David versus Goliath: The Relation between Auditor Size and Audit Quality for U.K. Private Firms

Figures & data

Table 1. Sample selection.

All U.K. private firms subject to mandatory audit with non-missing total assets in year t and year t-1 and one registered auditor (2010–2016)	392,844
Less:
Financial firms (one-digit SIC code 6) and firms with missing industry classification	(77,530)
Subsidiaries of publicly listed and private firms	(165,116)
Firms with missing data to calculate AQ and ADA	(124,542)
Firms with missing data to calculate various control variables	(3,867)
Final sample	21,789

Table 2. Descriptive statistics and correlation matrix.

Panel A: Descriptive statistics
Variables	Mean	St. Deviation	25%	Median	75%
AQ	−0.044	0.038	−0.055	−0.033	−0.020
ADA	−0.081	0.093	−0.102	−0.052	−0.023
B4	0.220	0.414	0.000	0.000	0.000
GROUP	0.521	0.500	0.000	1.000	1.000
LNASSETS	16.760	1.205	15.910	16.500	17.340
LEV	0.216	0.250	0.026	0.141	0.322
ROA	0.086	0.116	0.026	0.065	0.125
ATURN	1.961	1.368	1.012	1.675	2.558
CURR	2.118	2.254	1.100	1.510	2.360
REC & INV	0.413	0.279	0.185	0.402	0.604
LOSS	0.108	0.310	0.000	0.000	0.000
GROWTH	0.103	0.367	−0.035	0.053	0.210
STDSALES	0.253	0.313	0.071	0.157	0.319
Z-SCORE	3.916	2.046	2.691	3.753	4.850
TA	−0.021	0.129	−0.079	−0.025	0.026
LNTAX_FEES	0.776	1.322	0.000	0.000	1.609
DISP	0.388	0.487	0.000	0.000	1.000
TENURE	0.229	0.420	0.000	0.000	0.000
IMPORTANCE	0.101	0.213	0.000	0.007	0.078
GAAP_ETR	−0.211	0.094	−0.261	−0.213	−0.193
CASH_ETR	−0.203	0.147	−0.277	−0.169	−0.111
BTD	0.000	0.019	−0.004	−0.001	0.002
CTD	0.001	0.011	−0.006	0.001	0.007
Panel B: Correlation matrix
Variables	AQ	ADA	B4	GROUP	LNASSETS	LEV	ROA	ATURN	CURR	REC & INV	LOSS	GROWTH	STDSALES	Z-SCORE	TA	LNTAX_FEES	DISP	TENURE	IMPORTANCE	BTD
AQ	1.00
ADA	0.37	1.00
B4	−0.18	−0.09	1.00
GROUP	0.09	0.08	−0.01	1.00
LNASSETS	−0.01	0.00	0.41	0.27	1.00
LEV	−0.10	−0.11	0.13	−0.01	0.20	1.00
ROA	−0.13	−0.13	−0.01	−0.06	−0.04	−0.19	1.00
ATURN	−0.05	−0.12	−0.11	−0.06	−0.26	−0.10	0.13	1.00
CURR	−0.05	−0.01	0.03	−0.04	0.04	−0.25	0.11	−0.20	1.00
REC & INV	0.03	−0.08	−0.13	0.00	−0.16	−0.08	0.10	0.50	−0.02	1.00
LOSS	−0.16	−0.02	0.10	0.04	0.05	0.16	−0.45	−0.10	−0.05	−0.11	1.00
GROWTH	−0.01	−0.11	−0.03	−0.01	0.01	0.04	0.21	0.38	−0.09	0.28	−0.14	1.00
STDSALES	−0.28	−0.22	−0.01	−0.07	−0.13	−0.01	0.07	0.54	−0.10	0.25	0.00	0.17	1.00
Z-SCORE	0.05	0.01	−0.14	−0.09	−0.21	−0.49	0.30	0.41	0.48	0.22	−0.22	0.09	0.20	1.00
TA	−0.05	−0.09	0.01	−0.04	0.03	0.11	0.13	0.05	0.11	0.22	−0.10	0.06	0.06	0.09	1.00
LNTAX_FEES	−0.03	0.01	0.35	0.24	0.43	0.05	−0.01	−0.04	−0.02	−0.03	0.05	0.00	−0.05	−0.07	−0.02	1.00
DISP	0.15	0.11	−0.16	0.12	−0.05	−0.10	−0.01	0.01	0.00	0.02	−0.05	0.00	−0.06	0.07	−0.03	−0.01	1.00
TENURE	−0.02	−0.02	−0.09	0.00	−0.08	0.00	0.02	0.03	0.00	0.02	0.00	0.00	0.03	0.01	0.02	−0.07	0.00	1.00
IMPORTANCE	0.07	0.02	−0.25	−0.12	−0.12	−0.05	0.02	0.13	−0.02	0.09	−0.05	0.04	0.06	0.10	0.01	−0.18	0.07	−0.05	1.00
BTD	−0.03	−0.02	0.01	−0.02	0.02	−0.04	0.34	−0.03	0.07	0.01	−0.20	0.02	−0.01	0.07	0.12	−0.02	0.02	0.02	0.00	1.00

