Using plot-level data from Cameroon, we document how gender disparity in productivity varies according to how plot headship is defined and distinguished by gender. We account for selectivity bias and obtain direct and indirect drivers of gender disparities through the implementation of an extended Oaxaca-Blinder decomposition and distributional decomposition using percentile-weighted regressions. We find that gender disparities differ by headship and result from unobserved factors with women’s structural disadvantage exceeding men’s structural advantage. Direct contributors to gender disparities are gender-specific: (i) such non-labor inputs as fertilizer (plot head and plot owner) and cost of irrigation (de jure plot head); (ii) age (migrant plot head); and (iii) plot size (plot manager). Factors that drive these major contributors and, thus, indirectly affect the components of gender disparities differ by gender and include (i) cost of fertilizer (plot head and plot owner); (ii) years of education and growing a single crop on the plot (all plot heads); (iii) access to subsidized inputs (plot head, plot owner and plot migrant); (iv) household tools and ethnicity (plot manager); (v) plot size (plot migrant) and (vi) age (plot owner). We also find that the endowment effect is more pronounced for the poorest and wealthiest farmers. The gender differences in the results suggest that policies should be gender specific.
