Accrual-based, real activities earnings management and corporate social responsibility: A virtuous circle? emerging market evidence

Figures & data

Figure 1. Research framework.

Table 1. Sample collection

Year	Basic Materials	Consumer Cyclicals	Consumer Non-Cyclicals	Energy	Healthcare	Industrials	Technology	Utilities
2014	56	39	12	7	9	76	7	19
2015	56	39	12	7	9	76	7	19
2016	56	39	12	7	9	76	7	19
2017	56	39	12	7	9	76	7	19
2018	56	39	12	7	9	76	7	19
Grand Total	280	195	60	35	45	380	35	95

Table 2. Descriptive statistics of variables

Variable	Obs	Mean	Std. Dev.	Min	Max
CSR_all	862	.2794417	.1467802	0	.9393939
aem_jones	862	.0198783	.1392275	−.5362092	1.231.337
aem_kothari	862	.0137309	.1386722	−.5147431	1.238.111
r_cfo	862	.0101958	.1543661	−1.110.523	.8216203
r_prod	862	−.0138581	.176287	−.6085648	1.729.609
r_disx	862	.001026	.0787152	−.1895448	.5085886
rm_proxy	862	−.0001438	.1496032	−1.112.264	1.834.393
rev_grow	862	.1956827	.7197505	−.8509265	920.336
size	862	28.28428	1.277.746	2.473.013	3.199.059
lev	862	.2505155	.1883887	0	.7356519
roa	862	.0991352	.0909397	−.2894412	.5891411
age	862	2.810525	.4473604	1.386.294	4.779.123

Table 3. Pearson’s correlation coefficient matrix

	CSR_all	aem_jones	aem_kothari	r_cfo	r_prod	r_disx	rm_proxy	rev_grow	size	lev	roa	age	VIF
CSR_all	1												1.15
aem_jones	−0.022	1											1.04
aem_kothari	−0.036	0.981*	1										1.02
r_cfo	0.035	−0.719*	−0.790*	1									1.32
r_prod	−0.114*	0.134*	0.241*	−0.541*	1								1.62
r_disx	0.133*	0.011	−0.038	0.172*	−0.521*	1							1.20
rm_proxy	−0.021	−0.599*	−0.588*	0.560*	0.287*	0.121*	1						1.11
rev_grow	−0.084*	0.037	0.046	0.114*	0.056	0.116*	0.288*	1					1.03
size	0.277*	−0.004	0.033	−0.137*	0.193*	−0.090*	−0.005	−0.013	1				1.28
lev	0.002	0.015	0.086*	−0.280*	0.303*	−0.214*	−0.072	−0.031	0.348*	1			1.38
roa	0.109*	0.135*	0.005	0.445*	−0.601*	0.260*	−0.055	0.006	−0.242*	−0.450*	1		1.35
age	0.119*	−0.005	−0.011	−0.019	−0.114*	0.220*	−0.050	−0.152*	0.008	0.061	0.024	1	1.04

Table 4. The OLS and FE regression models used to assess Granger causality between AEM and CSR (H1a)

Variables	OLS	OLS	FE	FE
	CSR_all	CSR_all	CSR_all	CSR_all
aem_jones	−0.039**		−0.034*
	[−2.00]		[−1.74]
L.aem_jones	−0.043***		−0.044***
	[−2.65]		[−2.71]
aem_kothari		−0.037*		−0.034*
		[−1.89]		[−1.74]
L.aem_kothari		−0.044***		−0.044***
		[−2.68]		[−2.70]
L.CSR_all	0.841***	0.841***	0.225***	0.225***
	[41.50]	[41.49]	[6.62]	[6.63]
rev_grow	−0.001	−0.001	−0.001	−0.001
	[−0.29]	[−0.29]	[−0.43]	[−0.42]
size	0.010***	0.010***	0.017*	0.016
	[4.37]	[4.36]	[1.67]	[1.61]
lev	0.01	0.011	−0.02	−0.02
	[0.61]	[0.65]	[−0.48]	[−0.47]
roa	0.095***	0.078**	0.077	0.075
	[2.76]	[2.30]	[1.21]	[1.20]
age	−0.009	−0.009	0.168***	0.169***
	[−1.42]	[−1.45]	[4.53]	[4.56]
_cons	−0.210***	−0.208***	−0.727***	−0.713***
	[−3.06]	[−3.03]	[−2.83]	[−2.78]
N	862	862	862	862
R-sq	0.713	0.713	0.223	0.222

