**4. Empirical findings**

#### **4.1. The valuation of CSR**

Models 1 and 2 of **Table 4** report the results of the cross-sectional regressions of 1-month forward Tobin's Q on CSR as shown in Eqs. (1) and (2), while Models 3 and 4 report the results of 2-month forward Tobin's Q on CSR.

Observing Models 1 through 4 of **Table 4**, we note that our results suggest that the aggregation of CSR dimensions has a confounding effect when examining the CSR-valuation relation. Specifically, Model 1 shows that the firm value and the overall CSR score are statistically significant at the 1% level but economically small. One standard deviation increase in the overall


This table displays full sample regression estimates of 1- and 2-month forward Tobin's Q on CSR from 2009 to 2014. The main independent variables are the firm's (lagged) environment (E), social (S), and governance (G) scores. Refer to Appendix A for variable definitions. Regressions include industry and country dummies as indicated. Models 1 and 2 report estimates of Fama-MacBeth [54] regressions of 1-month forward Tobin's Q on aggregated and disaggregated CSR. Models 3 and 4 report estimates of Fama-MacBeth [54] regressions of 2-month forward Tobin's Q on aggregated and disaggregated CSR. Standard errors are clustered on year and country. T-Statistics are reported in parentheses. \*, \*\*, and \*\*\* indicate the significance level at the 10, 5, and 1%, respectively.

**Table 4.** The relationship between CSR and firm value.

*CSRi,n,t-1 × IFVi,t* is the individual dimension index measures of CSR for firm *i* relative to dimen-

This table displays both the full sample and subsample (i.e., developed/emerging market) summary statistics for key

**Variable Obs. Mean Std. Min. Max.** Tobin's Q 134,823 1.63 0.93 0.83 5.07 ESG 134,823 43.62 28.77 1 100 E 134,823 52.62 33.30 1 100 S 134,823 52.21 32.13 1 100 G 134,823 44.96 28.40 1 100

CINV 134,823 21.93 6.52 10.86 48.66 FREE 134,823 71.78 8.54 50.30 90 GDPPC 134,823 10.36 0.67 3.02 11.27 GFDDB 134,823 104.80 58.19 17.28 302.74 GFDDP 134,823 52.59 35.66 0.05 193.41 GFDDS 134,823 94.79 50.85 15.17 524.41 WGIRQ 134,823 81.60 16.36 26.32 100 WGIRL 134,823 82.09 17.76 23.70 100 WGIGE 134,823 84.28 14.75 19.62 100 WGIPS 134,823 62.77 21.25 5.19 98.58 WGICC 134,823 80.21 19.15 11.48 100 WGIVA 134,823 77.07 20.65 4.74 100

Models 1 and 2 of **Table 4** report the results of the cross-sectional regressions of 1-month forward Tobin's Q on CSR as shown in Eqs. (1) and (2), while Models 3 and 4 report the results

Observing Models 1 through 4 of **Table 4**, we note that our results suggest that the aggregation of CSR dimensions has a confounding effect when examining the CSR-valuation relation. Specifically, Model 1 shows that the firm value and the overall CSR score are statistically significant at the 1% level but economically small. One standard deviation increase in the overall

sion *n* (i.e., environment, social, governance) interacted with the dummy.

**4. Empirical findings**

variables for the time period of 2009 to 2014.

**Table 3.** Summary statistics of the main variables.

*Country-level institutional void measures*

86 Firm Value - Theory and Empirical Evidence

**4.1. The valuation of CSR**

of 2-month forward Tobin's Q on CSR.

CSR score is related to an increase in Tobin's Q of about 0.0029, representing an increase of about 0.18% from the mean of 1.63.

