Inequality and female labour force participation in West Africa

This study examined the impact of income inequality on female labour force participation in West Africa for the period 2004 to 2016. The study employed the Gini coefficient, the Atkinson index and the Palma ratio as measures of income inequality. For robustness, the study also utilises female employment and female unemployment as measures of female labour force participation. The study employed the instrumental variable fixed effects model with Driscoll and Kraay standard errors to account for simultaneity/reverse causality, serial correlation, groupwise heteroskedasticity and crosssectional dependence. The empirical results reveal that the three measures of income inequality significantly reduce the participation of women in the labour force in West Africa. The study also revealed that domestic credit, remittances and female education are positively associated with female labour force participation in the sub-region. Further findings reveal that economic development reduces the participation of women in the labour force in West Africa with the U-shaped feminisation theory not valid for the West African region. The study, however, revealed an inverted U-shaped relationship between inequality and female unemployment. Policy recommendations based on these findings are discussed.


Introduction
Achieving food security, poverty alleviation, gender equality, full employment and economic development for all in Africa, and particularly West Africa, will be a mirage if the challenge of poor female labour participation is not addressed. To achieve the Sustainable Development Goals (SDGs) by 2030, closing the gender gap in the labour force is quite indispensable (United Nations, 2018), and explains why researchers are now paying greater attention to female labour force participation. Women in West Africa, like their counterparts elsewhere, are very resourceful economic agents whose full participation in the economy's labour force can contribute greatly to economic growth. In many developed economies, women make up almost half of the labour force (World Development Indicators (WDI), 2020), which could partly be the reason for the fast developmental pace of these economies, and implies that women make many significant contributions to economic development when harnessed.
According to the International Monetary Fund (IMF, 2018), increasing women's engagement in the economy's labour force leads to significant increases in productivity and impact positively on economic growth. Akinyemi, Solanke and Odimegwu (2018) also observe that female employment reduces the level of infant and child mortality. Despite its positive impact, data from WDI (2020) has consistently revealed that across West African countries, female labour force participation has remained dismal. Abney and Laya (2018) has noted that there are fewer women in the formal economic sector, which is particularly unfavourable for economic and human prosperity, as economies are unable to reap the trickle down positive spill overs of female labour force participation on other socio-economic indicators.
In addition to the low participation in the labour market, women in West Africa face significant discriminations. In the world of labour in West Africa, patriarchal ideology, educational attainment, household responsibilities and maternal commitment have led to the historic marginalisation of women. These factors automatically transmit to unequal opportunities for men and women. The feminization hypothesis also posits that in the early stages of economic development, female labour force participation declines. This is as a result of the adjustments in the structure of the economy from one which is based on agriculture to one based on industrialisation. However, in the later stages of economic development, with economies transforming into modern economies, fertility rates tend to decline accompanied with the rise in female education, leading to the rise in female labour force participation. This denotes a U-shaped relationship between economic development and female labour force participation.
However, one economic fundamental that has not gained much prominence in economic discourse pertaining to female labour force participation in West Africa is income inequality.
Several studies have established links between income inequality and women participation in the labour force. Studies such as (Gronau, 1982;Nelissen, 1990;Björklund, 1992;Del Boca & Pasqua 2003;Amin & DaVanzo 2004;Western, Bloome & Percheski 2008;Harkness, 2010) found a strong link between income inequality and female employment. These studies reflect that while female employment may reduce income inequality, income inequality can also act as a determining factor which could deteriorate female labour participation. A consistent theory on the effect of gender inequality on employment by Klasen and Lamanna (2009) is that income inequality changes economic realities negatively and motivates externalities that limit opportunities for women, leading to increase in female unemployment and participation in the labour market. This belief pushed further by Asongu and Odhiambo (2019) are broadly consistent with the literature on relationships between unemployment, income inequality and economic prosperity (Witte & Witt, 2001;Brush, 2007;Odedokun & Round, 2001;Perugini & Martino, 2008; Van der Hoeven, 2010). Furthermore, income inequality also drives poverty which remains a determining factor for not just the participation of women in the labour market but also the participation of men in the labour market. Hence, to improve the female labour force participation rate, it becomes necessary to study how income inequality influences the labour force participation of women.
There are relatively few studies on income inequality and the economic participation of women in Africa. One such study is that of Asongu and Odhiambo (2019). This study, however, drifts away from their study by (1) accounting for cross-sectional dependence in the modelling exercise, and (2) employing three different measures of female labour force participation to account for robustness. This study also concentrates on West Africa, a region where poverty is substantially prevalent with a high rate of infant and child mortality. Another study that investigated the influence of inequality on female labour force participation in Africa is that of Idowu and Owoeye (2019). Their study, however, also failed to account for cross-sectional dependence which, according to Iheonu (2019), can lead to estimation bias.
The present study whose scope covers 16 West African economies from 2004 to 2016 is novel in several ways. First, the study analysed the relationship between income inequality and female labour force participation in West Africa, a region where there is a dearth of studies despite the vulnerable nature of the region, and the need to attain the SDGs. Second, the study employed three measures of income inequality-the Gini coefficient, the Palma ratio and the Atkinson index, and female labour force participation rate to capture female labour participation. For robustness purpose, we utilise female employment and the female unemployment rate as they are measures of the economic participation of women (Asongu and Odhiambo (2019), (3) the study applied the instrumental variable Fixed Effects (FE) model to account for simultaneity/reverse causality as well as the Driscoll and Kraay (1998) standard errors to account for heteroskedasticity, serial correlation and cross sectional dependence. Based on the objective of this study, the testable hypothesis is that income inequality has no significant influence on female labour force participation in West Africa. The remainder of this study is structured thus: section two presents a review of the related theoretical and empirical literature, section three discusses the methodology and, while section four presents the estimation results and discusses the findings.
Section five concludes the study and provides policy recommendations.

