

Xs are all the control variables used in the second stage of the estimation. To capitalize on the longitudinal
nature of our data, we calculate the IMR using annual Probit estimation following the suggestion of Wool-
dridge (1995). In our second stage model, we regress the measures of tax aggression on Comply, other control vari-
ables, IMR and also include firm and year fixed effect, μi and δt respectively.

TaxAit = β0 + β1 Complyit + β2 IMRit + γX it + μi + δt + ɛit : ð2Þ

We use two measures of tax aggressiveness (TaxAit). First, is the effective tax rate (ETR) that is defined as total tax expense
divided by profit before tax (PBT). The higher the ETR, the lower is the tax aggressiveness of the firm and it measures the
firm's tax burden. However, ETR does not capture aggressive tax planning either through investment in a tax haven or in a
tax-favored asset which reduces taxable income but not the book income. Moreover, ETR is influenced by accounting for
tax-related accruals. To overcome these concerns, we use the book-tax difference (BTD) measure of Desai and
Dharmapala (2006) (BTD_DD), which controls the effect of accrual by taking the residual of the firm-fixed effect regression
of BTD on the accruals. Higher BTD_DD implies higher tax aggression of the firm. β1 captures the causal effect of manda-
tory CSR compliance on tax aggressiveness and β2 captures the direction of the selection bias.

4 | RESULTS

4.1 | Univariate analyses

In Table 1, we provide descriptive statistics for our sample. The mean value of the ETR is 0.28 which is statistically
different from the statutory corporate tax rate of 0.3 and the mean value of BTD_DD is 0.021 which is statistically
different from zero. Thus, the firms in our sample exhibit tax aggressive behavior. In the un-tabulated correlation
RAITHATHA AND SHAW 5

TABLE 1 Descriptive statistics

Variables Mean SD Variables Mean SD Variables Mean SD

BTD_DD 0.021 0.043 Aud_TaxServ 0.196 0.397 PPE 0.464 0.336
ETR 0.281 0.115 Leverage 0.489 0.212 ROA 0.113 0.079
Comply (dummy) 0.418 0.493 PB 3.146 3.392 Board_Size 10.39 4.002
Family (dummy) 0.500 0.500 Advertisement 0.017 0.093 Prop_Indep 0.423 0.141
Foreign_Income 0.747 0.435 R&D 0.004 0.013 Age 3.461 0.567
Op_Loss 0.085 0.279 Intangible 0.022 0.075 Size 8.365 2.041

Note: BTD_DD is book-tax difference based on Desai and Dharmapala (2006). ETR is tax expense divided by profit before tax.
Comply is equal 1 if firms CSR spending is more than the mandated requirement. Family is equal 1 if firm is family controlled and
family managed. Foreign_Income is Ln(foreign exchange earnings). Op_Loss is 1 if firm reports operating loss. Aud_TaxServ is 1 if firm
avail tax services from the auditors. Leverage is borrowings divided by total assets. PB is price to book ratio. Advertisement is
advertising intensity. R&D is research and development intensity. Intangible is intangible assets divided by total assets. PPE is
tangible assets divided by total assets. ROA is earnings before interest and tax scaled by total assets. Board_Size is Ln(Board size).
Prop_Indep total independent directors divided by board size. Age is Ln(firm age since incorporation). Size is Ln(total sales).

matrix we found that Comply is positively and significantly correlated with BTD_DD and negatively correlated with
ETR that is inconsistent with hypothesis 1. However, this correlation is not causation as it may be subjected to sam-
ple selection bias which is addressed in our multivariate analysis.