Panel A presents summary statistics for all key variables. Panel B presents the Pearson correlation matrix. Bold values indicate significance at the two-tailed 5% level or better. See the Appendix for variable definitions.

Table 3. The relation between auditor size and audit quality for private client firms.

Panel A: Differences in audit quality between Big 4 and non-Big 4 private client firms
	(1)		(2)		(3)		(4)
	Dep. Var. = AQ		Dep. Var. = ADA		Dep. Var. = DA+		Dep. Var. = DA-
Variables	Coeff.	t-stat	Coeff.	t-stat	Coeff.	t-stat	Coeff.	t-stat
B4	−0.006***	−8.54	−0.014***	−6.08	−0.006***	−2.87	−0.011***	−6.26
GROUP	0.003***	6.11	0.009***	6.23	−0.002	−1.18	0.004***	3.58
LNASSETS	0.001***	4.92	0.002**	2.06	0.001	0.59	−0.000	−0.44
LEV	−0.005***	−4.52	−0.043***	−8.36	0.076***	12.86	0.027***	5.72
ROA	−0.042***	−20.25	−0.129***	−12.18	−0.081***	−4.71	−0.180***	−23.77
ATURN	−0.001*	−1.70	−0.003***	−2.71	−0.001	−1.26	−0.001	−0.86
CURR	−0.001***	−6.51	−0.004***	−4.87	0.001	1.37	−0.003***	−3.71
REC & INV	0.000	0.18	−0.006	−1.50	−0.039***	−9.73	−0.025***	−7.52
LOSS	−0.014***	−20.45	−0.020***	−7.74	0.013***	4.93	−0.007***	−3.38
GROWTH	0.002***	3.96	−0.011***	−3.99	0.002	0.79	−0.017***	−6.98
STDSALES	−0.020***	−25.37	−0.050***	−13.37	0.021***	5.67	−0.021***	−6.50
Z-SCORE	0.001***	5.83	0.004***	5.10	0.000	0.06	0.002***	2.74
TA	−0.002*	−1.77	−0.024*	−1.71	1.101***	70.58	1.169***	88.70
LNTAX_FEES	−0.000	−0.10	0.001	1.56	−0.000	−0.34	0.000	0.19
DISP	0.004***	7.86	0.010***	7.21	0.002**	2.02	0.005***	5.12
TENURE	−0.001	−1.25	−0.003	−1.64	0.002	1.45	0.001	1.10
IMPORTANCE	0.002**	2.34	0.006*	1.90	−0.003	−1.12	0.002	0.85
Constant	−0.054***	−12.44	−0.084***	−5.42	0.022	1.49	0.044***	3.34
Year FE	Yes		Yes		Yes		Yes
Industry FE	Yes		Yes		Yes		Yes
# of obs.	21,789		21,789		10,490		11,299
R-squared	0.255		0.125		0.642		0.661
B4	−0.010***	−9.85	−0.025***	−7.16	−0.005*	−1.65	−0.018***	−6.83
GROUP	0.002***	2.82	0.005***	3.09	−0.001	−0.88	0.002	1.23
B4×GROUP	0.008***	6.10	0.022***	5.32	−0.002	−0.50	0.014***	4.29
LNASSETS	0.001***	4.94	0.002**	2.08	0.001	0.59	−0.000	−0.40
LEV	−0.005***	−4.33	−0.043***	−8.27	0.076***	12.86	0.027***	5.76
ROA	−0.042***	−19.94	−0.129***	−12.15	−0.081***	−4.71	−0.180***	−23.75
ATURN	−0.001*	−1.71	−0.003***	−2.71	−0.001	−1.26	−0.001	−0.88
CURR	−0.001***	−6.41	−0.003***	−4.80	0.001	1.36	−0.003***	−3.65
REC & INV	−0.000	−0.27	−0.008*	−1.84	−0.039***	−9.62	−0.025***	−7.71
LOSS	−0.014***	−20.28	−0.020***	−7.62	0.013***	4.92	−0.007***	−3.31
GROWTH	0.002***	4.06	−0.010***	−3.94	0.002	0.78	−0.017***	−7.00
STDSALES	−0.020***	−25.08	−0.049***	−13.27	0.021***	5.67	−0.020***	−6.41
Z-SCORE	0.001***	5.77	0.004***	5.04	0.000	0.07	0.002***	2.71
TA	−0.002	−1.57	−0.023*	−1.66	1.101***	70.63	1.168***	88.85
LNTAX_FEES	−0.000	−0.95	0.001	0.85	−0.000	−0.27	−0.000	−0.32
DISP	0.003***	7.24	0.009***	6.56	0.002**	2.06	0.005***	4.59
TENURE	−0.001	−1.26	−0.003	−1.64	0.002	1.45	0.001	1.10
IMPORTANCE	0.002*	1.90	0.005	1.49	−0.003	−1.08	0.001	0.51
Constant	−0.053***	−12.15	−0.080***	−5.21	0.021	1.47	0.046***	3.50
Year FE	Yes		Yes		Yes		Yes
Industry FE	Yes		Yes		Yes		Yes
# of obs.	21,789		21,789		10,490		11,299
R-squared	0.260		0.127		0.642		0.662

