Gender and the tax compliance puzzle: does gender influence taxpayers’ behaviour towards tax compliance? Evidence from Rwanda

Figures & data

Table 1. Sampling frame.

Category	Population	Female owned/managed	Male owned/Managed	Sample
Large	10,056	2011	8,045	33
Medium	20,567	8,227	12,340	68
Small	31,751	15,240	16,511	105
Micro	58,586	35,151	23,435	195
Total	120,960	60,629	60,331	401

Source: RRA Statistical Reports 2020/National Institute of Statistics Rwanda (2020).

Table 2. Measurement of study of variables.

Variable	Type	Data
Tax Audit (TA)	Categorical using Likert Scale	Survey data
Tax Penalty (TP)	Categorical using Likert Scale	Survey data
Tax Fairness (TF)	Categorical using Likert Scale	Survey data
Enforcement level (EL)	Categorical using Likert Scale	Survey data
Corruption (C)	Categorical using Likert Scale	Survey data
Government Spending (GS)	Categorical using Likert Scale	Survey data
Economic Factors (EF)	Categorical using Likert Scale	Survey data
Tax Education (TE)	Categorical using Likert Scale	Survey data
Tax Knowledge	Categorical using Likert Scale	Survey data
Roles of tax authority (RTA)	Categorical using Likert Scale	Survey data
Tax Rate (TR)	Categorical using Likert Scale with 1 male and 2 female	Survey data
Gender	Dummy data	Survey data
Tax compliance	Categorical using Likert Scale	Survey data

Tax compliance (TC) = F (Taxpayer’s behaviour TB).

Figure 1. Theoretical framework of analysis modified from Serkan et al. (2011).

Figure 2. Group statistics on the determinants of taxpayers’ behaviour towards tax compliance between male and female taxpayers.

Table 3. Regression models.

	Model 1 (Male & Female included)	Model 2 (if gender = Male)	Model 3 ((if gender ‘=Female)
Tax knowledge	0.0420826	0.0151814	0.130855***
	(1.74)	(0.72)	(3.62)
Enforcement level	−1.534415***	−0.2363028***	0.1353184***
	(-6.09)	(-8.01)	(4.54)
Perception of government spending	−0.0235381	0.1176839***	−0.0929172**
	(1.00)	(5.12)	(2.56)
Role of tax authority	0.0386073***	0.0316607	0.941621***
	(-2.11)	(1.29)	(-3.52)
Tax fairness	0.1523574***	0.260066***	0.01237835
	(6.75)	(-10.79)	(0.32)
Tax rates	−0.0887628***	−0.979553***	−0.0820734
	(-4.37)	(5.11)	(-2.70)
Economic factors	0.015617	−0.0626991**	−0.001488
	(0.92)	(-2.51)	(0.06)
Corruption	−1.504722***	−0.1720719***	−0.1365769***
	(-5.22)	(-6.84)	(-3.40)
Tax education	0.1209815***	0.12244073***	0.02062315***
	(3.60)	(3.83)	(5.58)
Tax audit	−0.2245682***	−0.844073***	−0.188144***
	(-9.71)	(4.24)	(-3.54)
Tax penalties	−0.24273	0.1564189***	0.0766483
	(-0.47)	(3.28)	(1.01)
Gender	0.1712104***	–	–
	(3.88)	–	–
Cons	4.549661***	3.090674***	5.90103***
	(14.87)	(8.77)	(7.26)
R^2	0.5100	0.8423	0.7186
Ad. R^2	0.4930	0.8327	0.6985

t statistics in parentheses, ∗p < .10, ∗∗p < .05, ∗∗∗p < .0 > 1.

Figure 3. Independent samples test for significance of the mean difference.

Figure 4. Multivariate tests^a determinants of taxpayers’ compliance and gender.

Figure 5. Levene’s test of equality of error variances^a.

Figure 6. Test between subject effects.

Table

	Very low extent	Low extent	Average extent	High extent	Very high extent
Tax knowledge
Enforcement levels
Perception of government spending
Role of tax authority
Tax fairness
Tax rates
Demographic factors
Economic factors
Cultural factors
Corruption
Tax education
Tax audit
Tax penalties

Table

	SD	D	N	A	SA
I always declare all my taxes on time
I always pay all my taxes on time
I Keep proper books of accounts
I register all my taxes on time

Table A1. Group statistics on the determinants of taxpayers’ beheviour towards tax compliance between male and female taxpayers.

