A gendered resource curse? Mineral ownership, female unemployment and domestic violence in Sub-Saharan Africa

Abstract

Several studies suggest that the extractive industry has negative consequences for gender equality despite the often positive growth impact of natural resources. We re-examine this claim at the sub-state level in sub-Saharan Africa and argue that we need to differentiate between ownership arrangements in the extractive industry. To test our argument on the gender dimension of the resource curse, this article employs unique data on the control rights of minerals within sub-Saharan countries as well as data from Afrobarometer and Demographic and Health Surveys (DHS). Our quantitative analyses explore how international vs. domestic ownership of copper, diamond and gold mines affects the labor market integration of females and intimate partner violence. The regression results suggest in line with our theoretical expectations that gender-specific structural labor market shifts within extractive industries are contingent on mineral control rights. Our models show that within mining areas, only domestic ownership reduces male unemployment. While domestic mining seems to reinforce the traditional male breadwinner model, internationally owned mineral extraction induces structural labor market changes: women

The rotating Presidency of the Council of the EU – Still an agenda-setter?

“…To the best of my knowledge, proper fractional dynamic panel models have not yet been fully developed. Alternatives include the Arellano–Bond estimator, as well as a generalised linear model (GLM) adapted for panel estimates of a fractional dependent variable (Krauser et al., 2019). While the Arellano–Bond estimator allows a proper inclusion of lagged dependent variables, it suffers from issues in dealing with datasets where the number of time periods is larger than the number of groups, such as in the specific case at hand.…”

“…For the purpose of this study, the modelling strategy of Krauser et al. (2019: 225) is applied adapting a GLM to accommodate both the fractional nature of the dependent variable as well as the panel nature of the dataset.…”

“…In the baseline model, Presidency salience indicates the level of salience in the Presidency programme of a given policy area p during Presidency t, government salience identifies the salience for the government of policy area p , ideological placement captures the placement of the government on the left/right scale, net recipient country is a binary variable identifying countries as either net recipient or net contributors to the EU budget, geographical position classifies countries as either Central–Eastern, Southern, or North-Western Europe, and ε is the error term. I also include the mean at Presidency t, policy p of the main explanatory variable of interest, government salience, to model the panel structure of the data as recommended by Papke and Wooldridge (2008) and Krauser et al. (2019).…”

“…(2019). 9 Similarly, given the fractional nature of the dependent variable, the equation is fit using a fractional model with a logit link (Krauser et al., 2019).…”

“…The resulting panel dataset, once augmented with information on domestic factors of interest, is analysed by means of panel model estimations. In the article, I present the estimates from a general linear model adapted for panel data, following Krauser et al. (2019).…”