The Quest for Female Economic Empowerment in Sub-Saharan African Countries: Implications on Gender-Based Violence

Abstract

Recent empirical literature has focused on the social aspect of gender-based violence regarding domestic violence and physical abuse while the implications of economic empowerment in an attempt to reduce gender-based violence remain under-researched. This study investigated the connection between female economic empowerment and factors that could reduce gender-based violence in sub-Saharan African countries. We used the panel fully modified least squares estimation method to investigate the long-run implications. The gender inequality index, the female genital mutilation prevalence, and the number of female children out of school were used as proxies for gender-based violence. Likewise, economic empowerment was a proxy for female economic participation; it was replaced by female employment for the robustness test. Evidence from the panel fully modified least squares estimation showed that female economic empowerment had a negative relationship with the gender inequality index, the number of female children out of primary school, and female genital mutilation. We concluded that an increase in the economic power of females through increased economic participation could reduce gender-based violence in the long run. Based on these findings, this study recommends policies to improve the situation. This study shifts attention to the macro-connection between factors that can reduce GBV and increase female economic empowerment in selected areas of sub-Saharan Africa.

1. Introduction

Gender-based violence (GBV) has long-term consequences on victims, their families, and broader societies. Furthermore, empowering women plays a pivotal role in reducing poverty. Consequently, there is a pressing need for global collaboration to safeguard women and girls, ensuring their full participation in society (Asaleye and Strydom 2022; Mbukanma and Strydom 2021; World Bank 2019). The official statistics show that GBV is a global epidemic that affects one out of every three women in her lifetime (World Bank 2019). Moreover, global statistics show that 35 per cent of women have experienced physical and/or sexually intimate relationship violence. Additionally, an intimate partner is responsible for up to 38 per cent of female deaths worldwide.

Similarly, the World Bank (2019) states that 200 million women have been subjected to female genital mutilation/cutting. This problem is traumatic for victims of domestic violence and their relatives, and has substantial societal and economic consequences. The World Bank’s (2019) report also stated that violence against women is projected to cost certain nations up to 3.7 per cent of their GDP, approximately half of what most governments spend on education. In addition, according to a World Health Organization’s (2021) report, approximately 37 per cent of women living in sub-Saharan African countries have experienced GBV, with a prevalence as high as one in two in certain countries.

Consequently, Goal 5 of the Sustainable Development Goals emphasises achieving gender equality and empowering all women and girls. This study focuses on female economic empowerment as a tool to reduce the risk of GBV, and this empowerment will also assist in promoting gender equality. According to the United Nations (2020), the COVID-19 epidemic has the potential to undo the minimal gains in gender equality and women’s rights. The coronavirus pandemic has exacerbated inequities for women and girls in many areas, from health and the economy to security and social protection. Given this, the United Nations designed a swift and targeted response focused on the effect of the COVID-19 crisis on women and girls to ensure that long-term recovery benefits them. The goals include the following: the ability to minimise and decrease GBV, especially domestic violence; to ensure that women and girls benefit from social protection and economic stimulus measures; to encourage and promote proportionate sharing of care work in the community; to increase the inclusiveness of women in decision-making and planning processes regarding the response to COVID-19; and to include gender views in data and coordinating processes. A vast number of recent studies have focussed on the ability to minimise and decrease GBV, especially domestic violence, but the other goals identified by the United Nations (2020) have been neglected (Bhana et al. 2021; Eger 2021; Nunbogu and Elliott 2022; Montserrat et al. 2022; del Mar Rodas-Zuleta et al. 2022; Freijomil-Vázquez et al. 2022; Mbukanma and Strydom 2022). Therefore, this study was designed to fill the gap in the literature by modelling the long-run implications of female economic empowerment on GBV in sub-Saharan African countries to reduce the prevalence of GBV.

2. Literature Review

Theoretically, the connection between GBV and economic participation has been explained in the resource model; this model predicts that all social connections are based on the assumption of force or the threat of force (Kenny et al. 2019). According to this theory, the greater the economic empowerment, the greater the danger of force but the less likely force/violence would be employed in that relationship (Walters et al. 2013). The theory does not imply that no threats would be employed to retain control. Still, it does imply that physical violence may not manifest in the process, but emotional violence is often displayed. The aggression that a woman may face in this circumstance is emotional rather than physical, causing GBV. If a person has fewer alternative resources, force/violence may be used to control the relationship (Tsai 2022; Walters et al. 2013). Although legislative regulations are primarily concerned with physical violence against women and do not adequately address emotional violence, resource theory should be applied to enhance social change to eradicate GBV caused by economic power (Jewel 2020). Emotional violence, primarily stemming from diminished economic power, is not yet universally acknowledged as a crime or unlawful act. Therefore, societal change in economic empowerment should focus on addressing emotional abuse against women. Because of this, many cultural norms are being challenged (Jewel 2020; Walters et al. 2013); this remains an identified gap in the empirical literature that necessitates attention.