Model 2 shows that the three subdimensions have different relations to the firm value. The environmental CSR has a statistically significant positive effect on firm value. These results are in line with prior findings (e.g., [2, 13, 14, 20, 51, 55]). Anecdotally, we also note that actors in the global business environment (i.e., policy-makers, activists, etc.) have long argued for the importance of environmental performance for shareholders, drawing significant attention to corporate environmental conscientiousness (e.g., the toughening of oil sands rules in Canada4 , China's renewed pledge to fight smog post-release of the viral documentary "Under the Dome,"5 and America's continued push for carbon emission reduction6 ). One standard deviation increase in environmental CSR is related to an increase in Tobin's Q of about 0.02, representing an increase of about 1.4% from the mean of 1.63.

<sup>4</sup> http://www.bloomberg.com/news/articles/2015-03-13/oil-sands-rules-get-tougher-as-alberta-seeks-less-damage, retrieved on 30 March 2015

<sup>5</sup> http://www.bloomberg.com/news/articles/2015-03-07/china-pollution-film-vanishes-as-xi-makes-pledge-on-environment, retrieved on 30 March 2015

<sup>6</sup> http://www.bloomberg.com/politics/articles/2015-03-19/obama-orders-40-reduction-in-carbon-emissions-by-u-s-agencies, retrieved on 30 March 2015

For social CSR, Model 2 reports a statistically significant negative effect in Tobin's Q. This finding is similar to the prior results. Indeed, Brammer, Brooks, and Pavelin [56] find a negative relation between social CSR and market value. A possible explanation for this result is the view that CSR has the potential to materialize as future benefits (e.g., [7, 43]) after stakeholders recognize that firm behavior as being genuine implies that firms have to consistently pursue socially responsible initiatives in subsequent periods before they are rewarded (e.g., [26]; Greening and Turban 2000).

*Economic development Financial market development IFV =* **CINV FREE GDPPC GFDDB GFDDP GFDDS** E 0.0004\*\*\* 0.0003\*\*\* 0.0003\*\*\* 0.0003\*\*\* 0.0004\*\*\* 0.0006\*\*\*

S −0.0005\*\*\* −0.0004\*\*\* −0.0005\*\*\* −0.0005\*\*\* −0.0004\*\*\* −0.0003\*\*\*

G −0.0001 −0.0005\*\*\* −0.0005\*\*\* −0.0002\*\*\* 0.0000 −0.0003\*\*\*

IFV 1.0252\*\*\* −0.1720 0.2777\* 0.7313\*\*\* 0.7070\*\*\* −0.1098

E × IFV 0.0009\*\*\* 0.0018\*\*\* 0.0022\*\*\* 0.0012\*\*\* 0.0020\*\*\* 0.0003\*

S × IFV 0.0007\*\*\* 0.0005\*\*\* 0.0011\*\*\* 0.0007\*\*\* 0.0009\*\*\* −0.0001

G × IFV −0.0003\* 0.0014\*\*\* 0.0020\*\*\* 0.0003\* −0.0010\*\*\* 0.0012\*\*\*

Obs. 134,823 134,823 134,823 134,823 134,823 134,823 R-Squared 0.71 0.71 0.71 0.71 0.71 0.71

Control Yes Yes Yes Yes Yes Yes Ind. dum Yes Yes Yes Yes Yes Yes Ctr. dum Yes Yes Yes Yes Yes Yes

*IFV =* **WGIRQ WGIRL WGIGE WGIPS WGICC WGIVA** E 0.0002\*\*\* 0.0002\*\*\* 0.0003\*\*\* 0.0003\*\*\* 0.0002\*\*\* 0.0001\*\*\*

S −0.0005\*\*\* −0.0005\*\*\* −0.0005\*\*\* −0.0006\*\*\* −0.0006\*\*\* −0.0006\*\*\*

G −0.0005\*\*\* −0.0005\*\*\* −0.0005\*\*\* −0.0004\*\*\* −0.0005\*\*\* −0.0006\*\*\*

IFV 0.0904 0.2355 0.3498\*\* 0.5049\*\*\* 0.1936 0.3543\*\*

(6.43) (6.86) (8.14) (6.14) (6.49) (2.86)