4.2 | Multivariate analyses

The un-tabulated first-stage results show that in all the annual estimations, the coefficient of Pro_Ind_Comply is posi-
tive and significant (year 2015: β = 3.75, p < .01; year 2016: β = 2.46, p < .01; year 2017: β = 2.41, p < .01; year
2018: β = 2.19, p < .01) and the coefficient of Ind_Avg_Sales is negative and significant (year 2015: β = − 0.19,
p < .01; year 2016: β = − 0.14, p < .01; year 2017: β = − 0.24, p < .01; year 2018: β = − 0.25, p < .01). The p-value
of the Sargan over-identification test is 0.48, indicating that we cannot reject the null hypothesis that Ind_Avg_Sales
and Pro_Ind_Comply are exogenous in the second-stage model.
We report the results from our second-stage estimation in Table 2. We have included full sets of firm-year fixed
effects and reported bootstrap standard errors, clustered at the industry-level. In models 1 and 2, we report the baseline
estimation without controlling for the IMR. The results show that without selection bias correction, as firms that comply
with the CSR regulation increases their ETR but have no significant effect on BTD_DD. Once we include the IMR to con-
trol for the self-selection bias in models 3 and 4, we find that firms that comply with the mandated CSR regulation have
higher ETR (Model 3: β = 0.04, p < .05) and lower BTD_DD (Model 4: β = − 0.014, p < .05) and they are statistically sig-
nificant. Economically, these values imply that for an average firm that complies with the CSR regulation has ₹1.7
crores2 higher tax expense and ₹ 8.7 crores lower book-tax-income difference than not complying with the CSR regula-
tion. In both cases, the IMR is statistically significant but of the opposite sign of the Comply variable implying that there
is a positive selection bias that dampens the true causal effect of compliance on tax aggression if not controlled for.

5 | ADDITIONAL TESTS

5.1 | Role of pre-regulation and voluntary CSR behavior on tax aggression

Mukherjee et al. (2018) documented that the purpose of the mandatory CSR regulation was to enforce CSR spend-
ing for firms that had resources but chose not to spend voluntarily before 2015. Building on their arguments, we
6 RAITHATHA AND SHAW

TABLE 2 Impact of CSR compliance on tax aggressiveness

Without IMR With IMR

ETR (1) BTD_DD (2) ETR (3) BTD_DD (4)

Variables Coeff SE Coeff SE Coeff SE Coeff SE

Comply (dummy) 0.009*** (0.003) −0.001 (0.001) 0.042** (0.021) −0.014** (0.007)
Family −0.010 (0.009) −0.003 (0.003) −0.010 (0.009) −0.003 (0.003)
Foreign_Income −0.002 (0.006) 0.001 (0.002) −0.003 (0.006) 0.001 (0.002)
Op_Loss −0.008 (0.007) 0.000 (0.004) −0.006 (0.008) −0.001 (0.004)
Aud_TaxServ −0.004 (0.007) −0.000 (0.002) −0.004 (0.005) −0.000 (0.002)
Leverage 0.039** (0.016) −0.029*** (0.009) 0.040** (0.016) −0.029*** (0.009)
PB 0.000 (0.001) −0.001* (0.000) 0.000 (0.001) −0.001* (0.000)
Advertisement −0.001 (0.036) 0.012 (0.018) −0.002 (0.036) 0.013 (0.018)
R&D −0.216 (0.299) 0.124 (0.156) −0.190 (0.300) 0.114 (0.154)
Intangible −0.026 (0.034) −0.004 (0.015) −0.030 (0.035) −0.003 (0.015)
PPE 0.008 (0.008) −0.001 (0.003) 0.010 (0.009) −0.002 (0.003)
ROA −0.113*** (0.027) 0.189*** (0.020) −0.122*** (0.028) 0.193*** (0.020)
Board_Size 0.001** (0.000) −0.000 (0.001) 0.001* (0.000) −0.000 (0.000)
Prop_Indep 0.015 (0.013) −0.002 (0.005) 0.010 (0.014) 0.001 (0.005)
Age 0.064* (0.036) −0.007 (0.014) 0.066* (0.036) −0.008 (0.013)
Size 0.005 (0.003) −0.000 (0.002) 0.004 (0.003) 0.000 (0.002)
IMR — — — — −0.022** (0.011) 0.008** (0.004)
Constant −0.005 (0.124) 0.049 (0.048) −0.013 (0.124) 0.052 (0.048)
R-square 0.018 0.129 0.018 0.130

Note: Bootstrap standard error clustered at the level of industry is reported in the parenthesis. The estimation has both the
firm-level and year-level fixed effect. Observations = 6,082, firms = 1,577.
*p < .1. **p < .05. ***p < .01.