According to resource theory, the distribution of resources plays a crucial role in determining power dynamics within engagements or relationships (Kenny et al. 2019; Walters et al. 2013). Concerning GBV, economic resources are particularly significant. When women have greater economic autonomy, they gain independence and negotiating power in their relationships (Asaleye and Strydom 2023a; Tsai 2022). In the context of this study, women’s increasing access to employment and economic opportunities is considered to be a valuable resource, aligning with the principles of resource theory. Economic empowerment has the potential to reshape traditional power dynamics within relationships, thereby reducing the likelihood of violence (Bashford-Squires et al. 2022; Tsai 2022). By altering these traditional power imbalances, economic empowerment may contribute to a decrease in the incidence of violence (Asaleye and Strydom 2023b).

Additionally, social exchange theory underscores the continuous transfer of assets in relationships, with individuals attempting to maximise advantages while minimising costs (Kreager et al. 2013). In the realm of economic independence for women, employment and financial autonomy are essential resources. According to social exchange theory, women’s financial contributions in marriages or relationships can contribute to a decrease in the probability of violence as a means of dispute resolution (Kreager et al. 2013). The theory suggests that reciprocal gains are crucial for relationship stability. Therefore, economic empowerment for women may lead to healthier and more equitable marriages. Jewel (2020) advocates the exploring of social exchange theory to address the societal reforms necessary to eliminate violence against women. However, an effective reduction in violence against women requires sufficiently severe consequences to deter offences. Unfortunately, this is often lacking in cases of emotional abuse, which may be undocumented by victims. Both resource theory and social exchange theory highlight the significance of economic resources in relationships (Kenny et al. 2019; Kreager et al. 2013; Walters et al. 2013). Resource theory emphasises the broader cultural environment, demonstrating that access to economic resources empowers women at a structural level. Conversely, social exchange theory focuses on the micro-dynamics of individual relationships, stressing the continuous exchange of resources and its impact on the stability of relationships.

Empirically, recent literature has focused on the social aspect of GBV in terms of domestic violence and physical abuse, among others, while the implications of economic empowerment remain under-researched (Eger 2021; Montserrat et al. 2022; Freijomil-Vázquez et al. 2022); other studies on GBV have been carried out at the micro-level (del Mar Rodas-Zuleta et al. 2022; Bhana et al. 2021; Oparinde and Matsha 2021; Okolo and Okolo 2018). Another strand of literature has considered the impact of GBV on public health issues (Nunbogu and Elliott 2022; Mutinta 2022). In addition, further studies relating to female economic empowerment tend towards economic reforms, political outcomes, and economic growth. For example, Dahlum et al. (2022) documented that women’s political empowerment positively affects the economic growth of a country. A similar study was conducted by Deininger et al. (2020) in India with almost the same outcome. Buvinic et al. (2022) reported that women’s agency directly contributes to increased business profits, whereas personal savings, business profits, and household income are indirect drivers of women’s economic empowerment. Both factors have a strong direct impact on household income, which can increase women’s economic empowerment; this is reinforced by the direct effects of increased business profits. Ambler et al. (2021) investigated the factors facilitating women’s access to an economic empowerment initiative in Uganda and concluded that couples invited to a workshop were less likely to decline the empowerment initiative that was later given. Furthermore, the study by Sari (2019) documented that the need for schools is rising substantially as the population increases. Finally, in the study by Topal (2019) in Saudi Arabia, it was concluded that women’s empowerment has been utilised as a code term to boost the economy’s competitiveness and expand international economic integration.

Evidence from previous studies has shown that most have concentrated on the ability to minimise or decrease GBV, especially domestic violence, with fewer studies focused on the role of economic empowerment and emotional violence (Bhana et al. 2021; Eger 2021; Nunbogu and Elliott 2022; Montserrat et al. 2022; del Mar Rodas-Zuleta et al. 2022; Freijomil-Vázquez et al. 2022). Another strand of studies focused on the micro-level (Bashford-Squires et al. 2022; Tsai 2022; Okpara and Anugwa 2022; Mcilongo and Strydom 2021), while other studies carried out systematic reviews (Keith et al. 2022; Sabri et al. 2022; van Daalen et al. 2022; Villardón-Gallego et al. 2023; Mbukanma and Strydom 2021). Further studies concentrated on health, music, and socioeconomic issues such as sexuality, water, sanitation, and hygiene (Nunbogu and Elliott 2022; Philbrick et al. 2022; Zhou et al. 2022; Mahamid et al. 2022; Raftery et al. 2022). However, the other goals identified by the United Nations (2020) have been neglected; that is, less attention has been given to the following goals: to ensure that women and girls benefit from social protection and economic stimulus measures; to encourage and promote the proportionate sharing of care work in the community; to increase the inclusiveness of women in decision-making and planning processes; and to include gender views in data and coordinating processes. Therefore, this study aimed to close a gap in the literature and strive for gender equality by examining the effects of economic empowerment on GBV. To achieve this, this study investigated the long-run effects of female economic empowerment on GBV in sub-Saharan African economies.

