Determinants of financial inclusion gender gap in Ethiopia: Evidence from decomposition analysis

Figures & data

Figure 1. Conceptual framework.

Source: Own Construction based on literature.

Table 1. Variables definition and expected sign

Variable Name	Description	Expected Sign
Dependent Variables
Account	The respondent owns (or not), alone or with someone, an account in a formal financial institution. It takes 1 if the individual owns an account and 0 otherwise.	Not applicable
Borrowing	The respondent has (or has not) saved in a formal account in the past 12 months. It takes 1 if the individual has saved in the past 12 months and 0 otherwise.	Not applicable
Saving	The respondent has (or has not) borrowed from a formal financial institution. It takes 1 if the individual has borrowed in the past 12 months and 0 otherwise.	Not applicable
Emergency	Dummy that takes 1 if the respondent came up with an emergency fund and 0 otherwise.	Not applicable
Debit Card Holding	Dummy takes 1 if the respondent accessed the account using a mobile phone or internet, and 0 otherwise.	Not applicable
Explanatory variables
Education	Education level of the respondent: Primary education or less; and Secondary education. It takes 1 if the respondent completed Secondary education and 0 otherwise.	+
Age	Age of the respondent in years.	+
Age square	Age in years of the respondent squared.	-
Income quintile	Income quintile of the respondent: Income quintile 1 for poorest (20%), Income quintile 2 for second poorest (20%), Income quintile 3 for third poorest (20%), Income quintile 4 for fourth poorest (20%), and Income quintile 5 for fifth poorest (20%).	+
Employment	Dummy takes 1 if the respondent is in the workforce and 0 otherwise.	+
Mobile	Dummy takes 1 if the respondent has a mobile and 0 otherwise.	+
NID	Dummy takes 1 if the respondent has a national ID and 0 otherwise.	+

Source: Authors’ Compilation.

Table 2. Two-sample t-test with unequal variances

Variable	Mean (male)	Mean (female)	Mean difference	St. Err.	t-value	P-value
Age	34.623	31.892	2.731	.918	2.95	.003
Age squared	1419.045	1190.215	228.831	79.239	2.90	.004

Table 3. Fairlie decomposition result

Variables	Formal Saving		Formal Account		Formal Borrowing		Debit Card Holding		Emergency
	Decomposition
	Coeff.	Std. err.	Coeff.	Std. err.	Coeff.	Std. err.	Coeff.	Std. err.	Coeff.	Std. err.
Age	−0.030*	0.018	−0.041***	0.014	−0.016	−0.600	−0.031	0.032	−0.012	0.017
Age square	0.021	0.017	0.023*	0.012	0.020	0.820	0.029	0.032	0.020	0.016
Income quintile 1	0.003	0.002	0.006	0.003	−0.002	−0.740			0.011***	0.003
Income quintile 2	−0.001	0.001	−0.002	0.003	0.000	0.280			0.002	0.001
Income quintile 3	−0.002	0.002	−0.004	0.003	0.001	0.580			−0.002*	0.001
Income quintile 4	−0.001	0.002	−0.003	0.002	0.000	0.060	−0.005*	0.003	0.001	0.002
Education	−0.005	0.006	−0.024***	0.006	0.009	1.240	−0.018***	0.007	−0.018***	0.007
Employment	−0.037***	0.009	−0.004	0.007	−0.031***	−3.880	−0.007*	0.004	0.002	0.008
Mobile	−0.031***	0.010	−0.033***	0.010	−0.006	−0.550	−0.007	0.007
NID			−0.041***	0.009	0.011	1.110	−0.005	0.004
Group (male)	.716		.530		.455		0.116		0.673
Group (female)	.646		.367		.357		0.058		0.589
Difference	.070**		.163		.098***		0.058***		0.084***
Explained gap	.0998		.140		.016		0.050		0.041
Observations	1000		1000		1000		1000		1000

*** p<0.01, ** p<0.05, * p<0.1.

Source: Authors’ Computation using Stata 17.