Literature review
Theories on female labour force participation can be traced back to the classical labour theory (CLT). Between 1960 and 1970, the fundamental speculations on female labour force participation rate rose significantly. The theories include the work-leisure hypothesis of Mincer (1962), the human capital theory of Becker (1964), and the time allocation hypothesis of Becker (1965). The theoretical underpinning of the study, however, follows that of Asongu and Odhiambo (2019), and Klasen and Lamanna (2009). This is based on the premise that income inequality alters the workings of the economy and propels negative externalities that limit the opportunities participation. This means that in the early stages of economic development, the participation of women in the labour market declines but starts increasing at later stages of economic development.

Method
In understanding the influence of income inequality on female labour force participation in West Africa, the study employs the instrumental variable fixed effects model due to the following reasons (1) the instrumental variable fixed effects model is robust to simultaneity/reverse causality -a notable feature in income inequality and female labour force participation relationship, and (2) it accounts for unobserved heterogeneity in the modelling exercise. Following the study of Efobi et al. (2019), the issue of simultaneity/reverse causality is tackled through the process of instrumenting the indicators of inequality as well as the control variables with their first lags.
where , is an explanatory variable (i.e. income inequality indicators and other control variables) in country i at time t. 0 is a constant, , −1 is the first lag of the explanatory variable and , is the error term. In obtaining the instruments for the explanatory variables for the fixed effects The study presents a fixed effects model, where , = 0 + 1 , + , + + , .
Here, , is female labour force participation rate. Female labour participation rate captures the ratio of total female labour force to total female population. For robustness checks, we re-define the female labour force participation as female employment and female unemployment rates, Azizi, 2018; Asiedu and Chimbar, 2020). In this study, we posit a positive relationship between remittances and the participation of women in the labour market, particularly in the informal sector.
We also include female education into the model as education is a significant determinant of female labour force participation (Ince, 2010) via the increase in human capital. is country specific effect and , is the error term. The estimation of the fixed effects model utilises the Driscoll and Kraay (1998) standard errors which is robust to serial correlation, groupwise heteroskedasticity, and cross-sectional dependence (see, Iheonu, 2019;Iheonu et al., 2019).

Data
The

Descriptive statistics
The descriptive statistics presents the means, the minimum and maximum values, the standard deviation of each of the variables as well as the total number of observations. First, Table 2 shows that the dataset is unbalanced. It is also revealed that there is no large difference in the minimum and maximum values of the income inequality indicators, which suggests similarities across space and time. The study, however, reveals substantial differences in the minimum and maximum values of the indicators of labour force participation, which invariably reflects variations across West Africa and the observed time period. Domestic credit to the private sector and female education exhibits similar characteristics. Remittances and the natural logarithm of GDP per capita, however, do not have a substantial difference between the minimum and maximum values which is clearly shown by their standard deviations of 0.89 and 1.48, respectively.