Consequently, our work can be distinguished from previous studies because it shifts attention to the macro-connection between factors that can reduce GBV and female economic empowerment in selected sub-Saharan African countries by using the panel fully modified least squares (panel FMOLS) estimation method. The panel FMOLS method is a reliable single-equation technique to analyse the long-term effect of independent variables on a dependent variable (Asaleye and Strydom 2023a). It is worth mentioning that GBV manifests itself in a variety of ways, including male rape, bisexuality, and transgenderisation; however, this study is limited to economic factors that hinder women’s and girls’ access to empowerment opportunities in sub-Saharan African countries (Asaleye and Strydom 2023b). In most cases, female learners can be abused, leading to them dropping out of secondary school or forcing them to cater for the family when their parents are engaged in business activities (Das 2023). Usman and Projo (2023) stressed the limit of the financial resources of parents, which is prevalent in Africa; parents often prefer to invest in a male child than a female child.

Based on the research gap identified in the empirical literature, this study aimed to examine the impact of female economic empowerment on induced GBV at the macro-level in sub-Saharan African countries. The specific objectives were as follows:

• To examine the connection between economic empowerment and female secondary school enrolment to reduce GBV.
• To investigate the impact of economic empowerment on female genital mutilation.
• To analyse the effect of economic empowerment on female students not enrolled in primary schools in sub-Saharan Africa in an attempt to reduce GBV.

3. Method

Theoretical Framework and Model Specification

The theoretical underpinning for the connection between GBV and economic empowerment is substantially built on the importance of female empowerment and socioeconomic development, as stressed by resource theory (Kenny et al. 2019). The model of the impact of economic employment on GBV outlined here included two channels through which female economic empowerment could influence GBV: female economic participation and female employment. This study was limited to these two indicators due to data availability. The indicators of female economic empowerment were separately analysed in each equation to avoid the presence of multicollinearity (Obadiaru et al. 2018). Therefore, the model was as follows:

$$GBV=f(ECP)$$

(1)

where $ECP$ is economic empowerment and GBV is gender-based violence. Explicitly, Equation (1) can be written as:

$$GB{V}_{i,t}=\alpha +{\alpha }_{2}EC{P}_{i,t}+{\epsilon }_{i,t}$$

(2)

The (i) represents the respective countries used in the study, (t) is the period of observation, and $\epsilon$ is the error term. According to resource theory, there is a positive relationship between economic empowerment and the danger of force, but it is less likely that force/violence is employed in the nexus (Walters et al. 2013). Based on the importance and relevance of other variables identified in the empirical literature, we added the following control variables: female population, female wages and salaries, productivity growth, and trade openness (Asaleye and Strydom 2023a). With the inclusion of control variables, Equation (2) was modified as follows:

$$GB{V}_{i,t}=\alpha +{\alpha }_{2}EC{P}_{i,t}+{\alpha }_{3}OT{P}_{i,t}+{\alpha }_{4}WA{S}_{i,t}+{\alpha }_{5}FP{P}_{i,t}+{\alpha }_{6}TR{O}_{i.t}+{\epsilon }_{i,t}$$

(3)

Equation (3) does not imply that no threats will be employed to retain control. On the contrary, it implies that threats may not necessarily manifest in physical violence, but emotional violence could be utilised. In other words, the aggression that a woman may face in this circumstance is emotional rather than physical. If someone lacks wealth, their sole option to control the relationship is often force or violence (Walters et al. 2013). Concerning this, the gender inequality index, the female genital mutilation prevalence, and female children out of school were used as proxies for GBV. It was expected that GBV would have a negative relationship with female economic empowerment. Hence, Equation (3) was adjusted to incorporate the dependent variables (indicators of GBV) as follows:

$$GI{I}_{i,t}={\beta }_1+{\beta }_{2}EC{P}_{i,t}+{\beta }_{3}OT{P}_{i,t}+{\beta }_{4}WA{S}_{i,t}+{\beta }_{5}FP{P}_{i,t}+{\beta }_{6}TR{O}_{i.t}+v_{i,t}$$

(4)

$$FG{M}_{i,t}={\delta }_1+{\delta }_{2}EC{P}_{i,t}+{\delta }_{3}OT{P}_{i,t}+{\delta }_{4}WA{S}_{i,t}+{\delta }_{5}FP{P}_{i,t}+{\delta }_{6}TR{O}_{i.t}+{\mu }_{i,t}$$