Note: *, **, *** denote the level of significance of 10%; 5% and 1% respectively;

Table 5. The OLS and FE regression models used to assess Granger causality between CSR and AEM (H1b)

Variables	Model 1	Model 2	Model 3	Model 4
	aem_jones	aem_kothari	aem_jones	aem_kothari
L.CSR_all	0.045	0.043	0.1	0.099
	[0.74]	[0.69]	[1.41]	[1.41]
CSR_all	−0.118**	−0.113*	−0.140*	−0.139*
	[−2.00]	[−1.89]	[−1.74]	[−1.74]
L.aem_jones	−0.016		−0.242***
	[−0.55]		[−7.60]
rev_grow	0.009	0.01	0.001	0.002
	[1.39]	[1.53]	[0.09]	[0.26]
size	0.002	0.002	0.02	0.014
	[0.57]	[0.48]	[0.95]	[0.70]
lev	0.073**	0.084***	0.390***	0.418***
	[2.53]	[2.89]	[4.67]	[5.06]
roa	0.353***	0.128**	0.957***	0.871***
	[6.02]	[2.16]	[7.77]	[7.13]
age	−0.007	−0.009	0.019	0.039
	[−0.67]	[−0.84]	[0.25]	[0.51]
L.aem_kothari		−0.037		−0.265***
		[−1.29]		[−8.49]
_cons	−0.06	−0.033	−0.758	−0.671
	[−0.50]	[−0.27]	[−1.45]	[−1.29]
N	862	862	862	862
R-sq	0.047	0.022	0.171	0.178

Note: *, **, *** denote the level of significance of 10%; 5% and 1% respectively;

Table 6. The OLS and FE regression models used to assess Granger causality between REM and CSR (H2a)

Variables	OLS	OLS	OLS	OLS	FE	FE	FE	FE
	CSR_all	CSR_all	CSR_all	CSR_all	CSR_all	CSR_all	CSR_all	CSR_all
r_cfo	0.034*				0.038*
	[1.69]				[1.92]
L.r_cfo	0.024				0.044**
	[1.41]				[2.53]
r_prod		−0.033				−0.035
		[−1.37]				[−1.32]
L.r_prod		−0.016				−0.043*
		[−0.77]				[−1.66]
r_disx			0.201***				0.155*
			[2.63]				[1.78]
L.r_disx			−0.118*				−0.033
			[−1.65]				[−0.42]
rm_proxy				0.042*				0.038*
				[1.91]				[1.72]
L.rm_proxy				0.02				0.037*
				[1.17]				[1.88]
L.CSR_all	0.844***	0.841***	0.839***	0.844***	0.222***	0.197***	0.222***	0.197***
	[41.60]	[40.23]	[41.17]	[40.50]	[6.54]	[5.63]	[6.49]	[5.63]
rev_grow	−0.003	0	−0.005	−0.005	−0.003	0.001	−0.004	−0.003
	[−0.73]	[−0.01]	[−1.28]	[−0.90]	[−0.86]	[0.20]	[−1.11]	[−0.59]
size	0.010***	0.011***	0.010***	0.010***	0.015	0.014	0.015	0.013
	[4.27]	[4.26]	[4.38]	[4.18]	[1.45]	[1.19]	[1.50]	[1.09]
lev	0.009	0.004	0.009	0.005	−0.014	−0.052	−0.04	−0.033
	[0.54]	[0.25]	[0.53]	[0.31]	[−0.32]	[−1.23]	[−0.98]	[−0.75]
roa	0.037	0.032	0.066*	0.090**	0.03	0.003	0.054	0.048
	[0.94]	[0.74]	[1.92]	[2.54]	[0.50]	[0.05]	[0.88]	[0.75]
age	−0.009	−0.011*	−0.012*	−0.009	0.177***	0.195***	0.166***	0.204***
	[−1.44]	[−1.69]	[−1.92]	[−1.32]	[4.76]	[4.89]	[4.45]	[5.10]
_cons	−0.201***	−0.207***	−0.198***	−0.213***	−0.687***	−0.712**	−0.666***	−0.708**
	[−2.91]	[−2.81]	[−2.87]	[−2.90]	[−2.68]	[−2.33]	[−2.59]	[−2.31]
N	863	817	863	817	863	817	863	817
R-sq	0.711	0.71	0.713	0.71	0.223	0.218	0.217	0.219

