Gender and ethnic earnings gaps in seven West African cities

Abstract

a b s t r a c t JEL classification: J31 J71 O15 O55 Keywords: Earnings equations Gender wage gap Ethnic wage gap West AfricaIn this paper, we measure, compare and analyse gender and ethnic earnings gaps in seven West African capitals using data from an original series of urban household surveys. Our results show that gender earnings gaps are large in all the cities in our sample with significant variations across cities. Cities with large gender earnings gaps are also where gender education gaps are wider and where the female labour market participation is highest. Decomposition of the gender gaps shows that differences in characteristics explain around 40% of the raw gender gap on average, but this varies somewhat across cities. The results of the full decomposition of the gender earnings gaps suggest that differences in sector allocation contribute, on average, to one third of gender earnings gaps. Gender gaps are very wide in the informal sector and differences in micro-firm characteristics also account for differences in self-employment earnings. In contrast to the large gender earnings gaps measured in the seven cities, majority ethnic groups do not appear to be in a systematically advantageous position on the urban labour markets in our sample of cities, and observed gaps are small compared with gender gaps. Looking at more detailed levels of ethnic disaggregation, ethnic earnings differentials are found to be systematically smaller than gender differentials. Moreover, none of the minority "favoured" groups seem to have any relation to the ethnicity of the Head of State at the time of the survey. Holding productive characteristics constant, some unexplained differences persist however.

Female leaders and gender gaps within the firm: Evidence from three sub-Saharan African countries

“…Nordman and Roubaud (2009) showed the importance of measuring women's actual experience in estimating the portion of the gender wage gap that is explained by observable characteristics. Nordman et al (2011) analysed the gender gap in earnings in seven cities in West Africa and found that it was larger in the informal sector than in the public and private formal sectors. As these studies showed how differences in workers, jobs, and firms' characteristics affect the gender wage gap in sub-Saharan Africa, we control for these variables in our empirical strategy.…”

“…Our paper contributes to a growing literature on the determinants of the gender wage gap in sub-Saharan Africa by showing new evidence of the association between female CEO and the gender wage gap. Some of the previous studies used household and labour force surveys (Appleton et al 1999;Nordman and Roubaud 2009;Nordman et al 2011). Since employment segregation by gender is one of the main sources of gender differences in earnings in developing countries, and firms that employ women differ from firms that employ men, examining the role of firm characteristics is imperative to a study of the factors that contribute to the gender wage gap in developing countries (World Bank 2012;Borrowman and Klasen 2015).…”

Polygamy and female labour supply in Senegal

“…Senegalese females' participation in the labour market is particularly low in the West African context (Nordman et al 2011). The female paid employment rate in Dakar was 35.7 per cent in the early 2000s, while it was about-or up to-50 per cent in Bamako, Cotonou, Abidjan, and Lomé.…”

Labor Market Outcomes of Youth and Women in Newly Industrialized and Developing Countries

“…Bhorat and Goga (2013) examine the gap along the wage distribution and find that 58.8% or more of the gap at all quantiles is due to difference in returns. Christophe J Nordman, Robilliard, and Roubaud (2011) compare the gender earnings gap in seven West African capital cities and find substantial raw gender earning gap that ranges between 0.500 and 0.792; 56.4% to 67.1% of this gap is due to difference in returns (with the exception of Lome where 45.7% of the gap is unexplained). When using different decomposition methods, the results may show a variation but it does not substantially change their implications.…”