Mineral Deposits and Economic Prosperity in Africa

By Leon Cherr, Cameron Grant and Nicole Sim

How do economic policies for natural resources impact economic growth? This article uses nighttime lights data to compare different policies to trends in surrounding economic development. Africa is rich in mineral deposits, and mining is a major industry across countries. As countries have further industrialized, new policies to regulate mining for sustained benefits have been introduced, impacting economic outcomes. 

Reproducible research: The code used in this study can be replicated using the code available on GitHub at the GitHub repository

INTRODUCTION

Much of Africa’s income is from natural resources, specifically mineral deposits. The infrastructure, socioeconomic policies and resulting trends surrounding these minerals are crucial in the short and long-term development of Africa. An understanding of these policies and what constitutes positive or negative governance could make a difference in the sustainable development of infrastructure across the continent. 

We chose three countries to examine this topic: Ghana, Tanzania, and Mauritania. The largest gold mines in these countries are comparable in their size of output, making them useful for data pairing. Additionally, their similar recent histories of socioeconomic instability suggest that their pre-existing infrastructures will also be comparable, and non-linear growth trends in one country would likely be seen in the others.

In 1992, Ghana held its first democratic election in more than ten years. Following this, there was substantial investment Ghana’s economic and social infrastructures. Initiatives included Vision 20, which had the goal of “decentralizing governmental support in the country, so that all could progress” (Business & Human Rights Resource Centre, 2015). However, policies may not be as successful in practice. Figure 1 from Karimu (2024) presents the trends of debt as a percent of GDP over the period 1960–2021, suggesting that the IMF programs usually have a short-term impact and no long-term effect on debt sustainability. This raises the possibility that the path of Ghana, and perhaps other countries with similar socioeconomic situations, need to be structured differently. 

Figure 1: Debt as a share of GDP in Ghana. Source: Karimu (2024)

Tanzania has been on a path of infrastructure development for the last several decades. Like Ghana, there have been significant changes to government and systems, and the country has received large amounts of foreign aid. Education rates have improved, but doubts remain regarding actual income improvements (Nord et al., 2009). However, Tanzania’s government expenditure now more closely reflects that of more stable governments in sub-saharan Africa as shown in Figure 2, suggesting their efforts could provide guidance for other developing countries.

Figure 2: Government Expenditure in Tanzania relative to SSA (source: IMF)

Mauritania shows a similar history as shown in Figure 3: periods of economic instability and efforts to stabilize. Growth has been volatile, increasing in 2006 with the country’s entry into the oil market, before dipping in 2009 following a coup and a financial crisis. In 2015, a drop in commodity prices correlated with a drop in overall growth – a demonstration of the dangers of being reliant on one or only a few primary commodities (World Bank, 2020). 

Figure 3. Mauritania’s economic growth. Source: World Bank

Nighttime Lights and Government Policies 

We study whether differences in nighttime light intensity, between 2014 and 2024 around the largest gold mines in Tanzania, Mauritania, and Ghana reflect variations in government economic policies affecting mining operations and surrounding communities. Nighttime lights are a useful proxy for economic growth, providing consistent, uniform data at relatively high resolution and temporal resolution. To  isolate the effects of mine-related policy, we analyzed nighttime lights around mines, including the sites of mineral exploitation and the primary connected communities (ones that would most reflect the positive or negative impact of policies on mining activities, providing understanding as to the value of those policies on nearby communities). Government policies will impact the productivity of mines, which in turn will impact the resources going towards infrastructural development. This, finally, should be reflected in the visible nightlights, providing a consistent proxy by which growth caused by those policies can be measured. 

Fafchamps, Koelle & Shilpi (2016) pair USGS mine locations, population census data and nighttime lights with Difference-in-Differences (DiD) to compare towns within 20–30 km of mines to those of farther controls, pre/post mine opening or expanded operations, thus isolating localized economic impacts. The study shows that large mines can catalyze proto-urbanization & local economic development. Mamo, Bhattacharyya, and Moradi (2019) supports our hypothesis of  an enclave effect, providing empirical evidence that increases in nighttime light intensity cluster tightly around mining districts, directly reflecting mining operations and related settlement growth – mining districts experience large spikes in night-lights after major mine openings, while adjacent districts do not. It also supports the importance of policy influence, highlighting the economic effects tied to mine expansion timing and operational status, typically regulated by government policies. However, Mamo, Bhattacharyya, and Moradi executed their project at a continent wide scale, leaving out a country-by-country breakdown of different policies. Our contribution takes the hypotheses and findings from these studies to validate our assumptions. Building on their findings, we provide a comparison of mining policy implementation across the three countries, linking the timing of these policies to observed changes in nighttime light intensity around major gold mines.