Table 4. Daymont and Andrisani decomposition results

	Saving			Account			Borrowing			Emergency			Debit Card
	Coef.	%	Std. err.	Coef.	%	Std. err.	Coef.	%	Std. err.	Coef.	%	Std. err.	Coef.	%
$\mathit{O}\mathit{\text{mega}}=1$
Char	.082***	117.1	0.017	.121***	74.4	0.022	.019	14.3	0.210	.019	23.2	0.021	.031***	54.2
Coef	−.030	−42.8	0.037	.023	14.3	0.031	.089**	83.7	0.039	.027	32.6	0.038	.006	10.4
Int	.018	25.7	0.027	.018	11.3	0.026	−.011	1.9	0.028	.037	44.2	0.03	.020	35.3
$\mathit{O}\mathit{\text{mega}}=0$
Char	.100***	142.8	0.025	.140***	85.7	0.025	.008	16.3	0.025	.056**	67.4	0.027	.051	89.6
Coef	−.012	−42.1	0.032	.041	25.6	0.032	.078**	85.7	0.034	.064**	76.8	0.031	.026	45.8
Int	−.018	−25.7	0.027	−.018	−11.3	0.026	.011	−1.9	0.028	−.037	−44.2	0.030	−.020***	35.3
$\mathit{O}\mathit{\text{mega}}=\mathit{\text{wgt}}$
Prod	.085***	121.4	0.017	.132***	80.9	0.027	.031**	16.9	0.013	.039**	46.3	0.017	.042***	73.1
Adv	−.009	−12.9	0.016	.019	11.5	0.015	.040**	43.3	0.017	.027*	32.3	0.016	.009	16.2
Disadv	−.006	−8.5	0.020	.012	7.6	0.010	.027**	39.8	0.011	0.018*	21.4	0.010	.006	10.7
Raw	.070**		0.031	0.163***		0.032	.098***		0.031	.084***		0.031	.057***

*** p<0.01, ** p<0.05, * p<0.1.

Source: Authors’ Computation using Stata 17.

Table 5. Detailed decomposition results

	Saving			Account			Borrowing
Decomposition Results
	Coef.	%	Std. err.	Coef.	%	Std. err.	Coef.	%	Std. err.
E	.099***	142.8	0.021	.140***	85.7	0.019	.016	16.3	0.024
C	−.030***	−42.8	0.036	.023	14.3	0.031	.082**	83.7	0.040
R	.070**		0.028	.163***		0.027	.098***		0.031
Due to Difference in Characteristics (E)
Age	.070***	99.9	0.019	.090***	55.5	0.023	.047	48.6	0.031
Age square	−.054***	−77.4	0.019	−.078***	−48.1	0.022	−.058*	−59.4	0.033
Income quintile 1	−.007***	−9.7	0.002	−.008***	−4.7	0.002	.002	1.8	0.003
Income quintile 2	.001**	1.5	0.001	.002**	1.0	0.0001	−.002***	−1.5	0.001
Income quintile 3	.002	2.8	0.002	.002*	1.3	0.001	−.000	−.5	0.002
Income quintile 4	.005***	6.9	0.002	.003**	2.2	0.002	−.001	−1.5	0.002
Education	.006	9.1	0.008	.027**	16.7	0.007	−.018**	−18.9	0.010
Employment	.048***	69.3	0.015	.036**	22.4	0.020	.006	6.1	0.019
Mobile	.028**	40.3	0.012	.037***	23.2	0.011	.022	22.1	0.015
NID				.026	16.2	0.014	.019	19.7	0.016
Due to Differences in Coefficients (C)
Age	.718*	434.1	0.477	.4005	245.7	0.390	0.249	407.7	255.97
Age square	−.291	−175.7	0.193	−.147	−90	0.162	−0.115	−177.8	−118.08
Income quartile 1	.001	.8	0.022	.006	3.4	0.014	−0.001	−6.4	−1.0639
Income quartile 2	−.006	−3.4	0.023	−.005	−2.9	0.017	0.029	23.6	29.50
Income quartile 3	.015	8.9	0.024	.022	13.7	0.016	−0.004	−7.5	−3.94
Income quartile 4	−.036**	−22.3	0.035	.002	1.2	0.016	0.010	9.3	10.70
Education	−.006	−3.9	0.025	−.006	−3.4	0.016	−0.013	−3.7	−13.35
Employment	.052	31.2	0.077	.150	92.3	0.084	−0.109	−90.5	−111.74
Due to Differences in Coefficients (C)
Mobile				.007	4.13	0.029	0.025	25.634	25.63
NID				−.060	−36.79	0.036	0.073	74.31	74.31
Constant	−.351	−212.1	0.282	−.353	−216.5	0.280	−0.063	−43.8	−64.24