This table presents the OLS regression results of testing H1 (Panel A) and H2 (Panel B). AQ is the Dechow and Dichev (2002) accrual quality measure, modified by McNichols (2002), multiplied by −1; ADA is the absolute value of discretionary accruals as in Kothari et al. (2005), multiplied by −1; DA+ (DA-) represents income-increasing (income-decreasing) discretionary accruals. Higher values of AQ and ADA indicate higher audit quality. GROUP equals 1 if a client firm is the parent company of a business group, and 0 if it is a standalone firm. T-statistics are calculated based on heteroscedasticity and autocorrelation robust standard errors clustered at the client-firm level. *, **, and *** indicate significance at the two-tailed 10%, 5%, and 1% levels, respectively. See the Appendix for variable definitions.

Table 4. Addressing self-selection of auditor choice.

Panel A: Auditor changes
	Changes from Big 4 to non-Big 4 relative to no auditor changes				Changes from non-Big 4 to Big 4 relative no auditor changes				Changes from Big 4 to non-Big 4 relative to changes from non-Big 4 to Big4
	(1)		(2)		(3)		(4)		(5)		(6)
	Dep. Var. =ΔAQ		Dep. Var. =ΔADA		Dep. Var. =ΔAQ		Dep. Var. =ΔADA		Dep. Var. =ΔAQ		Dep. Var. =ΔADA
Variables	Coeff.	t-stat	Coeff.	t-stat	Coeff.	t-stat	Coeff.	t-stat	Coeff.	t-stat	Coeff.	t-stat
Δ B4	−0.003*	−1.95	0.001	0.05	−0.003**	−2.05	−0.020*	−1.73	−0.003**	−2.03	−0.008	−0.67
ΔControls	Yes		Yes		Yes		Yes		Yes		Yes
Year FE	Yes		Yes		Yes		Yes		Yes		Yes
Industry FE	Yes		Yes		Yes		Yes		Yes		Yes
Observations	13,201		13,201		13,192		13,192		189		189
R−squared	0.026		0.044		0.026		0.044		0.391		0.286
Panel B: Controlling for firm fixed effects
	(1)		(2)		(3)		(4)
	Dep. Var. =ΔAQ		Dep. Var. =ΔADA		Dep. Var. =ΔAQ		Dep. Var. =ΔADA
Variables	Coeff.	t-stat	Coeff.	t-stat	Coeff.	t-stat	Coeff.	t-stat
B4	−0.002*	−1.88	−0.013**	−2.26	0.010	1.36	−0.007	−1.13
Controls	Yes		Yes		Yes		Yes
Year FE	Yes		Yes		Yes		Yes
Firm FE	Yes		Yes		Yes		Yes
# of obs.	21,789		21,789		10,490		11,299
R−squared	0.837		0.540		0.806		0.805
B4	−0.006**	−2.02	−0.010*	−1.94	0.005	1.26	−0.008***	−2.89
Controls	Yes		Yes		Yes		Yes
Year FE	Yes		Yes		Yes		Yes
Industry FE	Yes		Yes		Yes		Yes
# of obs.	21,789		21,789		10,490		11,299
R−squared	0.336		0.209		0.667		0.708

