by Degol Hailu and Admasu Shiferaw

High economic growth rates were observed in mineral and oil exporting African countries since the turn of the 21^st century.  However, it is difficult to be sanguine about their sustained growth potentials. This skepticism is largely based on the unfavorable performance of the manufacturing sector.

Figure 1 shows the trends in manufacturing comparing resource-rich countries with resource-poor ones. While the intensity of resource extraction varies substantially over the years in resource-rich countries, it has been rising on average with sharp increases since 2005. In the meantime, the share of manufacturing in these countries remained relatively static at about 12% of GDP up until the early 2000s and started to decline thereafter. It is clear that the rising resource rents have not been used effectively to structurally transform resource-rich economies. There seems to be an important Dutch-Disease type mechanism that works through weak manufacturing activities. In contrast, the share of manufacturing in resource-poor countries has been rising from about 5% of GDP in the mid-1970s to about 15% in recent years.

                     Figure 1: Resource rents and manufacturing in Sub-Saharan Africa[1]

To capture the above undesirable outcome in a more reliable fashion, we run a panel data regression model for 47 countries with varying degrees of resource rents over the period 1970-2015. The estimation results are broadly consistent with the patterns observed in Figure 1. There is a negative and statistically significant association between manufacturing value added relative to GDP (MV/GDP), and the intensity of resource extraction. This is after controlling for differences across countries in per capita income and other unobserved country characteristics and common trends that may influence the performance of manufacturing. Accordingly, a country that experiences a ten-percentage point increase in resource rents to GDP ratio (RR), say due to a resource discovery, would experience a 1.2 percentage point reduction in the GDP share of manufacturing on average.[2]

It is also important to understand the extent to which reliance on extractive activities are persistent at the country level. To this effect, Figure 2 maps resource rents during 2010-16 among resource-rich countries against their resource rents during the 1980s. Countries in which the intensity of resource extraction during 2010-16 remained at the same level as in the 1980s will be located on the solid red line, while countries with rising intensity of resource extraction relative to the 1980s will be located above the solid line. The positive correlation observed in the scatter plot suggest strong persistence in resource-dependence in our sample while the fact that most countries are located above the equality line underscores that resource-dependence has actually been rising since the 1980s.

Figure 2: Persistence in Resource Dependence

In terms of policy, our findings suggest that there have been missed opportunities, and that resource rents on their own may not lead to economic transformation.

However, it shouldn’t be all gloom and doom. Natural resources can support growth in the manufacturing sector if backed by deliberate interventions. First, by securing a fair share of resource rents, governments can use fiscal policy to finance carefully designed industrial policy, develop the requisite human capital and acquire appropriate technology that are critical to promoting manufacturing. This is the experience of Chile, Indonesia and Malaysia. Second, local content, beneficiation and valued addition activities can motivate investment in manufacturing activities. South Africa and Brazil are examples of this. Third, infrastructure (roads, rails, utilities, communication etc.) built to support resource extraction can also facilitate export-oriented manufacturing and agro-processing. Fourth, private sector incomes earned through employment and profits from extractive activities can generate the impetus for industrialization through higher domestic effective demand.

Policy makers in resource-rich countries can exploit the above pathways for manufacturing development and reverse the de-industrialisation trend to meet the SDG target 9.2—achieving inclusive and sustainable industrialization.

The longer paper, on which this blog is based, is forthcoming.

[1] We distinguish resource-rich African countries from their resource-poor counterparts using a 3% GDP share of resource rents as a cutoff point.

[2] The estimated equation is (MV/GDP)it = g + βRRit + αln(GDP-PC)it + d + t + eit. The subscripts i and t index country and year. d is a time invariant country specific effect. t is a time fixed effect. e is an equation error term. The results are g=5.18 (10.43); β=-0.12 (0.04); α=0.85 (1.46); with R2=0.12 and 1,434 number of observations. The numbers in parenthesis are standard errors. The coefficients are statistically significant at the 1% level.