Note: . *, **, *** denote the level of significance of 10%; 5% and 1% respectively;

Table 7. The OLS and FE regression models used to assess Granger causality between CSR and REM (H2b)

Variables	OLS	OLS	OLS	OLS	FE	FE	FE	FE
	r_cfo	r_prod	r_disx	rm_proxy	r_cfo	r_prod	r_disx	rm_proxy
L.CSR_all	−0.068	0.02	−0.016	−0.073	−0.076	−0.014	0.001	−0.076
	[−1.12]	[0.38]	[−1.04]	[−1.27]	[−1.09]	[−0.25]	[0.06]	[−1.14]
CSR_all	0.100*	−0.071	0.040***	0.106*	0.152*	−0.084	0.032*	0.130*
	[1.69]	[−1.37]	[2.63]	[1.91]	[1.92]	[−1.32]	[1.78]	[1.72]
L.r_cfo	0.035				−0.246***
	[1.20]				[−7.47]
L.r_prod		0.325***				−0.209***
		[11.58]				[−5.39]
L.r_disx			0.812***				0.201***
			[51.78]				[5.72]
L.rm_proxy				−0.036				−0.286***
				[−1.31]				[−8.38]
rev_grow	0.021***	0.009	0.015***	0.079***	0.030***	0.003	0.017***	0.079***
	[3.17]	[1.23]	[9.18]	[9.73]	[4.38]	[0.44]	[10.97]	[8.67]
size	0.001	0.002	0	0	0.008	0.061***	−0.009**	0.039*
	[0.17]	[0.59]	[−0.45]	[−0.06]	[0.39]	[3.34]	[−1.97]	[1.75]
lev	−0.058**	0.014	−0.008	−0.068**	−0.407***	−0.110*	−0.009	−0.539***
	[−2.02]	[0.54]	[−1.12]	[−2.48]	[−4.94]	[−1.67]	[−0.47]	[−6.90]
roa	0.689***	−0.743***	0.034**	−0.201***	0.117	−0.706***	−0.029	−0.641***
	[10.92]	[−12.75]	[2.19]	[−3.62]	[0.96]	[−7.39]	[−1.05]	[−5.59]
age	0.012	−0.031***	0.009***	0.003	−0.104	−0.06	0.002	−0.152**
	[1.18]	[−3.16]	[3.29]	[0.31]	[−1.38]	[−0.96]	[0.09]	[−2.03]
_cons	−0.113	0.099	−0.026	0.007	0.145	−1.456***	0.245**	−0.496
	[−0.95]	[0.90]	[−0.85]	[0.06]	[0.28]	[−3.10]	[2.08]	[−0.87]
N	863	817	863	817	863	817	863	817
R-sq	0.228	0.498	0.801	0.12	0.131	0.143	0.201	0.271

Note: *, **, *** denote the level of significance of 10%; 5% and 1% respectively;