*** p<0.01, ** p<0.05, * p<0.1.

Source: Authors Computation using Stata 17.

Table A1. Chi-square test of association between gender and discrete explanatory variables

Explanatory Variables
Income level 1 (poorest 20%)
Gender	Income level 1 (dummy)			Pearson ${\chi }^2$
	0	1	Total
Male	83.2%	16.8%	100%	1.5509
Female	86%	14%	100%
Total	84.9%	15.1%	100%
Income level 2 (second 20%)
Gender	Income level 2 (dummy)			Pearson ${\chi }^2$
	0	1	Total
Male	84.2%	15.8%	100%	0.0670*
Female	83.6%	16.4%	100%
Total	83.8%	16.2%	100%
Income level 3 (third 20%)
Gender	Income level 3 (dummy)			Pearson ${\chi }^2$
	0	1	Total
Male	83.9%	16.1%	100%	0.5819
Female	82.1%	17.9%	100%
Total	82.8%	17.2%	100%
Income level 4 (fourth 20%)
Gender	Income level 4 (dummy)			Pearson ${\chi }^2$
	0	1	Total
Male	80.4%	19.6%	100%	0.8959
Female	77.9%	22.1%	100%
Total	78.9%	21.1%	100%
Educational status of respondents
Gender	Education (dummy)			Pearson ${\chi }^2$
	Primary	Secondary	Total
Male	63.8%	36.2%	100%	18.0810***
Female	76.3%	23.7%	100%
Total	71.3%	28.7%	100%
Employment status of respondents
Gender	Employment (dummy)			Pearson ${\chi }^2$
	Out of workforce	In workforce
Male	10.6%	89.4%	100%	54.1174***
Female	30.4%	69.6%	100%
Total	22.5%	77.5	100%
National ID Ownership
Gender	Own National ID			Pearson ${\chi }^2$
	Yes	No	Total
Male	79.1%	20.9%	100%	50.9135***
Female	57.3%	42.7%	100%
Total	66%	44%	100%
Mobile Ownership
Gender	Owns Mobile			Pearson ${\chi }^2$ 47.5588***
	Yes	No	Total
Male	66.6%	33.4%	100%
Female	44.3%	55.7%	100%
Total	53.2%	46.8%	100%

Table A2. Chi-Square test of association between gender and dependent variables

Dependent Variables
Formal Account
Gender	Account (dummy)			Pearson ${\chi }^2$
	0	1	Total
Male	47%	53%	100%	25.9562***
Female	63.3%	36.7%	100%
Total	56.8%	43.2%	100%
Formal Savings
Gender	Saving (dummy)			Pearson ${\chi }^2$
	0	1	Total
Male	57.5%	42.5%	100%	29.9102***
Female	74.1%	25.9%	100%
Total	67.5%	32.5%	100%
Formal Borrowing
Gender	Borrowing (dummy)			Pearson ${\chi }^2$
	0	1	Total
Male	54.5%	45.5%	100%	9.5482***
Female	64.3%	35.7%	100%
Total	60.4%	39.6%	100%
Debit Card Holding
Gender	Debit Card (dummy)			Pearson ${\chi }^2$
	0	1	Total
Male	88.4%	11.6%	100%	10.6190***
Female	94.2%	5.8%	100%
Total	91.9%	8.1%	100%
Possibility of coming up with emergency fund
Gender	Emergency fund (dummy)			Pearson ${\chi }^2$
	0	`1	Total
Male	32.7%	67.3%	100%	7.1408 ***
Female	41%	59%	100%
Total	71.3%	28.7%	100%