Panel A presents the OLS regressions of changes in accruals quality (ΔΑQ) and changes in absolute discretionary accruals (ΔADA) on contemporaneous changes in the Big 4 indicator (ΔB4) and changes in all control variables of equation (3). We present three alternative specifications: changes from Big 4 to non-Big 4 (ΔB4 = −1) relative to no auditor changes (ΔB4 = 0) (column 1), changes from non-Big 4 to Big 4 (ΔB4 = 1) relative no auditor changes (ΔB4 = 0) (column 2), and changes from non-Big 4 to Big 4 (ΔB4 = 1) relative to changes from Big 4 to non-Big 4 (ΔB4 = −1) (column 3). Panel B presents the OLS regression results of re-estimating model (3) after including client-firm fixed effects. Panel C presents the OLS regressions of re-estimating equation (3) for an entropy-balanced sample. We use the control variables of equation (3) as well as industry and year fixed effects as covariates to create the balanced sample. T-statistics are calculated based on heteroscedasticity and autocorrelation robust standard errors clustered at the client-firm level. *, **, and *** indicate significance at the two-tailed 10%, 5%, and 1% levels, respectively. See the Appendix for variable definitions.

Table 5. The relation between auditor size and tax avoidance for private client firms.

Panel A: Differences in tax avoidance between Big 4 and non-Big 4 private client firms
	(1)		(2)		(3)		(4)
	Dep. Var. = GAAP_ETR		Dep. Var. = CASH_ETR		Dep. Var. = BTD		Dep. Var. = CTD
Variables	Coeff.	t-stat	Coeff.	t-stat	Coeff.	t-stat	Coeff.	t-stat
B4	0.005***	2.67	0.010***	2.62	0.000***	3.51	0.001**	2.13
GROUP	0.001	0.91	0.012***	4.35	−0.000**	−2.02	0.001***	3.81
LNASSETS	−0.000	−0.12	0.006***	2.94	−0.000	−0.61	0.001***	5.77
LEV	0.005	1.35	0.025***	3.75	−0.000	−0.06	0.003***	4.98
ROA	0.079***	7.58	0.453***	27.53	0.008***	17.00	0.081***	46.88
LOSS					−0.004***	−24.39	−0.006***	−15.12
CASH	−0.017***	−3.98	−0.020**	−2.35	−0.002***	−5.28	−0.000	−0.49
ATURN	−0.007***	−10.32	0.000	0.32	−0.000***	−11.09	0.000**	2.45
STDSALES	0.006**	2.10	−0.017***	−3.45	0.001***	3.16	−0.002***	−4.06
INTAN	−0.098***	−11.05	−0.103***	−6.17	−0.009***	−16.43	−0.007***	−5.68
TANG	0.008**	2.21	0.036***	5.47	0.001***	3.30	0.005***	8.50
ADA	−0.030***	−3.95	0.009	0.58	−0.001	−1.53	−0.004***	−2.83
LNTAX_FEES	−0.000	−0.83	0.002	1.40	−0.000***	−4.97	−0.001*	−1.93
DISP	−0.002	−1.15	0.002	0.84	0.000	1.54	0.000	1.54
Constant	−0.241***	−13.58	−0.390***	−11.89	0.001	1.24	−0.022***	−7.90
Year FE	Yes		Yes		Yes		Yes
Industry FE	Yes		Yes		Yes		Yes
# of obs.	19,442		14,384		21,789		15,800
R-squared	0.097		0.111		0.124		0.242
B4	0.010***	3.86	0.007	1.02	0.001***	5.37	0.000	0.36
GROUP	0.003**	1.97	0.012***	3.98	−0.000	−0.10	0.001***	3.29
B4×GROUP	−0.010***	−2.80	0.004	0.56	−0.001***	−4.11	0.001	0.99
LNASSETS	−0.000	−0.32	0.006***	2.95	−0.000	−0.72	0.001***	5.76