(−12.62) (−12.56) (−13.65) (−14.26) (−13.40) (−15.88)

(−14.02) (−15.34) (−15.09) (−9.22) (−13.99) (−13.60)

(0.54) (1.44) (2.27) (3.14) (1.15) (2.36)

*Government quality*

(10.94) (7.02) (7.16) (8.09) (11.34) (18.15)

Can Corporate Social Responsibility Fill Institutional Voids?

http://dx.doi.org/10.5772/intechopen.76164

89

(−11.32) (−10.28) (−12.22) (−12.43) (−11.36) (−6.70)

(−1.21) (−13.38) (−15.31) (−4.02) (0.23) (−7.51)

(8.30) (−0.93) (1.75) (5.08) (5.48) (−0.77)

(11.80) (21.26) (23.70) (12.91) (27.23) (1.95)

(6.54) (4.65) (8.43) (7.17) (8.40) (−0.37)

(−1.92) (8.99) (11.43) (1.81) (−4.99) (5.07)

For governance CSR, Model 2 reports a statistically significant but economically negligible negative effect in Tobin's Q. This result is also in line with prior findings (e.g., [13, 14, 24]). In particular, Cheng et al. [14] postulate that the weaker effect of corporate governance stems from the fact that the main driver of corporate governance is the country-level institutional structures that firms operate in.

We also include all the control variables (including LROAW, LLEVW, LCAPXW, LCASHW, LSGRW, LADW, LASSET, LDDUM), industry, and country-fixed effects (e.g., [57]). In unreported results, we find that the effects of our controls are similar to the findings in the literature (e.g., [13, 23]). Specifically, across Models 1 to 4, we find a positive relation with return on assets, leverage, capital expenditure, cash, sales growth, advertising expenditure, and a negative relation with firm size and dividend payout.

In summary, these results provide empirical support for our first two hypotheses, whereby CSR creates value for the firm on average and that the CSR-valuation relation is heterogeneous in nature and CSR dimension is dependent, such that there is significant heterogeneity in valuation effects across different groups of stakeholders.

### **4.2. The moderating effect of institutional voids on the CSR-firm value relation**

Next, we investigate how the CSR-firm value relation changes in the context of different institutional frameworks by modeling the marginal valuation effect of CSR in the presence of institutional voids. **Table 5** reports the results of the cross-sectional regressions of Tobin's Q on CSR with the inclusion of institutional framework dummies and their interaction terms as shown in Eq. (3).

Across all models with different specifications of institutional void (*IFV*), we observe that the CSR-firm value relation (i.e., firms in strong institutional frameworks) is generally consistent with our earlier findings. All coefficients except one for environmental, social, and governance CSR remain generally statistically significant at the 1% level.

In line with our expectations, we find significant differences in the CSR-firm value relation across institutional frameworks in our institutional void analysis. For environmental CSR, all regression models show that environmental CSR has a statistically significant positive effect on firm value for firms in weak institutional frameworks. On average, one standard deviation increase in environmental CSR predicts an increase in Tobin's Q of about 0.067, representing an increase of about 4.1% (given the mean is at 1.63). The average effect is


For social CSR, Model 2 reports a statistically significant negative effect in Tobin's Q. This finding is similar to the prior results. Indeed, Brammer, Brooks, and Pavelin [56] find a negative relation between social CSR and market value. A possible explanation for this result is the view that CSR has the potential to materialize as future benefits (e.g., [7, 43]) after stakeholders recognize that firm behavior as being genuine implies that firms have to consistently pursue socially responsible initiatives in subsequent periods before they are rewarded (e.g.,

For governance CSR, Model 2 reports a statistically significant but economically negligible negative effect in Tobin's Q. This result is also in line with prior findings (e.g., [13, 14, 24]). In particular, Cheng et al. [14] postulate that the weaker effect of corporate governance stems from the fact that the main driver of corporate governance is the country-level institutional