	Gender	N	Mean	Std. Deviation	Std. Error Mean
Tax knowledge	Male	209	4.3158	1.06779	.07386
	Female	181	4.1602	.81088	.06027
Enforcement levels	Male	239	3.8086	.90521	.09028
	Female	151	2.9558	1.12952	.08396
Perception of government spending	Male	209	3.7416	1.38683	.09593
	Female	181	4.1050	1.09799	.08161
Role of tax authority	Male	209	3.5694	.99335	.06871
	Female	181	3.9834	.95728	.07115
Tax fairness	Male	209	4.0096	1.34447	.09300
	Female	181	3.8895	1.23331	.09167
Tax rates	Male	200	3.6416	1.48397	.10265
	Female	190	3.6243	1.37124	.10192
Economic factors	Male	229	4.2052	1.13506	.07851
	Female	161	3.6441	1.20774	.08977
Corruption	Male	209	4.0670	.98318	.06801
	Female	181	4.5912	.49298	.03664
Tax education	Male	209	4.5024	.68728	.04754
	Female	181	4.3646	.74511	.05538
Tax audit	Male	209	4.3349	.84511	.05846
	Female	181	4.3702	1.14552	.08515
Tax penalties	Male	209	4.8066	.39604	.02944
	Female	181	4.1944	.65905	.04589

Source: Survey data.

Table A2. Independent test.

		Levene’s Test for Equality of Variances		t-test for Equality of Means
		F	Sig	t	df	Sig. (2-tailed)	Mean Difference	Std. Error Difference	95% Confidence Interval of the Difference
									Lower	Upper
Tax knowledge	Equal variances assumed	2.493	.115	1.601	388	.110	.1556	.09719	−.03552	.34666
	Equal variances not assumed			1.632	381.7	.104	.1556	.09533	−.03187	.34301
Enforcement levels	Equal variances assumed	7.293	.007	−1.182	388	.238	−.1472	.12457	−.39210	.09772
	Equal variances not assumed			−1.194	388.0	.233	−.1472	.12329	−.38958	.09521
Perception on government spending	Equal variances assumed	4.841	.028	−2.838	388	.005	−.3634	.12805	−.61510	−.11160
	Equal variances not assumed			−2.885	384.9	.004	−.3634	.12595	−.61098	−.11571
Role of tax authority	Equal variances assumed	4.131	.043	−4.175	388	.000	−.4141	.09918	−.60904	−.21905
	Equal variances not assumed			−4.186	383.6	.000	−.4141	.09892	−.60853	−.21956
Tax fairness	Equal variances assumed	1.486	.224	.914	388	.361	.1201	.13140	−.13827	.37841
	Equal variances not assumed			.919	386.7	.358	.1201	.13058	−.13668	.37681
Tax rates	Equal variances assumed	5.356	.021	.806	388	.420	.1173	.14548	−.16871	.40334
	Equal variances not assumed			.811	386.4	.418	.1173	.14465	−.16709	.40173
Economic factors	Equal variances assumed	36.33	.000	.338	388	.735	.0402	.11873	−.19328	.27359
	Equal variances not assumed			.337	372.2	.737	.0402	.11926	−.19436	.27466
Corruption	Equal variances assumed	46.80	.000	−6.499	388	.000	−.5242	.08065	−.68274	−.36560
	Equal variances not assumed			−6.785	315.6	.000	−.5242	.07725	−.67617	−.37218
Tax education	Equal variances assumed	3.102	.079	1.898	388	.050	.1378	.07257	−.00492	.28042
	Equal variances not assumed			1.887	369.4	.060	.1378	.07299	−.00577	.28128
Tax audit	Equal variances assumed	15.73	.000	−.349	388	.728	−.0352	.10111	−.23403	.16356
	Equal variances not assumed			−.341	326.8	.733	−.0352	.10328	−.23842	.16794
Tax penalties	Equal variances assumed	44.8	.000	−10.96	388	.000	.6152	.05613	−.72560	−.50488
	Equal variances not assumed			−11.34	347.8	.000	−.6152	.05427	−.72197	−.50851

Table A3. Multivariate tests^a determinants of taxpayers’ compliance and gender.