Results
Empirical results in table 3, table 4 and table 5  The negative relationship between GDP per capita and female labour force participation is in line with the study of Altuzarra, Galvez-Galvez and Gonzalez-Flores (2019). In West Africa, it is plausible that high female labour force participation falls with economic development and rises with an economic decline. This is based on the intuition that women tend to support their families when the household is faced with financial constraint, particularly in times when the economy is on a decline. However, it is revealed that a further increase in GDP per capita tends to reduce the negative impact of GDP per capita on female labour force participation, as revealed by the coefficient value. This result contradicts the U-shaped hypothesis on the relationship between economic development (as measured by GDP per capita) and female labour force participation.
In West Africa, later stages of economic development also have a negative consequence on female labour force participation.  Note: *** and ** denotes statistical significance at 1% and 5% respectively.
In able 4, the results reveal that the three indicators of income inequality significantly reduce female employment in West Africa. This transmits to the fact that income inequality leads to a fall in the participation of women in the labour force. The result supports the finding of Asongu and Odhiambo (2019). Their study revealed that income inequality reduces the participation of women in economic activities. Domestic credit, remittances and female education have a positive and significant impact on female employment in West Africa. However, GDP per capita and the square of GDP per capita significantly reduces female employment in the sub-region. These results highlight the importance of credit access, remittances and education to female employment.
Employing the female employment rate as a proxy for female labour force participation supports the findings in table 3, where a U-shaped relationship between GDP and female labour force participation do not exist. Instead, it is revealed that the rise in GDP per capita initially has a more pronounced negative consequence on female employment in West Africa. However, the negative consequence becomes less pronounced over time. In table 5, female unemployment is utilised as the proxy for female labour force participation.

It is revealed that female unemployment is an increasing function of income inequality in West
Africa. However, we do not see any statistically significant relationship between income inequality and female unemployment rate in West Africa. Note: *** and ** denotes statistical significance at 1% and 5% respectively. Furthermore, it is also revealed that remittances significantly reduce female unemployment while domestic credit and female education reduces female unemployment insignificantly.
Furthermore, GDP per capita is revealed to significantly increase the female unemployment rate in West Africa. However, the square of GDP is seen to significantly reduce the female unemployment rate in the region. This result shows the existence of an inverted U-shaped relationship between economic development and female unemployment. Earlier stages of economic development results in the increase in the female unemployment rate. However, later stages of economic development eventually result in a fall in the female unemployment rate.

Conclusion and relevant policy recommendations
The study has investigated the impact of income inequality on female labour force participation in West Africa utilising three measures of income inequality. The study employed the instrumental variable fixed effects model in order to correct for the possibility of simultaneity/reverse causality and individual heterogeneity. The study finds that the Gini coefficient, Atkinson index and the Palma ratio significantly influences female labour force participation rate and female employment negatively. However, the three indicators of income inequality increase female unemployment rate but not at any level of statistical significance. This result is consistent with the findings of Asongu and Odhiambo (2019). Further findings from the control variable show that domestic credit to the private sector, remittances and female education significantly increases the female labour force participation rate and female employment in West Africa.
Also, remittances and female education reduce female unemployment in West Africa. While remittances are revealed to be significant, female education is seen to be insignificant. The result has revealed the importance of migrant remittance to female labour force participation and as such, the importance of removing bottlenecks to enable the smooth flow of remittances into West Africa. Findings also reveal that GDP per capita and the square of GDP per capita decreases female labour force participation rate and female employment which disproves the U-shaped hypothesis for West Africa. We, however, found an inverted U-shaped relationship between GDP per capita and female unemployment, revealing that an initial response to economic development is the rise in female unemployment with a subsequent decline in female unemployment in West Africa at later stages of economic development. The study recommends the following based on findings, (1) there is an essential need for the adoption of inequality reducing policies in West Africa. The governments in the sub-region should also invest in sectors where women have a greater likelihood of participation. There is also need for quality infrastructure that drives economic activities as well as the adoption of a tax system that is progressive in nature to reduce income inequality, (2) it is necessary that the sub-region concentrates in developing the financial sector to improve access to credit. This can be achieved through a significant reduction in interest rates in financial institutions to enable borrowing, (3) the development of the financial sector would also reduce bottlenecks which would aid the increase in remittances flow into the West African subregion, (4) human capital development in the form of female education should also be prioritised in West Africa. The positive linkage between female education and female labour force participation should be a focal point for West African countries on the need to pursue the girl child education. Finally, the study also reflects on the importance of changing the socio-cultural norms in Africa, where many women participate more in the labour market only when economic activities are declining in order to support their spouses. As highlighted, in periods of rising per capita GDP, women draw out of the labour market and tend to cater more for their husbands and children leaving their husbands to cater for the family. Conclusively, the feminisation theory fails to hold in West Africa.