claim that firms that complied with the mandatory CSR regulation but did not spend on CSR before 2015 would
be more visible and therefore would be more under the scrutiny of the tax authority. To test our claim, we sub-
divided the variable Comply into two categories: CSRBf_Comply and NoCSRBf_Comply. CSRBf_Comply
(NoCSRBf_Comply) is one if Comply is one and the firms had (not) spent on social and community expenditure in
any year between 2011 and 2014. The estimated results are reported in Models 1 and 2 of Table 3 for ETR and
BTD_DD respectively and it shows that the ETR (BTD_DD) has increased (decreased) both for NoCSRBf_Comply
firms and CSRBf_Comply firms. Furthermore, in the coefficient difference test for CSRBf_Comply and
NoCSRBf_Comply we fail to reject the hypothesis that the coefficients are equal. Thus, these results imply that
after implementing the mandatory CSR regime the firms CSRBf_Comply and NoCSRBf_Comply have a similar level
of tax aggression, albeit less than the firm that did not comply with the regulation. Following Mukherjee
et al. (2018) we also estimated the change in tax aggression in the period 2015–2018 from 2011 to 2014 for each
category of firms CSRBf_Comply, NoCSRBf_Comply, and No_Comply. We compiled the data from 2011 to 2018 and
created a variable Post that takes the value one for years after 2014. We estimated a fixed effect model both with
ETR and BTD_DD as the dependent variable. The results are reported in Table 4 and we found that only the sub-
category of firms NoCSRBf_Comply has a positive(negative) significant coefficient for Post in the estimation
with ETR(BTD_DD). These results imply that only NoCSRBf_Comply firms have decreased their tax aggression
significantly after 2015 due to higher scrutiny by the tax regulators.
RAITHATHA AND SHAW 7

TABLE 3 The effect of pre-regulation and voluntary CSR behavior on tax aggression

ETR (1) BTD_DD (2) ETR (3) BTD_DD (4)

Variables Coeff SE Coeff SE Coeff SE Coeff SE

CSRBf_Comply 0.040* (0.023) −0.013* (0.007) — — — —
NoCSRBf_Comply 0.047** (0.021) −0.014** (0.007) — — — —
Comply — — — — 0.042** (0.018) −0.011* (0.006)
NotComply_CSRNZero — — — — 0.009* (0.004) 0.0004 (0.002)
Voluntary CSR — — — — 0.025 (0.021) −0.007 (0.007)
IMR −0.023** (0.011) 0.008** (0.004) −0.016* (0.009) 0.006** (0.003)
Obs (firms) 6,082 (1577) 6,082 (1577) 6,743 (1758) 6,743 (1758)
R-square 0.019 0.130 0.018 0.173

Note: Industry clustered bootstrap standard error in the parenthesis, other control variables, firm-level and year-level fixed
effects are included.
*p < .1. **p < .05. ***p < .01.

Further, we studied the tax aggressive behavior of the firms that spent on CSR voluntarily during the period
2015–2018. Voluntary CSR firms are those firms that were not mandated to spend on CSR activities but have vol-
untarily undertaken CSR activity in any year between 2015 and 2018. In this process, we have added 661 observa-
tions for 181 firms to our original data. We consider four categories of the firm: first, firms that complied with the
mandatory CSR regulation (Comply), second those who were mandated to undertake CSR but the amount spend
did not meet the requirement (NotComply_CSRNZero), third, those firms that were mandated but did not spend on
CSR (Not_Comply) and finally, firms that have spent voluntarily during this period (Voluntary_CSR). In our estima-
tion, we have re-calculated the IMR where any firm that had undertaken CSR expenditure was coded as one. The
estimation result is reported in columns 3 and 4 of Table 3 and we found that the ETR (BTD_DD) is positive (nega-
tive) and significant only for the Comply variable. The increase in ETR for NotComply_CSRNZero is small and only
significant at 10%. The coefficient of Voluntary_CSR is insignificant in all the estimations. If we estimate the model
for only the sub-sample of 181 firms, Voluntary_CSR is not significant. Thus, given these tests, it can be concluded
that the decrease in tax-aggressive behavior is observed in those firms that have complied with the mandatory
CSR requirement.