Effect		Value	F	Hypothesis df	Error df	Sig	Partial Eta Squared	Noncent. Parameter	Observed Power^c
Intercept	Pillai’s Trace	.840	179.481^b	11.000	377.000	.000	.840	1974.288	1.000
	Wilks’ Lambda	.160	179.481^b	11.000	377.000	.000	.840	1974.288	1.000
	Hotelling’s Trace	5.237	179.481^b	11.000	377.000	.000	.840	1974.288	1.000
	Roy’s Largest Root	5.237	179.481^b	11.000	377.000	.000	.840	1974.288	1.000
Tax compliance	Pillai’s Trace	.503	34.744^b	11.000	377.000	.000	.603	382.183	1.000
	Wilks’ Lambda	.497	34.744^b	11.000	377.000	.000	.603	382.183	1.000
	Hotelling’s Trace	1.014	34.744^b	11.000	377.000	.000	.603	382.183	1.000
	Roy’s Largest Root	1.014	34.744^b	11.000	377.000	.000	.603	382.183	1.000
Gender	Pillai’s Trace	.551	42.113^b	11.000	377.000	.000	.551	463.242	1.000
	Wilks’ Lambda	.449	42.113^b	11.000	377.000	.000	.551	463.242	1.000
	Hotelling’s Trace	1.229	42.113^b	11.000	377.000	.000	.551	463.242	1.000
	Roy’s Largest Root	1.229	42.113^b	11.000	377.000	.000	.551	463.242	1.000

a. Design: Intercept + Tax compliance + Gender.

b. Exact statistic.

c. Computed using alpha = .05.

Table A4. Levene’s test of equality of error variances^a.

	F	df1	df2	Sig
Tax knowledge	1.621	1	388	.204
Enforcement levels	16.977	1	388	.000
Perception on government spending	13.201	1	388	.000
Role of tax authority	.555	1	388	.457
Tax fairness	19.560	1	388	.000
Tax rates	17.389	1	388	.000
Economic factors	14.404	1	388	.000
Corruption	42.559	1	388	.000
Tax education	49.783	1	388	.000
Tax audit	14.427	1	388	.000
Tax penalties	16.090	1	388	.000

Tests the null hypothesis that the error variance of the dependent variable is equal across groups.

Design: Intercept + Tax compliance + Gender.

Table A5. Tests of between-subject effects.

Source		Type III Sum of Squares	df	Mean Square	F	Sig	Partial Eta Squared	Noncent. Parameter	Observed Power^l
Corrected Model	Tax knowledge	4.865^a	2	2.433	2.667	.071	.14	5.334	.528
	Enforcement levels	13.312^b	2	6.656	4.497	.012	.23	8.994	.767
	Perception of government spending	33.087^c	2	16.543	10.728	.000	.53	21.457	.990
	Role of the tax authority	25.675^d	2	12.837	13.756	.000	.66	27.512	.998
	Tax fairness	111.625^e	2	55.813	40.033	.000	.171	80.065	1.000
	Tax rates	145.521^f	2	72.761	43.167	.000	.182	86.334	1.000
	Economic factors	44.132^g	2	22.066	17.551	.000	.83	35.102	1.000
	Corruption	37.698^h	2	18.849	31.206	.000	.139	62.411	1.000
	Tax education	16.835^i	2	8.417	17.782	.000	.84	35.564	1.000
	Tax audit	5.384^j	2	2.692	2.745	.065	.14	5.491	.541
	Tax penalties	44.151^k	2	22.076	76.869	.000	.284	153.738	1.000

^aR Squared = .132 (Adjusted R Squared = .114).

^bR Squared = .430 (Adjusted R Squared = .418).

^cR Squared = .363 (Adjusted R Squared = .349).

^dR Squared = .256 (Adjusted R Squared = .240).

^eR Squared = .439 (Adjusted R Squared = .427).

^fR Squared = .330 (Adjusted R Squared = .316).

^gR Squared = .275 (Adjusted R Squared = .259).

^hR Squared = .292 (Adjusted R Squared = .277).

ⁱR Squared = .360 (Adjusted R Squared = .346).

^jR Squared = .224 (Adjusted R Squared = .208).

^kR Squared = .507 (Adjusted R Squared = .497).

^lComputed using alpha = .05.

Serkan, B., Ahmet, F. C., & Tamer, B. (2011). An Investigation of tax compliance intention: A theory of planned behavior approach. European Journal of Economics, Finance and Administrative Science, 28(28), 180–188.

 Google Scholar

Availability of data

Raw data were generated from different business enterprises in Rwanda. Derived data supporting the findings of this study are available from the corresponding author (Dr Daniel Twesige) on request