35 pages, Mozambique remains predominantly poor. The official statistics show that poverty incidence barely changed from 54% in 2002–03 to 55% in 2008–09, which stands way above the government's target of 45% by the year 2009. This places the country off-target to cut hunger and poverty by half by 2015, despite an annual economic growth of about 7% in the period 1994–2010. In rural areas, poverty levels have slightly increased, due to the underperformance of the agricultural sector. Extension services can have a significant impact on poverty reduction through stimulating growth in agricultural productivity. Based on a nationally representative household survey from Mozambique, this paper uses three econometric models, namely an OLS regression, the doubly robust estimator and matching and regression to estimate the economic impact of receipt of extension. The results suggest that the receipt of extension increases farm incomes by 12%. However, rather than crafting resource-poor technologies, extension services tend to target wealthier households who are relatively more likely to adopt the existing technologies. This might increase income inequality. The impact of extension, and therefore its contribution to poverty reduction, can be enhanced through several mechanisms (e.g., programme design and the number of staff).
22 pages, In this paper, we investigate the link between windfall gains and losses of income associated with commodity exports and economic performance in a panel of 45 sub-Saharan African (SSA) countries over the period from 1990 to 2019. Windfall gains and losses of income are measured in terms of fluctuations in a country-specific commodity terms of trade (CTOT) index in which each commodity is weighted by the ratio of exports of that commodity in the country’s gross domestic product (GDP). The CTOT index therefore reflects the commodity export specialisation for individual countries. The data on CTOT are taken from the International Monetary Fund. Additionally, we use changes in real GDP per capita as our SSA economic performance measure. We employ a random coefficient model that yields individual estimates for each of the countries included in the analysis. Our approach is based on the assumption that the effect of windfall gains and losses on real GDP per capita growth varies across different SSA countries. Our main conclusion can be elaborated as follows: first, natural resources have undoubtedly contributed to higher economic growth in SSA countries since 1990. Second, when SSA countries are analytically divided into two groups depending on their commodity export specialisation, we find that resource-rich countries—in particular oil rich—are the best economic growth performers during the observation period. Finally, we find that windfall gains from commodity exports are not significantly associated with increased real GDP per capita growth in most agriculture-exporting countries.