LEV	0.004	1.23	0.025***	3.75	−0.000	−0.18	0.003***	4.99
ROA	0.079***	7.56	0.453***	27.52	0.008***	16.98	0.081***	46.87
LOSS					−0.004***	−24.47	−0.006***	−15.11
CASH	−0.017***	−3.93	−0.020**	−2.36	−0.002***	−5.19	−0.000	−0.50
ATURN	−0.007***	−10.23	0.000	0.32	−0.000***	−10.90	0.000**	2.44
STDSALES	0.005**	2.05	−0.017***	−3.46	0.001***	3.05	−0.002***	−4.06
INTAN	−0.097***	−10.90	−0.103***	−6.17	−0.009***	−16.18	−0.008***	−5.70
TANG	0.008**	2.30	0.036***	5.46	0.001***	3.47	0.005***	8.50
ADA	−0.029***	−3.82	0.009	0.56	−0.001	−1.34	−0.004***	−2.86
LNTAX_FEES	−0.000	−0.46	0.002	1.36	−0.000***	−4.67	−0.001*	−1.92
DISP	−0.001	−0.92	0.002	0.81	0.000*	1.89	0.000	1.49
Constant	−0.239***	−13.46	−0.390***	−11.89	0.001	1.18	−0.022***	−7.86
Year FE	Yes		Yes		Yes		Yes
Industry FE	Yes		Yes		Yes		Yes
# of obs.	19,442		14,384		21,789		15,800
R-squared	0.098		0.111		0.125		0.242

Panel A (B) presents the OLS regression results of estimating the relation between auditor size and tax avoidance for private firms (auditor size and tax avoidance for group versus standalone private firms). We measure tax avoidance using the GAAP effective tax rate multiplied by −1 (GAAP_ETR, column 1), the cash effective tax rate multiplied by −1 (CASH_ETR, column 2), the level of book-tax differences (BTD, column 3), and the level of book-cash tax differences (CTD, column 4). Higher values of these measures indicate greater tax avoidance. GROUP equals 1 if a client firm is the parent company of a business group, and 0 if it is a standalone firm. T-statistics are calculated based on heteroscedasticity and autocorrelation robust standard errors clustered at the client-firm level. *, **, and *** indicate significance at the two-tailed 10%, 5%, and 1% levels, respectively. See the Appendix for variable definitions.

Table 6. Implications of auditor-provided tax services for audit quality.

Panel A: Auditor-provided tax services and audit quality – Limiting the sample to Big 4 clients with LNTAX_FEES > 0
	(1)		(2)		(3)		(4)
	Dep. Var. = AQ		Dep. Var. = ADA		Dep. Var. = DA+		Dep. Var. = DA-
Variables	Coeff.	t-stat	Coeff.	t-stat	Coeff.	t-stat	Coeff.	t-stat
LNTAX_FEES	−0.002**	−2.27	−0.005*	−1.91	0.002	0.86	−0.001*	−1.92
Controls	Yes		Yes		Yes		Yes
Year FE	Yes		Yes		Yes		Yes
Industry FE	Yes		Yes		Yes		Yes
# of obs.	2,556		2,556		1,153		1,403
R-squared	0.276		0.150		0.650		0.650
Panel B: Auditor-provided tax services and audit quality – Limiting the sample to non-Big 4 clients with LNTAX_FEES > 0
	(1)		(2)		(3)		(4)
	Dep. Var. = AQ		Dep. Var. = ADA		Dep. Var. = DA+		Dep. Var. = DA-
Variables	Coeff.	t-stat	Coeff.	t-stat	Coeff.	t-stat	Coeff.	t-stat
LNTAX_FEES	−0.002***	−2.86	−0.001	−0.45	−0.005**	−2.21	−0.002**	−2.57
Controls	Yes		Yes		Yes		Yes
Year FE	Yes		Yes		Yes		Yes
Industry FE	Yes		Yes		Yes		Yes
# of obs.	4,274		4,274		1,990		2,284
R-squared	0.262		0.117		0.655		0.686