5.2 | Alternate measures of tax aggression and effect of firm heterogeneity

We controlled for a mechanical increase in the ETR or a reduction in BTD_DD by adding CSR expenditure to the PBT
and re-calculating the adjusted ETR, adj_ETR, and adjusted BTD_DD, adj_BTD_DD. The estimated coefficient of Com-
ply both for adj_ETR and adj_BTD_DD are similar to the previous estimation and are not reported for the sake of brev-
ity. Further, following Huseynov and Klamm (2012, pg 814) we re-estimated the model using excess tax aggressive
which is the difference between actual and the natural level of tax aggression. The natural level of tax aggression is
the estimated level of tax aggression determined by regressing the actual level of tax aggressiveness on firm-level
factors and fixed effects. We found the estimated coefficient of Comply is β = 0.044 (p < .01) with excess ETR and
β = − 0.011 (p < .01) with excess BTD_DD. In all these estimations the IMR has the expected sign. This implies that
firms that comply with the CSR regulation decrease tax aggression. The tabulated results are not reported for the
sake of brevity. To study the effect of firm-level heterogeneity on the relationship of CSR compliance and tax aggres-
sion we interacted the variable Comply with previous tax disputes of firm, firm ownership variable Family which takes
the value one if the firm is owned and managed by a family and board-level variables. In these estimations, the
8 RAITHATHA AND SHAW

TABLE 4 The change in tax aggression in 2015–2018 compared to 2011–2014

ETR BTD_DD

Variables CSRBf_Comply NoCSRBf_Comply No_Comply CSRBf_Comply NoCSRBf_Comply No_Comply

Post 0.002 0.017*** 0.006 0.002 −0.002** −0.003
(0.010) (0.003) (0.004) (0.003) (0.001) (0.003)
Constant −0.485** 0.232*** 0.120** 0.275*** 0.016 0.056***
(0.213) (0.037) (0.050) (0.064) (0.012) (0.022)
Obs(firms) 831(105) 7,933(1021) 3,406(449) 831(105) 7,933(1021) 3,406(449)
R-square 0.071 0.022 0.026 0.103 0.064 0.121

Note: Industry clustered bootstrap standard error in the parenthesis other control variables, and firm fixed effects are
included. Post takes value 1 for the year 2015–2018 and 0 otherwise.
*p < .1. **p < .05. ***p < .01.

interaction term is insignificant implying that the firm's ownership and governance do not moderate the relationship
between CSR and tax aggressiveness.

6 | C O N CL U S I O N

Our research provides evidence of the causal relationship of how firms complying with the mandatory CSR expendi-
ture decrease their tax aggression after controlling for the self-selection bias. Aguilera, Rupp, Williams, and
Ganapathi (2007) suggested that the government enacted CSR rules are effective in establishing social expectations
about responsible corporate behavior. On the similar lines, our research attests that a firm's compliance with manda-
tory CSR expenditure that is contingent on its profitability and size exposes it to more visibility and scrutiny resulting
in lower tax aggression of the firm.
However, our paper has a few limitations. First, our use of the two-step Heckman correction was to address
the important endogeneity bias due to the self-selection issue arising from the “comply or explain” clause of the
mandatory CSR regulation. For example, in a KPMG survey of 2017 of the top 100 companies of India, it was
found that 37 companies self-selected to spent less than the mandated amount by citing the reason of “exploring
opportunities”. Although we have addressed the self-selection bias in our analysis, endogeneity bias can also arise
if a firm's taxation policy drives CSR policy. For instance, firms that follow an aggressive taxation policy might
comply with the CSR regulation to dampen the negative effect on their reputation (Col & Patel, 2019). Our empiri-
cal strategy is less likely to account for these types of endogeneity. However, if tax aggression would have driven
CSR then firms complying with CSR regulation would have higher tax aggression. Since our result shows that CSR
compliance reduced tax aggression, we believe that the reverse causality bias would be negligible if any. Second,
in our empirical estimation, the Comply dummy captures the mandatory aspect of CSR compliance. Even those
companies that comply with the regulation might vary in their CSR policy or just take a “tick the box” approach.
However, our research shows that such approaches will also increase firms' visibility that incentivizes them to
undertake socially responsible behavior but future research can delve into how CSR level variations affect a firm's
taxation and other firm-level policies.