To locate mineral deposits and mines across Africa, we used a package from the US Geological Survey from 2021¹. The package contains information on mineral deposits, mineral development sites, and various infrastructures, such as roads (Padilla et al., 2021). 

Figure 4. Visual of USGS survey data package layers.

To find three comparable countries to analyze, we used a source from Statista, choosing the three largest gold mines with similar annual production. Mauritania, Tanzania, and Ghana’s largest gold mines are all within 18,000 oz. of produced gold. For our second set of mines, we used three separate datasets from Global Data that ranked the five largest gold mines in each of the countries.

Figure 5. Largest gold mines in Africa. Source: Statista.

We used Annual masked averages from the Earth Observation Group, downloading masked (removing all gas fires and extreme outliers) average .tif files.² Specifically, we downloaded v21 and v22 files from 2014-2024. Our variables were NTTL Lights (dependent variable), using Mean Radiance to calculate floating point radiance in nW/cm²/sr, averaged for each mine’s respective bounding box. Unit’s represent radiance since post-2014. Our independent variable, time is over a 10 year span from 2014-2024.

Ghana overall showed increasing trends from 2014 to 2024. With the sharpest periods of increase between 2016-2018, 2019-2021, and post COVID-19. Two policies stood out during this period. First was the Minerals Development Fund Act (2016) (Parliament of Ghana, 2016), updating and enforcing strengthened provisions from the Minerals and Mining Act (2006) (Republic of Ghana, 2006) on environmental protection, community development, and local content requirements. Mining companies were mandated to increase investment in local infrastructure, including electrification and social amenities, leading to improved lighting and economic activity detectable via NTTL. The second policy was an expansion of  the National Electrification Scheme, a plan initiated in 1990 (Arthur-Mensah, 2019). The policy set in motion a major government initiative to extend electricity access to rural and mining areas, resulting in more households, businesses, and public infrastructure gaining access to electricity, directly increasing nighttime illumination.

Figure 6. Ghana’s first and second largest mine grew from 2014-2024 and policies.

Tanzania had three primary events, showing an increasing trend from 2014 to 2024 with the sharpest periods of increase from 2020-2024 for Geita gold mine, and 2014-2016 for North Mara. First was the Mining Act of 2010, introducing new regulations such as including government participation and local content requirements (United Republic of Tanzania, 2010). North Mara was under greater scrutiny by the government, due to social issues like abuse cases and deaths (Business & Human Rights Resource Centre, 2013), leading to tighter enforcement of the 2010 Mining Act. Geita was likely more stable and cooperative and thus had a more gradual increase. Second, the updated Mining Act amendments effective 2019 placed strong emphasis on local content compliance, requiring greater Tanzanian participation in mining operations, for suppliers, contractors, and employees, leading to a sharp increase in NTTL around Geita (United Republic of Tanzania, 2010). Finally, an export ban in 2017 was placed over Acacia Mining Plc (North Mara’s operator) while the company was embroiled in governmental lawsuits and tax disputes, leading to a distinct (and temporary) decrease in NTTL on the mine and surrounding region (Business & Human Rights Resource Centre, 2013).

Figure 7. Tanzania’s first and second largest mine growth from 2014-2024 with policies.

Mauritania’s two largest mines show a shaky, but increasing trend from 2014-2024. The noisy data reflects frequently changing regulations and weaker institutional capacity, unlike Ghana or Tanzania where mining governance and regulatory frameworks are somewhat more stable and policy implementation is comparatively consistent. The primary policy seen to affect NTTL lights was the Mining Policy Framework Renewal (2016) (Intergovernmental Forum on Mining, Minerals, Metals and Sustainable Development, 2019), increasing organization that allowed for concentration of economic activity around official mining sites. The government formalized and regulated previously chaotic artisanal mining by introducing licensing requirements (mandating that license holders be Mauritanian citizens, and requiring gold sales through designated state bodies). This led to significant investments, and, visualized by a decrease in NTTL as activity slowed for it, expansions. Guelb Moghrein’s growth after 2017 appears to be interrupted, before rising again in 2018, likely a result of a development project (First Quantum Minerals Ltd., 2025). 