Table 8. Bootstrap test analysis

Dependent Variable	Independent Variable	Coef.	S.E.	Z-value	P>Z	LL 95% CI	UL 95% CI
CSR_all	aem_jones	−0.039	0.017	−2.28	0.022	−0.073	−0.006
	L. aem_jones	−0.043	0.018	−2.41	0.016	−0.078	−0.008
	aem_kothari	−0.037	0.017	−2.14	0.032	−0.071	−0.003
	L. aem_jones	−0.044	0.018	−2.43	0.015	−0.079	−0.008
	r_cfo	0.034	0.018	1.87	0.061	−0.002	0.069
	L.r_cfo	0.024	0.020	1.2	0.232	−0.015	0.062
	r_prod	−0.033	0.023	−1.39	0.165	−0.078	0.013
	L.r_prod	−0.016	0.020	−0.81	0.417	−0.054	0.022
	r_disx	0.201	0.090	2.23	0.026	0.024	0.378
	L.r_disx	−0.118	0.084	−1.4	0.163	−0.283	0.047
	rm_proxy	0.042	0.020	2.14	0.033	0.004	0.081
	L.rm_proxy	0.020	0.017	1.17	0.242	−0.014	0.055
aem_jones	CSR_all	−0.118	0.049	−2.43	0.015	−0.213	−0.023
	L. CSR_all	0.045	0.051	0.89	0.376	−0.055	0.145
aem_kothari	CSR_all	−0.113	0.050	−2.23	0.026	−0.212	−0.014
	L. CSR_all	0.043	0.053	0.8	0.425	−0.062	0.147
r_cfo	CSR_all	0.100	0.051	1.94	0.052	−0.001	0.201
	L. CSR_all	−0.068	0.055	−1.24	0.213	−0.176	0.039
r_prod	CSR_all	−0.071	0.046	−1.55	0.121	−0.160	0.019
	L. CSR_all	0.020	0.048	0.43	0.67	−0.074	0.115
r_disx	CSR_all	0.040	0.017	2.43	0.015	0.008	0.072
	L. CSR_all	−0.016	0.015	−1.09	0.275	−0.046	0.013
rm_proxy	CSR_all	0.106	0.049	2.17	0.03	0.010	0.203
	L. CSR_all	−0.073	0.052	−1.4	0.161	−0.174	0.029

Table 9. Robustness results for testing all hypotheses

Variables	OLS	FE	OLS	FE
	CSR_all	CSR_all	EM_all	EM_all
EM_all	−0.041*	−0.032
	[−1.94]	[−1.46]
CSR_all			−0.112*	−0.114
			[−1.94]	[−1.46]
L.EM_all	−0.033**	−0.036**	−0.045*	−0.249***
	[−2.09]	[−2.08]	[−1.73]	[−8.12]
L.CSR_all	0.842***	0.199***	0.044	0.093
	[40.40]	[5.67]	[0.73]	[1.37]
rev_grow	−0.003	−0.001	−0.039***	−0.037***
	[−0.54]	[−0.22]	[−4.62]	[−3.96]
size	0.011***	0.014	0.003	−0.027
	[4.26]	[1.20]	[0.68]	[−1.19]
lev	0.006	−0.04	0.084***	0.545***
	[0.34]	[−0.91]	[2.96]	[6.79]
roa	0.146***	0.083	1.071***	1.632***
	[3.40]	[1.16]	[17.71]	[13.91]
age	−0.008	0.198***	0.002	0.111
	[−1.24]	[4.98]	[0.17]	[1.45]
_cons	−0.224***	−0.730**	−0.158	0.201
	[−3.04]	[−2.38]	[−1.30]	[0.35]
N	816	816	816	816
R-sq	0.711	0.219	0.323	0.329

Table 10. 2SLS results for hypothesis H1a and H2a

Variables	CSR_all	CSR_all	CSR_all
aem_jones	−0.037*
	[−1.68]
L.aem_jones	−0.025
	[−1.37]
aem_kothari		−0.037*
		[−1.68]
L.aem_kothari		−0.027
		[−1.43]
rm_proxy			0.057**
			[2.17]
L.rm_proxy			0.005
			[0.24]
L.CSR_all	0.916***	0.915***	0.928***
	[31.52]	[31.43]	[31.70]
rev_grow	0.005	0.005	0.004
	[1.27]	[1.28]	[0.69]
size	0.009***	0.009***	0.008**
	[3.14]	[3.16]	[2.52]
lev	−0.007	−0.006	−0.011
	[−0.36]	[−0.34]	[−0.54]
roa	0.035	0.022	0.022
	[0.95]	[0.61]	[0.56]
age	−0.008	−0.008	−0.009
	[−1.12]	[−1.13]	[−1.23]
_cons	−0.215***	−0.215***	−0.171**
	[−2.65]	[−2.64]	[−1.97]
N	656	656	596
R-sq	0.760	0.760	0.769
Hausman test for endogeneity	Fail to reject	Fail to reject	Fail to reject
Sargan test for overidentification	Fail to reject	Fail to reject	Fail to reject