Panel A (B) presents the OLS regression results of estimating the relation between auditor-provided tax service fees and audit quality for Big 4 (non-Big 4) client firms that buy tax services from their auditor. AQ is the Dechow and Dichev (2002) accrual quality measure, modified by McNichols (2002), multiplied by −1; ADA is the absolute value of discretionary accruals as in Kothari et al. (2005), multiplied by −1; DA+ (DA-) represents income-increasing (income-decreasing) discretionary accruals. Higher values of AQ and ADA indicate higher audit quality. T-statistics are calculated based on heteroscedasticity and autocorrelation robust standard errors clustered at the client-firm level. *, **, and *** indicate significance at the two-tailed 10%, 5%, and 1% levels, respectively. See the Appendix in the manuscript for variable definitions.

Table 7. The relation between auditor size and audit quality and tax avoidance for private and public client firms – Using a propensity-score matched sample.

Panel A: Auditor size and audit quality – Matching private and public firms on B4 and all control variables
	(1)		(2)		(3)		(4)
	Dep. Var. = AQ		Dep. Var. = ADA		Dep. Var. = DA+		Dep. Var. = DA-
Variables	Coeff.	t-stat	Coeff.	t-stat	Coeff.	t-stat	Coeff.	t-stat
B4	−0.004***	−3.43	−0.010**	−2.09	−0.005*	−1.91	−0.008***	−2.80
LISTED	−0.001	−0.78	0.000	0.03	−0.000	−0.17	0.001	0.17
B4×LISTED	0.006***	3.44	0.016**	2.42	0.005	1.55	0.012***	2.97
Controls	Yes		Yes		Yes		Yes
Year FE	Yes		Yes		Yes		Yes
Industry FE	Yes		Yes		Yes		Yes
# of obs.	2,256		2,256		1,089		1,167
R-squared	0.347		0.099		0.461		0.434
Panel B: Auditor size and tax avoidance – Matching private and public firms on B4 and all control variables
	(1)		(2)		(3)		(4)
	Dep. Var. = GAAP_ETR		Dep. Var. = CASH_ETR		Dep. Var. = BTD		Dep. Var. = CTD
Variables	Coeff.	t-stat	Coeff.	t-stat	Coeff.	t-stat	Coeff.	t-stat
B4	0.017**	2.05	0.042***	3.63	0.001*	1.69	0.002**	2.00
LISTED	0.056***	6.66	0.088***	7.72	0.004***	6.73	0.004***	4.69
B4×LISTED	−0.028**	−2.38	−0.053***	−3.21	−0.002***	−2.61	−0.003**	−2.47
Controls	Yes		Yes		Yes		Yes
Year FE	Yes		Yes		Yes		Yes
Industry FE	Yes		Yes		Yes		Yes
# of obs.	1,346		1,094		1,704		1,364
R-squared	0.159		0.211		0.305		0.523

Panel A (B) presents the OLS regressions of estimating the relation between auditor size and audit quality (tax avoidance) for a sample of propensity-score matched private and public firms. In Panel A (B), we propensity-score match private and public firms on B4 and all control variables in equation (3) (equation (5)). AQ is the Dechow and Dichev (2002) accrual quality measure, modified by McNichols (2002), multiplied by −1; ADA is the absolute value of discretionary accruals as in Kothari et al. (2005), multiplied by −1; DA+ (DA-) represents income-increasing (income-decreasing) discretionary accruals. GAAP_ETR is the GAAP effective tax rate, multiplied by −1; CASH_ETR is the cash effective tax rate, multiplied by −1; BTD is the level of book-tax differences; CTD is the level of book-cash tax differences. Higher values of AQ and ADA (GAAP_ETR, CASH_ETR, BTD, CTD) indicate higher audit quality (greater tax avoidance). T-statistics are calculated based on heteroscedasticity and autocorrelation robust standard errors clustered at the client-firm level. *, **, and *** indicate significance at the two-tailed 10%, 5%, and 1% levels, respectively. See the Appendix for variable definitions.