(5)

$$FC{S}_{i,t}={\lambda }_1+{\lambda }_{2}EC{P}_{i,t}+{\lambda }_{3}OT{P}_{i,t}+{\lambda }_{4}WA{S}_{i,t}+{\lambda }_{5}FP{P}_{i,t}+{\lambda }_{6}TR{O}_{i.t}+e_{i,t}$$

(6)

where $GII$ represents the gender parity index of secondary school enrolment, $FGM$ represents female genital mutilation prevalence, and $FCS$ is female children out of school. ${\beta }_1$, ${\delta }_1$, and ${\lambda }_1$ are the respective intercepts or constant terms for Equations (4)–(6). Also, ${\beta }_1,\dots ,{\beta }_6$, ${\delta }_1,\dots ,{\delta }_6$, and ${\lambda }_1,\dots ,{\lambda }_6$ are parameters. $v_{i,t}$, ${\mu }_{i,t}$, and $e_{i,t}$ are the respective error terms. From Equations (4)–(6), it was presumed that empowerment of the female gender could determine ‘out of female school students’ because of the assumptions that men generally take responsibility for fees and settle bills in the African context and that fewer efforts are made towards investing in female education than male education. Likewise, when female children are forced into early marriages, it affects their education. However, if their mothers are economically empowered, their children will be in school, thus reducing GBV.

As mentioned earlier, ECP was proxied by female economic participation and female employment. Therefore, the first set of estimations involved using female economic participation $EP1$ as a proxy for economic empowerment. For robustness, the second set of estimations used a close substitute for female economic participation: female employment $EP2$. Equations (4)–(6) were estimated using the panel FMOLS method introduced by Pedroni (2004) to achieve the objective of this study. Before the estimation of the panel FMOLS, preliminary analyses were carried out on the series; this included the descriptive statistics, correlation analysis, and unit root. Then, based on the unit root’s outcome, the study estimated the panel FMOLS, a reliable single-equation technique used to analyse long-term effects (Obadiaru et al. 2018). In Equations (4)–(6), the series was not stationary at any level, as shown in Table 3, but integrated within an order of one to become stationary. The data sources and descriptions are presented in

Table 1. Data sources and measurements.

Symbol	Variable	Description	Source
$EP1$	Female labour force participation rate	Percentage of people between the ages of 15 and 64 who are employed and active in production activities	International Labour Organization (ILO)
$EP2$	Female employment in industry	Working-age individuals involved in any activity to produce items or provide services for pay or profit	Same as EP1
$FCS$	Children out of school, female	Percentage of primary school-age children not enrolled in primary or secondary school	UNESCO Institute for Statistics
$GII$	School enrolment, secondary gender parity index	Ratio of girls to boys enrolled in public and private institutions at the secondary level	Same as FCS
$FGM$	Female genital mutilation prevalence rate	Percentage of women aged 15 to 49 who have partially or entirely removed the female external genitalia or other harm to female genital organs for cultural or non-therapeutic reasons	UNICEF data
$OTP$	GDP per capita growth	Percentage growth rate of a country in a constant local currency	World Bank national account data.
$WAS$	Female wage and salaries	Worker classifications for ‘paid employment occupations’, including wages and salaries	Same as EP1
$FPP$	Female population	Proportion of the total population that is female	Same as OTP
$TRO$	Trade openness	Aggregate inflow of goods divided by nominal GDP	Same as OTP

Source: authors’ compilation.

.

4. Results

The preliminary results are shown in

Table 2. Summary statistics.

	GII	FCS	FGM	EP1	EP2	FPP	OTP	TRO	WAS
Mean	0.7550	1.6874	1.4101	1.7735	0.7468	1.7018	1.8620	0.2268	0.978219
Median	0.7862	1.5191	1.6474	1.8157	0.7619	1.7006	2.1708	0.2194	1.009662
Maximum	1.1316	11.651	1.9863	1.9474	1.5109	1.7189	28.676	0.3353	1.873437
Minimum	0.2866	−0.3838	−0.5229	1.4582	−0.2676	1.6930	−31.333	0.1541	−0.207608
Std. Dev.	0.1729	1.3955	0.6214	0.1246	0.3878	0.0051	4.6007	0.0326	0.358484
S. Sq. Dev.	10.192	664.02	131.67	5.2966	51.273	0.0090	7217.7	0.3621	43.82220
Obs.	342	342	342	342	342	342	342	342	342

Source: authors’ computation.

,

Table 3. Correlation analysis.