Figure 8. Mauritania’s first and second largest mines grew from 2014 to 2024, with policies

We then combined the results into a single chart to compare the growth of all six mines, shown in Figure 13. Keep in mind that there are two different scales shown on the Y-axis – the left side shows values for solid lines, while the right shows values for dotted lines. 

Figure 9. Growth of Tanzania, Ghana, and Mauritania’s first and second largest gold mines.

CONCLUSION

To summarize our results, we find a correlation between government policy implementation and changes in nighttime lights intensity around major gold mines in Ghana, Tanzania, and Mauritania from 2014 to 2024. We also find that governance capacity and stability play a critical role in a country’s ability to enforce mining policies consistently across all mines. For example, Ghana, ranked 7th in overall governance in Africa, has clear, stable, and well-enforced mining policies, which correspond with steady and strong increases in NTTL. In contrast, Mauritania, ranked 41st in governance, experiences irregular and unpredictable NTTL levels that fluctuate. Frequent changes in mining laws and weaker enforcement have caused uneven local development. In Mauritania, mining companies often exercise greater operational control, sometimes overshadowing governmental policies, leading to volatile outcomes. Looking at Tanzania (ranked 15th), the NTTL increases were less consistent overall. Unlike the Geita mine, the North Mara mine experienced slower and more uneven growth due to regulatory conflicts, export bans, and lawsuits. This highlights the challenges of balancing policy enforcement with complex social issues, resulting in variable economic impacts across mines within the same country.

Our study faced numerous limitations, the first being attribution challenges. It is hard (if not impossible) to perfectly attribute economic growth in surrounding communities to the mine: there could be external factors, such as foreign aid, influencing NTTL levels. Local data was also limited – some mines house workers, meaning there aren’t nearby communities to look to for NTTL-based-estimates on how much money is going to the region. Next, our study assumes direct and isolated impact of a single national policy on NTL changes at mines. However, this assumption does not fully capture the complexity of policy implementation (e.g. simultaneous effect of multiple policies or lags in impact) nor take into consideration local or social context, such as company/firm decisions. There is also a lack of quantitative econometric analysis, meaning one is unable to establish causality or quantify strength of relationship between NTTL intensity and government policies. Finally, reliance on secondary and online sources could introduce errors to datasets and our understanding of the various aspects (i.e. policy or development). 

Further steps could be taken in the future to improve the quality of results. First, one could dive deeper into local context and challenges. A more rigorous and focused literature review into case studies would provide nuanced insights into specific policies and firm strategies (e.g. corruption, outflow of economic benefits to foreign companies, etc.). Next, further quantitative analysis would improve our understanding of different elements at play – for example, comparing NTTL data with other indicators for economic growth to determine relationships (e.g. amount of gold extracted, GDP of settlements, income per capita of settlements, etc). Potential Difference-in-Differences (DiD) designs would explore “controlled” areas with mines that have known mineral deposits that are unextracted. Finally, one could Explore Simple econometric models (e.g. OLS). 

References

1. Fafchamps, M. (2003). Rural poverty, risk and development. Stanford University. https://web.stanford.edu/~fafchamp/proto_all.pdf
2. Henderson, J. V., Storeygard, A., & Weil, D. N. (2012). Measuring economic growth from outer space. American Economic Review, 102(2), 994–1028. https://www.sciencedirect.com/science/article/abs/pii/S0304387818303936
3. Karimu, S. (2024). Ghana and the IMF: Policy shifts, economic bailouts, and macroeconomic outcomes. Journal of Policy Modeling, 46(6), 1146–1164. https://doi.org/10.1016/j.jpolmod.2024.07.006
4. Padilla, A. J., Otarod, D., Deloach-Overton, S. W., Kemna, R. F., Freeman, P. A., Wolfe, E. R., Bird, L. R., Gulley, A. L., Trippi, M. H., Dicken, C. L., Hammarstrom, J. M., & Brioche, A. S. (2021). Compilation of geospatial data (GIS) for the mineral industries and related infrastructure of Africa [Data release]. U.S. Geological Survey. https://doi.org/10.5066/P97EQWXP

Footnotes:

1. https://www.sciencebase.gov/catalog/item/607611a9d34e018b3201cbbf ↩︎
2. https://eogdata.mines.edu/nighttime_light/annual/v21/2014/VNL_v21_npp_2014_global_vcmslcfg_c202205302300.average_masked.dat.tif.gz ↩︎