Table

ADA	The absolute value of performance-adjusted discretionary accruals, as in Kothari et al. (2005), multiplied by −1;
AQ	The Dechow and Dichev (2002) accruals quality measure, as modified by McNichols (2002), multiplied by −1;
ATURN	The ratio of sales to lagged total assets;
B4	An indicator variable that equals 1 if a firm is audited by a Big 4 auditor in a specific fiscal year, and 0 otherwise;
BTD	The level of book-tax differences calculated as pre-tax income times the statutory tax rate less tax expense. We scale this variable by the level of lagged total assets. Higher values of BTD indicate more book-tax differences;
CASH	The ratio of cash to lagged total assets;
CASH_ETR	The ratio of cash taxes paid to pre-tax income. We multiply CASH_ETR by −1, so that higher values of this measure indicate more tax avoidance;
CTD	The level of book-cash tax differences calculated as pre-tax income times the statutory tax rate less cash taxes paid. We scale this variable by the level of lagged total assets. Higher values of CTD indicate more book-cash tax differences;
CURR	The ratio of current assets to current liabilities;
DA+	Income-increasing performance-adjusted discretionary accruals;
DA-	Income-decreasing performance-adjusted discretionary accruals;
DISP	An indicator variable that equals 1 if no shareholder owns more than 50% of the firm (BvD Independence Indicator = ‘A’ or ‘B’), and 0 otherwise (BvD Independence Indicator = ‘C’ or ‘D’);
GAAP_ETR	The ratio of tax expense to pre-tax income. We multiply GAAP_ETR by −1, so that higher values of this measure indicate more tax avoidance;
GROUP	An indicator variable that equals 1 if a client firm is the parent company of a business group, and 0 if it is a standalone firm;
GROWTH	The percentage change in sales in year t relative to year t-1;
IMPORTANCE	The ratio of sales of a client firm in a given year to the total amount of sales of all client firms of the audit firm in that year;
INTAN	The ratio of intangible assets to lagged total assets;
LEV	The ratio of long-term debt to lagged total assets;
LISTED	An indicator variable that equals 1 if a firm is publicly listed, and 0 otherwise;
LNASSETS	The natural logarithm of total assets;
LNTAX_FEES	The natural logarithm of tax fees;
LOSS	An indicator variable that equals 1 if the firm reports negative pre-tax income in a specific year, and 0 otherwise;
REC & INV	The ratio of receivables and inventory to lagged total assets;
ROA	The ratio of pre-tax income to lagged total assets;
STDSALES	The five-year standard deviation of sales;
TA	The ratio total accruals to lagged total assets;
TANG	The ratio of PPE to lagged total assets;
TENURE	An indicator variable that equals 1 if auditor tenure is three years or less, and 0 otherwise (Francis & Yu, 2009);
Z-SCORE	Altman’s Z-score calculated as 1.2 × (net working capital/total assets) + 1.4 × (retained earnings/total assets) + 3.3 × (EBIT/total assets) + 0.6 × (book value of equity/book value of total liabilities) + sales/total assets;
ΔVARIABLE	The difference between the value of a given variable in year t relative to its value in year t-1.

Dechow, P. M., & Dichev, I. D. (2002). The quality of accruals and earnings: The role of accrual estimation errors. The Accounting Review, 77(2002), 35–59. https://doi.org/10.2308/accr.2002.77.s-1.35

 Web of Science ®Google Scholar

McNichols, M. F. (2002). The quality of accruals and earnings: The role of accrual estimation errors: Discussion. The Accounting Review, 77(2002), 61–69. https://doi.org/10.2308/accr.2002.77.s-1.61

 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

Francis, J. R., & Yu, M. D. (2009). Big 4 office size and audit quality. The Accounting Review, 84(5), 1521–1552. https://doi.org/10.2308/accr.2009.84.5.1521

 Web of Science ®Google Scholar