We also include all the control variables (including LROAW, LLEVW, LCAPXW, LCASHW, LSGRW, LADW, LASSET, LDDUM), industry, and country-fixed effects (e.g., [57]). In unreported results, we find that the effects of our controls are similar to the findings in the literature (e.g., [13, 23]). Specifically, across Models 1 to 4, we find a positive relation with return on assets, leverage, capital expenditure, cash, sales growth, advertising expenditure, and a

In summary, these results provide empirical support for our first two hypotheses, whereby CSR creates value for the firm on average and that the CSR-valuation relation is heterogeneous in nature and CSR dimension is dependent, such that there is significant heterogeneity

Next, we investigate how the CSR-firm value relation changes in the context of different institutional frameworks by modeling the marginal valuation effect of CSR in the presence of institutional voids. **Table 5** reports the results of the cross-sectional regressions of Tobin's Q on CSR with the inclusion of institutional framework dummies and their interaction terms as

Across all models with different specifications of institutional void (*IFV*), we observe that the CSR-firm value relation (i.e., firms in strong institutional frameworks) is generally consistent with our earlier findings. All coefficients except one for environmental, social, and gover-

In line with our expectations, we find significant differences in the CSR-firm value relation across institutional frameworks in our institutional void analysis. For environmental CSR, all regression models show that environmental CSR has a statistically significant positive effect on firm value for firms in weak institutional frameworks. On average, one standard deviation increase in environmental CSR predicts an increase in Tobin's Q of about 0.067, representing an increase of about 4.1% (given the mean is at 1.63). The average effect is

**4.2. The moderating effect of institutional voids on the CSR-firm value relation**

[26]; Greening and Turban 2000).

88 Firm Value - Theory and Empirical Evidence

structures that firms operate in.

shown in Eq. (3).

negative relation with firm size and dividend payout.

in valuation effects across different groups of stakeholders.

nance CSR remain generally statistically significant at the 1% level.


average, one standard deviation increase in governance CSR predicts an increase in Tobin's Q of about 0.02, representing an increase of about 1.3% (given the mean is at 1.63). In the strong institutional frameworks, the effect is about −0.6% in Tobin's Q with a one standard deviation

Can Corporate Social Responsibility Fill Institutional Voids?

http://dx.doi.org/10.5772/intechopen.76164

91

Observing the pattern of coefficient significance, we note that our results suggest that the positive valuation effect of environmental and governance CSR is driven by the country's economic and financial sector development, while the positive return effect of social CSR is driven by the country's quality of law and government effectiveness. The degree of the variation is likely caused by the inherently complex and multidimensional nature of

In summary, we find support for our third hypothesis, whereby the CSR-valuation relation is moderated by the institutional frameworks that firms operate in, such that the presence of greater (lesser) institutional voids in financial, economic, and governmental institutions will

We perform two robustness tests. First, we examine the longevity of value creation attributed to CSR to test if our observations are driven by market reaction rather than material value creation. According to theory, CSR should create long-term value for the firm, and as such, we expect that there are no reversions in firm value over a longer time frame. In these specifications, we re-estimate regression specifications (1), (2), and (3) by using 3-month forward values of Tobin's Q. Our results are robust with different forward measures of the firm valuation. This suggests that our observations are likely not driven by market over- or under-reaction.

Second, we examine the possibility that our findings are driven by firms in regulated industries or "sin" stocks (e.g., [3]). As such, we rerun our models excluding firms in regulated industries, which constitute banking, energy, insurance, telecommunication, transportation, and utility companies, and those under the Triumvirate of Sin (e.g., [23]), which constitute

We find that our results are similar and lend themselves to the same conclusions and omit these results for brevity. This test provides evidence for the robustness of our results and suggests that the underlying mechanism driving the CSR-firm value relation is not likely due to

This study advances the ongoing research on the effect of CSR on firm value by integrating an institution-based view with an institutional void perspective. We draw on institutional void theory to argue for country-level institutional frameworks as a systemic, institutional-level

increase in the governance CSR.

result in a greater (lesser) valuation effect.

alcohol, gambling, and tobacco companies.

firms in regulated industries or "sin" firms.

governance.