ENDNOTES
1
Previous studies have used political donation at the central level to measure of political connectedness. However, this
does not provide any information of political connections at local level which is important for ours study.
RAITHATHA AND SHAW 9

2
The average PBT in our sample firm is Rs 39.3 Cr, so firm that complied with the CSR regulation on average have
0.04 × 39.3 ≈ 1.7 Cr higher tax expense.

RE FE R ENC E S
Aguilera, R. V., Rupp, D. E., Williams, C. A., & Ganapathi, J. (2007). Putting the S back in corporate social responsibility: A
multilevel theory of social change in organizations. Academy of Management Review, 32(3), 836–863.
Brammer, S., Millington, A., & Pavelin, S. (2006). Is philanthropy strategic? Analysis of charitable giving in UK companies.
Business Ethics: A European Review, 15(3), 234–245.
Carroll, A. B. (1991). The pyramid of corporate social responsibility: Toward the moral management of organizational stake-
holders. Business Horizons, 34(4), 39–48.
Col, B., & Patel, S. (2019). Going to haven? Corporate social responsibility and tax avoidance. Journal of Business Ethics, 154
(4), 1033–1050.
Davis, A. K., Guenther, D. A., Krull, L. K., & Williams, B. M. (2015). Do socially responsible firms pay more taxes? The
Accounting Review, 91(1), 47–68.
Desai, M. A., & Dharmapala, D. (2006). Corporate tax avoidance and high-powered incentives. Journal of Financial Economics,
79(1), 145–179.
Heckman, J. J. (1979). Sample selection bias as a specification error. Econometrica, 47(1), 153–161.
Huseynov, F., & Klamm, B. K. (2012). Tax avoidance, tax management and corporate social responsibility. Journal of Corpo-
rate Finance, 18(4), 804–827.
Kanagaretnam, K., Lee, J., Lim, Y., & Lobo, J. (2018). Cross-country evidence of independent media constraining corporate
tax aggressiveness. Journal of Business Ethics, 150(3), 879–902.
King, W. R. (2008). An integrated architecture for an effective knowledge organization. Journal of Knowledge Management,
12(2), 29–41.
Lanis, R., & Richardson, G. (2012). Corporate social responsibility and tax aggressiveness: A test of legitimacy theory.
Accounting, Auditing and Accountability Journal, 26(1), 75–100.
Lanis, R., & Richardson, G. (2015). Is corporate social responsibility performance associated with tax avoidance? Journal of
Business Ethics, 127(2), 439–457.
Lennox, C. S., Francis, J. R., & Wang, Z. (2011). Selection models in accounting research. The Accounting Review, 87(2),
589–616.
Lys, T., Naughton, J. P., & Wang, C. (2015). Signalling through corporate accountability reporting. Journal of Accounting and
Economics, 60(1), 56–72.
Marquis, C., Toffel, M. W., & Zhou, Y. (2016). Scrutiny, norms, and selective disclosure: A global study of greenwashing.
Organization Science, 27(2), 483–504.
Mukherjee, A., Bird, R., & Duppati, G. (2018). Mandatory corporate social responsibility: The Indian experience. Journal of
Contemporary Accounting and Economics, 14(3), 254–265.
Sánchez, J. L. F., & Sotorrío, L. L. (2007). The creation of value through corporate reputation. Journal of Business Ethics, 76(3),
335–346.
Scholes, M. S., Wolfson, M. A., Erickson, M., Maydew, E., & Shevlin, T. (2014). Taxes & business strategy. Upper Saddle River,
NJ: Prentice Hall.
Sikka, P. (2010). Smoke and mirrors: Corporate social responsibility and tax avoidance. Accounting Forum, 34(3–4), 153–168.
Watson, L. (2015). Corporate social responsibility, tax avoidance, and earnings performance. The Journal of the American Tax-
ation Association, 37(2), 1–21.
Watts, R. L., & Zimmerman, J. L. (1978). Towards a positive theory of the determination of accounting standards. The
Accounting Review, 53(1), 112–134.
Wooldridge, J. M. (1995). Selection corrections for panel data models under conditional mean independence assumptions.
Journal of Econometrics, 68(1), 115–132.