	GII	FCS	FGM	EP1	EP2	FPP	OTP	TRO	WAS
GII	1
FCS	−0.6143	1
FGM	−0.3664	0.5283	1
EP1	−0.3039	0.1109	−0.1185	1
EP2	0.0252	0.1431	−0.202	−0.249	1
FPP	−0.0605	0.1829	0.1543	−0.1695	−0.2594	1
OTP	0.0846	0.0717	−0.2947	0.2499	0.386	−0.3073	1
TRO	0.1524	−0.2387	−0.0927	−0.0482	0.1479	−0.5459	0.1871	1
WAS	0.7314	−0.4975	−0.177	−0.5717	0.0737	−0.0328	−0.1899	0.0801	1

Source: authors’ computation.

and

Table 4. Unit root test results.

Variable	LLL		IPS		F-ADF		I (R)
	I (0)	I (1)	I (0)	I (1)	I (0)	I (1)
GII	6.76553	−18.5414 ***	5.66979	−12.0551 ***	5.53282	−10.1504 ***	I (1)
FCS	−5.08054 ***	−22.1620 ***	1.67156	−12.0072 ***	3.29429	−8.80165 ***	I (1)
FGM	−1.23369	−2.20136 ***	−0.41143	−0.61566	−0.78077	−1.56710 **	I (1)
EP1	1.94075	−2.57954 ***	9.75083	−1.72871 **	−0.12970	−1.51070 *	I (1)
EP2	−13.3959 ***	−12.1609 ***	−0.51205	−15.7824 ***	1.43064	−13.4946 ***	I (1)
FPP	−36.9737	−5.10694 ***	3.14433	0.67299	7.27942	1.89519 ***	I (1)
OTP	−16.0399	−34.2723 ***	−13.5096	−33.8109 ***	−13.8195	−23.5893 ***	I (1)
TRO	53.2142	116.522 **	6.67546	15.5786 **	7.02251	15.3989 *	I (1)
WAS	2.43381	−12.1141 ***	8.35779	−11.0905 ***	8.10897	−10.0850 ***	I (1)

*, **, and *** indicate significance at 10 per cent, 5 per cent, and 1 per cent, respectively. Source: authors’ computation. I (R): order of integration (summary); LLL: Levin, Lin, and Chu; IPS: Im, Pesaran, and Shin; F-ADF: Fisher-ADF.

, while

Table 5. Cointegration results.

Model 1: Kao Residual Cointegration Test
Null Hypothesis: No Cointegration
	t-Statistic	Prob.	Spectral Estimation
ADF	5.152268 ***	0.0000	Bartlett
ADF	4.744271 ***	0.0000	Parzen
ADF	4.885887 ***	0.0000	Quadratic Spectral
Model 2: Kao Residual Cointegration Test
Null Hypothesis: No Cointegration
	t-Statistic	Prob.	Spectral Estimation
ADF	4.218169 **	0.0033	Bartlett
ADF	2.766044 **	0.0218	Parzen
ADF	1.848954 *	0.0980	Quadratic Spectral
Model 3: Kao Residual Cointegration Test
Null Hypothesis: No Cointegration
	t-Statistic	Prob.	Spectral Estimation
ADF	4.707647 **	0.0396	Bartlett
ADF	2.982547 *	0.0629	Parzen
ADF	1.808226 *	0.0567	Quadratic Spectral

*, **, and *** indicate significance at 10 per cent, 5 per cent, and 1 per cent, respectively. Source: authors’ computation.

presents the long-run regression model for the panel FMOLS estimation.

Table 2 presents the descriptive statistics of the study, providing insights into the nature of the data used in this study. It can be seen that female wages and salaries had the highest mean value of 0.978219. The mean values for the ratio of girls to boys in public and private secondary school enrolment (GII), female children out of primary school (FCS), female genital mutilation (FGM), female labour force participation (EP1), female employment in industry (EP2), female population (FPP), gross domestic product per capita (OTP), and trade openness (TRO) were 0.7550, 1.6874, 1.4101, 1.7735, 0.7468, 1.7018, 1.8620, and 0.2268, respectively. The standard deviations for GII, FCS, FGM, EP1, EP2, FPP, OTP, TRO, and WAS were 0.1729, 1.3955, 0.6214, 0.1246, 0.3878, 0.0051, 4.6007, 0.0326, and 0.358484, respectively. The statistics revealed that most women between the age of 15 and 49 were genitally mutilated within the region because the mean value of FGM was closer to the maximum than the minimum value and the standard deviation was not fundamentally different from the mean. This situation was similar to that of children out of school, while the standard deviation of the female wages and salaries showed a large disparity from the mean, an indication that a few females are potentially paid an above-average income, with the mean hovering around the median wages and salaries.