Note: *, **, *** denote the level of significance of 10%; 5% and 1% respectively;

Table 11. 2SLS results for hypothesis H1b and H2b

Variables	aem_jones	aem_kothari	rm_proxy
CSR_all	−0.191**	−0.198**	0.148*
	[−2.20]	[−2.23]	[1.91]
L.CSR_all	0.061	0.050	−0.080
	[0.67]	[0.54]	[−0.99]
L.aem_jones	−0.637***
	[−11.96]
L.aem_kothari		−0.695***
		[−12.69]
L.rm_proxy			−0.651***
			[−12.59]
rev_grow	0.008	0.010	0.103***
	[0.91]	[1.07]	[9.32]
size	−0.004	−0.003	0.007
	[−0.64]	[−0.46]	[1.31]
lev	0.149***	0.163***	−0.093**
	[3.61]	[3.88]	[−2.52]
roa	0.459***	0.111	−0.264***
	[5.72]	[1.36]	[−3.72]
age	−0.034**	−0.036**	0.004
	[−2.20]	[−2.27]	[0.28]
_cons	0.152	0.163	−0.191
	[0.83]	[0.87]	[−1.16]
N	637	637	596
R-sq	0.760	0.760	0.769
Hausman test for endogeneity	Fail to reject	Fail to reject	Fail to reject
Sargan test for overidentification	Fail to reject	Fail to reject	Fail to reject

Note: *, **, *** denote the level of significance of 10%; 5% and 1% respectively;

Table 12. FsQCA findings regarding CSR-EM relationship

	Configurations	AEM	REM	REV_GROW	SIZE	LEV	ROA	AGE
CSR	1	●	~
	2	~	●		~
	3	~			●		~
	4	~				●	~
	5		●			~	●
	6	●			●		●
	7			●	~		~	~
	8		●	~			●	●
	9		~	~		~
	10	~		~		~
	11	~	●					~
	12		●		●		●	~
	Solution coverage	0.659071
	Solution consistency	0.998065

Note: ● indicate the presence of causal conditions. ~ indicate the absence or negation of causal conditions. The blank cells indicate “don´t care conditions” (Kraus et al., 2017)

Table 13. FsQCA results in terms of EM-CSR relationship

	Configurations	CSR	REV_GROW	SIZE	LEV	ROA	AGE
EM	1	●
	Solution coverage	0.998694
	Solution consistency	0.601341

Note: ● indicate the presence of causal conditions. ~ indicate the absence or negation of causal conditions. The blank cells indicate “don´t care conditions” (Kraus et al., 2017)

Table

Variable	Description	Source
AEM	$\frac{TA{C}_{i,t}}{A_{i,t-1}}={\beta }_1\frac{1}{A_{i,t-1}}+{\beta }_2\frac{\mathrm{\Delta }RE{V}_{i,t}}{A_{i,t-1}}+{\beta }_3\frac{PP{E}_{i,t}}{A_{i,t-1}}+{\epsilon }_{i,t}$ $\frac{TA{C}_{i,t}}{A_{i,t-1}}={\beta }_1\frac{1}{A_{i,t-1}}+{\beta }_2\frac{\mathrm{\Delta }\left(RE{V}_{i,t}-A{R}_{i,t}\right)}{A_{i,t-1}}+{\beta }_3\frac{PP{E}_{i,t}}{A_{i,t-1}}+{\beta }_{4}RO{A}_{i,t}+{\epsilon }_{i,t}$	Jones (1991); Kothari et al. (2005)
REM	$\frac{CF{O}_{i,t}}{A_{i,t-1}}={\beta }_1\frac{1}{A_{i,t-1}}+{\beta }_2\frac{RE{V}_{i,t}}{A_{i,t-1}}+{\beta }_3\frac{\mathrm{\Delta }\mathrm{R}\mathrm{E}{\mathrm{V}}_{i,t}}{A_{i,t-1}}+{\epsilon }_{i,t}(7)$ $\frac{PRO{D}_{i,t}}{A_{i,t-1}}={\beta }_1\frac{1}{A_{i,t-1}}+{\beta }_2\frac{RE{V}_{i,t}}{A_{i,t-1}}+{\beta }_3\frac{\mathrm{\Delta }\mathrm{R}\mathrm{E}{\mathrm{V}}_{i,t}}{A_{i,t-1}}+{\beta }_4\frac{\mathrm{\Delta }\mathrm{R}\mathrm{E}{\mathrm{V}}_{i,t-1}}{A_{i,t-1}}+{\epsilon }_{i,t}\left(8\right)$ $\frac{DISCEX{P}_{i,t}}{A_{i,t-1}}={\beta }_1\frac{1}{A_{i,t-1}}+{\beta }_2\frac{RE{V}_{i,t-1}}{A_{i,t-1}}+{\epsilon }_{i,t}$	Roychowdhury (2006)
CSR1	$\frac{\sum X_{i200}+\sum X_{i300}+\sum X_{i400}}{n_{i200}+n_{i300}+n_{i400}}$	Thuy et al. (2022)
REV_GROW	The ratio of current-year sales to previous-year sales	Ghaleb et al. (2021); Khuong and Anh (2022); Khuong et al. (2022); Rezaee et al. (2020); Shabbir and Wisdom (2020); Thuy et al. (2022)
SIZE	The natural logarithm of the end-of-year book value of total assets
LEV	Total debt to total assets
ROA	Return on assets
AGE	The natural logarithm of the difference between the current and foundation years