**4.3. Robustness tests**

**5. Conclusion**

This table displays full sample regression estimates of Tobin's Q on CSR from 2009 to 2014. The main independent variables are the firm's (lagged) environment (E), social (S), and governance (G) scores. The interaction effect models the marginal valuation effect of CSR in the presence of institutional voids across 12 different measures of institutional framework strength. Regressions include industry and country dummies as indicated. T-Statistics are reported in parentheses. \*, \*\*, and \*\*\* indicate the significance level at the 10, 5, and 1%, respectively.

**Table 5.** The link between institutional environment, CSR, and Tobin's Q.

about 0.6% in the strong institutional frameworks. Hence, it indicates an increase of 3.5% in Tobin's Q.

For social CSR, all regressions show a statistically significant positive effect on firm value for firms in weak institutional frameworks. In addition, the interaction between social CSR and weak institutional frameworks is positive such that the joint effect transforms the negative base case effect into a positive one. Interestingly, this suggests that the market recognizes the benefit to the firm upon filling these institutional voids and, thus, actively rewards firms who are working to fill them. On average, one standard deviation increase in social CSR predicts an increase in Tobin's Q of about 0.015, representing an increase of about 0.9% (given the mean is at 1.63). In the strong institutional frameworks, the effect is about −0.9% in Tobin's Q with a one standard deviation increase in the social CSR.

For governance CSR, most regressions show that governance CSR generally has a statistically significant positive effect on firm value for firms in weak institutional frameworks. Similarly, the significant and positive effect of governance CSR also suggests that the market recognizes and rewards firms in weak institutional frameworks who work to fill institutional voids. On average, one standard deviation increase in governance CSR predicts an increase in Tobin's Q of about 0.02, representing an increase of about 1.3% (given the mean is at 1.63). In the strong institutional frameworks, the effect is about −0.6% in Tobin's Q with a one standard deviation increase in the governance CSR.

Observing the pattern of coefficient significance, we note that our results suggest that the positive valuation effect of environmental and governance CSR is driven by the country's economic and financial sector development, while the positive return effect of social CSR is driven by the country's quality of law and government effectiveness. The degree of the variation is likely caused by the inherently complex and multidimensional nature of governance.

In summary, we find support for our third hypothesis, whereby the CSR-valuation relation is moderated by the institutional frameworks that firms operate in, such that the presence of greater (lesser) institutional voids in financial, economic, and governmental institutions will result in a greater (lesser) valuation effect.

#### **4.3. Robustness tests**

We perform two robustness tests. First, we examine the longevity of value creation attributed to CSR to test if our observations are driven by market reaction rather than material value creation. According to theory, CSR should create long-term value for the firm, and as such, we expect that there are no reversions in firm value over a longer time frame. In these specifications, we re-estimate regression specifications (1), (2), and (3) by using 3-month forward values of Tobin's Q. Our results are robust with different forward measures of the firm valuation. This suggests that our observations are likely not driven by market over- or under-reaction.

Second, we examine the possibility that our findings are driven by firms in regulated industries or "sin" stocks (e.g., [3]). As such, we rerun our models excluding firms in regulated industries, which constitute banking, energy, insurance, telecommunication, transportation, and utility companies, and those under the Triumvirate of Sin (e.g., [23]), which constitute alcohol, gambling, and tobacco companies.

We find that our results are similar and lend themselves to the same conclusions and omit these results for brevity. This test provides evidence for the robustness of our results and suggests that the underlying mechanism driving the CSR-firm value relation is not likely due to firms in regulated industries or "sin" firms.