Table 3 presents the correlation analysis of the variables used in this study. Evidence from the results showed that GII had a negative correlation with FCS, FGM, EP1, and FPP, with values of −0.6143, −0.3664, −0.3039, and −0.0605, respectively; it had a positive correlation with EP2, OTP, TRO, and WAS, with values of 0.0252, 0.0846, 0.1524, and 0.7314, respectively. FCS positively correlated with FGM, EP1, EP2, FPP, and OTP, with values of 0.5283, 0.1109, 0.1431, 0.1829, and 0.0717, respectively; it had a negative relationship with TRO and WAS, with values of −0.2387 and −0.4975, respectively. The FGM variable was negatively correlated with EP1, EP2, OTP, and TRO, with values of −0.1185, −0.202, −0.2947, −0.0927, and −0.177, respectively; it had a positive correlation with FPP, with a value of 0.1543. EP1 had a positive correlation with OTP, with a value of 0.2499, and a negative correlation with EP2, FPP, TRO, and WAS, with values of −0.249, −0.1695, −0.0482, and −0.5717, respectively. Furthermore, EP2 had a negative correlation with FPP, with a value of −0.2594, and a positive correlation with OTP, TRO, and WAS, with values of 0.386, 0.1479, and 0.0737, respectively. FPP had a negative correlation with OTP, TRO, and WAS, with values of −0.3073, −0.5459, and −0.0328, respectively. OTP had a positive relationship with TRO, with a value of 0.1871, and a negative correlation with WAS, with a value of −0.1899. Finally, a positive correlation existed between TRO and WAS, with a value of 0.0801. A cursory interpretation of the correlation analysis revealed that wages showed a high positive association with the parity index at 73.14%, which conformed a priori as it was expected that an increase in wage would increase the number of school enrolments. This supported the hypothesis of this study, asserting that economic empowerment reduces GBV. Likewise, female labour force participation, EP1, was positively associated with school enrolments and negatively correlated with out of school children, FCS, which further supported the study hypothesis. It could be argued that, despite the level of wages of the female gender evidenced by the descriptive statistics, the impact of economic empowerment was vivid. Although the correlation analysis did not suggest relationships, it provided an insight into the data behaviour, with the potential to predict the direction of relationships, as seen in the regression analysis.

Table 4 presents the unit roots of the variables used in this study. The null hypothesis was that the variables were stationary at a significance level of five per cent. Evidence from the results showed that when using IPS and F-ADF, the school enrolment, secondary (GII); children out of school, female (FCS); female genital mutilation prevalence (FGM); labour force participation rate, female (EPI); employment in industry, female (EP2); population, female (FPP); GDP per capita growth (OTP); trade openness (TRO); and WAS (wage and salaried workers, female) were not stationary at the level form. The result of LLL showed that FCS and EP2 were stationary at a one per cent significance level at the level form, while the others were not stationary. Therefore, the null hypothesis that the variables were stationary at the level was rejected.

However, this study could not reject the hypothesis at the first difference; evidence from the results using LLL showed that GII, FCS, FGM, EP1, EP2, FPP, OTP, and WAS were all stationary at a one per cent significance level. At the same time, TRO was stationary at a five per cent significance level. Likewise, the results using IPS showed that GII, FCS, EP2, OTP, and WAS were stationary at a one per cent significance level. In comparison, EP1 and TRO were stationary at five per cent. In addition, FGM and EPP were not stationary. Finally, the F-ADF results showed that GII, FCS, EP2, EPP, OTP, and WAS were stationary at one per cent, while FGM was at a five per cent significance level and TRO was at a ten per cent significance level. The general conclusion from the unit root results was that all the variables were integrated by an order of one to become stationary. Based on the evidence of these results, we estimated the long-run relationship using the panel FMOLS method. This study employed three tests of unit roots to ensure robustness and reduce the chances of errors.

Table 5 presents the cointegration results for the variables used in this study. The null hypothesis was that there was no cointegration between the series; this hypothesis was rejected at various levels of significance in the three models investigated in this study. Based on the evidence of these results (unit root and cointegration tests), we estimated the long-run relationship for the three models’ relationship using the panel FMOLS method.

Table 6. Long-run regression models: panel FMOLS.

Model 1: Using GII as a Dependent Variable
Variable	Coefficient	Std. Error	t-Statistic	Prob.
EP1	−0.595488 ***	0.221959	−2.682871	0.0075
FPP	−5.179956 ***	1.625269	−3.187137	0.0015
OTP	00000.2	0.000747	0.032217	0.9743
TRO	1.292531 ***	0.199398	6.482167	0.0000
WAS	0.155069 ***	0.037625	4.121423	0.0000
R-Squared	0.913034
Adjusted R-Squared	0.906564
Model 2: Using FCS as a Dependent Variable
Variable	Coefficient	Std. Error	t-Statistic	Prob.
EP1	−1.750749 ***	0.515848	−3.393922	0.0007
FPP	3.523820 ***	0.628012	5.611071	0.0000
OTP	−0.005213	0.010587	−0.492394	0.6226
TRO	−2.321002	1.576645	−1.472115	0.1414
WAS	−0.694139 ***	0.139553	−4.974033	0.0000
R-Squared	0.695458
Adjusted R-Squared	0.656769
Model 3: Using FGM as a Dependent Variable
Variable	Coefficient	Std. Error	t-Statistic	Prob.
EP1	−0.809446 **	0.401126	−2.017933	0.0442
FPP	12.13309 ***	4.403087	2.755587	0.0061
OTP	−0.000727	0.001042	−0.697822	0.4857
TRO	0.248654	0.280783	0.885574	0.3764
WAS	−0.236893 ***	0.057068	−4.151068	0.0000
R-Squared	0.988778
Adjusted R-Squared	0.987986