Kraus, S., Ribeiro-Soriano, D., & Schüssler, M. (2017). Fuzzy-set qualitative comparative analysis (fsQCA) in entrepreneurship and innovation research – the rise of a method. International Entrepreneurship and Management Journal, 14(1), 15–33. https://doi.org/10.1007/s11365-017-0461-8

 Web of Science ®Google Scholar

Jones, J. J. (1991). Earnings management during import relief investigations. Journal of Accounting Research, 29(2), 193–228. https://doi.org/10.2307/2491047

 Web of Science ®Google Scholar

Kothari, S. P., Leone, A. J., & Wasley, C. E. (2005). Performance matched discretionary accrual measures. Journal of Accounting and Economics, 39(1), 163–197. https://doi.org/10.1016/j.jacceco.2004.11.002

 Web of Science ®Google Scholar

Roychowdhury, S. (2006). Earnings management through real activities manipulation. Journal of Accounting and Economics, 42(3), 335–370. https://doi.org/10.1016/j.jacceco.2006.01.002

 Web of Science ®Google Scholar

Thuy, H. X., Khuong, N. V., Anh, L. H. T., & Quyen, P. N. (2022). Effect of corporate governance on corporate social responsibility in Vietnam: State-ownership as the moderating role. Journal of Financial Reporting and Accounting. https://doi.org/10.1108/jfra-10-2021-0367

 Web of Science ®Google Scholar

Ghaleb, B. A. A., Qaderi, S. A., Almashaqbeh, A., Qasem, A., & Ntim, C. G. (2021). Corporate social responsibility, board gender diversity and real earnings management: The case of Jordan. Cogent Business & Management, 8(1). https://doi.org/10.1080/23311975.2021.1883222

 Web of Science ®Google Scholar

Khuong, N. V., & Anh, L. H. T. (2022). The nexus between corporate social responsibility and firm value: The moderating role of life-cycle stages. Social Responsibility Journal, 19(5), 949–969. https://doi.org/10.1108/srj-09-2021-0370

 Web of Science ®Google Scholar

Khuong, N. V., Anh, L. H. T., & Van, N. T. H. (2022). Firm life cycle and earnings management: The moderating role of state ownership. Cogent Economics & Finance, 10(1). https://doi.org/10.1080/23322039.2022.2085260

 Web of Science ®Google Scholar

Rezaee, Z., Dou, H., & Zhang, H. (2020). Corporate social responsibility and earnings quality: Evidence from China. Global Finance Journal, 45, 45. https://doi.org/10.1016/j.gfj.2019.05.002

 Web of Science ®Google Scholar

Shabbir, M. S., & Wisdom, O. (2020). The relationship between corporate social responsibility, environmental investments and financial performance: Evidence from manufacturing companies. Environmental Science and Pollution Research, 27(32), 39946–39957. https://doi.org/10.1007/s11356-020-10217-0

 PubMed Web of Science ®Google Scholar