**, and *** indicate significance at 5 per cent, and 1 per cent, respectively. Source: authors’ computation.

presents the long-run regression results using the ratio of girls to boys in public and private secondary school enrolment (GII), female children out of primary school (FCS), and female genital mutilation (FGM) as dependent variables, which are referred to as models 1, 2, and 3, respectively. Using GII as the dependent variable, it could be seen that it had a significant relationship with female labour force participation (EP1), female population (FPP), trade openness (TRO), and female wages and salaries (WAS), respectively, at a 1 per cent significance level, while trade openness (OTP) was not statistically significant. Keeping all variables constant, a 1 percentage point change in EP1 caused a 0.6 per cent reduction in GII. Likewise, a 1 percentage point change in FPP resulted in a 5.2 reduction in GII. A 1 percentage point change in TRO caused a 1.3 per cent increase in GII. Finally, a 1 percentage point change in WAS caused a 0.2 per cent decrease in GII.

In model 2, using FCS as the dependent variable, it could be seen that it had a significant relationship with EP1, FPP, and WAS, respectively, at a 1 per cent significance level. At the same time, OTP and TRO were statistically insignificant. A 1 percentage point change in EP1 caused a 1.2 per cent reduction in FCS. Likewise, a 1 percentage point change in FPP resulted in a 3.5 per cent increase in FCS. Finally, a 1 percentage point change in WAS caused a 0.69 per cent decrease in FCS.

In model 3, using FGM as the dependent variable, it could be seen that it had a significant relationship with EP1, FPP, and WAS, respectively. The significance was at a 1 per cent level for FPP and WAS, and 5 per cent for EP1. Therefore, a 1 percentage point change in EP1 resulted in a reduction of 0.81 per cent for FGM. Likewise, a 1 percentage point change in FPP resulted in a 12.1 per cent increase in FGM. Conversely, a 1 percentage point change in WAS caused a 0.24 decrease in FGM.

5. Discussion

In model 1 of this study, it was shown that an increase in female labour participation reduced the ratio of girls to boys in secondary school enrolment. In model 2, it was observed that the increase in female labour force participation reduced the number of female children out of primary school; this finding was in line with the study of Das (2023). An increase in economic participation should increase female economic power, which should reduce GBV. Likewise, models 1 and 2 depicted that an increase in the female population reduced the ratio of girls to boys in school enrolment and the dropping out of female students from secondary school. As the population increases, learners—especially girls—have a high risk of dropping out of school; our result aligned with the results of the study by Usman and Projo (2023). In addition, Sari (2019) stressed that the need for schools substantially rises as the population increases. Models 1 and 2 in our study also showed that increasing female wages and salaries promoted the ratio of girls to boys in secondary and primary school. This outcome aligned with the findings of Usman and Projo (2023), who stated that the limits of parents’ financial resources often reduce investment in the education of a female. The finding was further supported by Buvinic et al. (2022).

Women’s empowerment, particularly through increasing economic participation and employment, has been shown to be a successful strategy in mitigating GBV (Asaleye and Strydom 2023b; Bashford-Squires et al. 2022; Tsai 2022). Economic empowerment grants women financial autonomy and a more substantial role in decision-making processes. This shift in power dynamics has the potential to significantly reduce instances of GBV. Studies consistently show that as women’s income increases, the prevalence of violence against them decreases (Sabri et al. 2022; Kreager et al. 2013; Walters et al. 2013). Economic empowerment equips women with the capacity to exit abusive relationships as they become less financially reliant on their spouses. Moreover, the availability of employment opportunities and access to resources empowers women to challenge cultural norms and expectations, fostering a more equal and just society (Isaacs et al. 2022; Netnou and Strydom 2020).

In model 3, this study observed that increased female labour participation reduced female genital mutilation caused for cultural or non-therapeutic reasons, although the results shown in model 3 were that as the female population increases, female mutilation also increases. This result aligned with resource theory (Walters et al. 2013). The conclusion to be drawn from this was that an increase in population without being empowered reduces the voice or impact that requires change and can increase the level of GBV.

Table A1. Robustness check (long-run regression model: panel FMOLS).

Model 1: Using GII as a Dependent Variable
Variable	Coefficient	Std. Error	t-Statistic	Prob.
EP2	−0.105195 ***	0.036818	−2.857188	0.0044
FPP	−4.851761 ***	1.626038	−2.983792	0.0030
OTP	0.000009	0.000750	0.120226	0.9043
TRO	1.332863 ***	0.199635	6.676492	0.0000
WAS	0.118178 ***	0.039536	2.989159	0.0029
R-Squared	0.912820
Adjusted R-Squared	0.906314
Model 2: Using FCS as a Dependent Variable
Variable	Coefficient	Std. Error	t-Statistic	Prob.
EP2	−0.491467 ***	0.146562	−3.353294	0.0008
FPP	1.715782 ***	0.226218	7.584649	0.0000
OTP	−0.004908	0.010591	−0.463360	0.6432
TRO	−2.854499 *	1.593028	−1.791870	0.0736
WAS	−0.288198 **	0.140483	−2.051480	0.0406
R-Squared	0.938864
Adjusted R-Squared	0.833598
Model 3: Using FGM as a Dependent Variable
Variable	Coefficient	Std. Error	t-Statistic	Prob.
EP2	−0.123756 *	0.069093	−1.791149	0.0740
FPP	13.18157 ***	4.300047	3.065447	0.0023
OTP	−0.000799	0.001009	−0.791265	0.4292
TRO	0.386332	0.280113	1.379203	0.1686
WAS	−0.255220 ***	0.055365	−4.609778	0.0000
R-Squared	0.988709
Adjusted R-Squared	0.987913

*, **, and *** indicate significance at 10 per cent, 5 per cent, and 1 per cent, respectively. Source: authors’ computation.

in the Appendix A section presents the robustness check of our panel FMOLS method. The female economic participation variable (EP1) was replaced by female employment (EP2) to perform the check. The evidence in Table A1 validated the outcome of our regression analysis in the results, except for a few changes. In model 1, the outcome of the estimation result in Table 5 was the same as in Table A1. In model 2, there was a slight difference between Table 5 and Table A1 regarding the significance of trade openness (TRO) at 10 per cent; the TRO was insignificant in Table 5. Also, in model 3, Table 5 differed a little from Table 6. In Table 6, it can be seen that EP2 was significant at 10 per cent compared with Table 5, where EP1 was significant at 5 per cent. However, all the signs of the variables were the same in Table 5 and Table 6, except for the insignificant variables. Therefore, the outcome of the robustness check validated the causal inference between the dependent and independent variables used in this study by the insensitivity of the replaced variable of female economic empowerment in the regression analysis.

6. Conclusions and Recommendations

Most studies relating to GBV have focused on the micro-impact or have paid less attention to the economic factors, which were the main thrust of this study. In addition, evidence from recent official statistics shows that one in two women in several sub-Saharan African nations has suffered GBV, affecting around 37 per cent of women globally. Therefore, this study has contributed to the empirical literature by investigating the long-term implications of economic empowerment on GBV in sub-Saharan African countries using a fully modified least squares panel. The link between women’s economic empowerment and a reduction in GBV was acknowledged. By addressing the underlying causes of inequality and providing women with equal opportunities in the workplace, we can potentially exert a positive influence on societal perceptions and behaviour.

The key findings in the study showed that an increase in female labour participation reduced the ratio of girls to boys in secondary school enrolment and reduced the number of female children out of primary school. The findings also showed that an increase in the female population increased the dropout rate of female children in primary and secondary schools. Finally, the findings indicated that female labour participation had a negative relationship with female genital mutilation. The main conclusion from this study was that an increase in the economic power of females through increased economic participation reduced GBV in the long run.

Based on the findings, this study recommends implementing stricter policies to protect victims, especially females, who are frequently subjected to emotional abuse, given the inherent challenge of identifying such forms of violence. Consequently, societal transformation is imperative to promote female economic empowerment as this initiative may serve as a means to counteract emotional abuse against women. It is further recommended that governments create awareness and forums to cater for young females in primary and secondary schools to increase the rate of human capital formation at this level; this will assist in decreasing GBV in the long run. Furthermore, incentives in the form of scholarships should be implemented to increase the enrolment of female scholars at schools in Africa. Finally, there is a need for more campaigns, support groups, and proactive women’s organisations to integrate and promote economic empowerment among fellow women and increase their collective voices against GBV. Moreover, employers should be encouraged to formulate policies supporting work–life balance, childcare, and flexible scheduling, thus facilitating women’s active participation in the labour force. Awareness campaigns should be implemented to challenge preconceptions and stigmas surrounding women’s responsibilities, both in the workplace and at home.

This research was limited by constraints in data-collection methods, particularly in obtaining data on domestic violence. This study relied on macro-data to investigate how female economic empowerment could mitigate factors contributing to GBV. For future studies, it is recommended that micro-data are incorporated to deepen our understanding of the relationship between GBV and female economic empowerment.