From Transparency to Transformation: NRGI’s NOC Work Past, Present and Future
Over the past two decades, national oil companies (NOCs) have moved from the margins of governance debates to the center of discussions to fix how countries manage resource wealth and navigate the energy transition.
NOCs produce more than half of the world’s oil and gas, and in many producing countries their decisions have major implications for public finances, energy policy and long-term development outcomes. That role is only growing in importance in an increasingly uncertain energy future. How NOCs can adapt to that uncertain future remains unresolved, and shocks such as today’s soaring and volatile prices serve as a potent reminder of the need to look ahead and plan for the future.
NRGI’s work on NOCs has evolved alongside these shifts, supporting more informed and inclusive decision-making about natural resources and the energy transition. As we mark our 20th anniversary in 2026, we are building on this experience with new insights on the future of NOCs.
Our new report National Oil Company Transformation: Strategic Choices for an Uncertain Energy Future is part of that effort. As we prepare to launch this work, we are also taking a moment to reflect on how our thinking around NOCs has evolved through four key phases, and what this means for the choices that countries face now and hence for the future of NRGI’s NOC work.
1. Building the foundations: from transparency to governance and country-driven change (2006-2017)
NRGI’s early work focused on extractive revenue transparency, helping citizens better understand how extractive revenues flowed into public coffers. NOCs were not initially central to this agenda, but concerns about country-level corruption risks and NOC opacity gradually brought them to the fore, particularly around their role in financial flows.
Sustained work on NOC oil sales transparency followed and led to tangible impacts. In Nigeria, work by NRGI and partners NEITI led to NNPC canceling criticized contracts. NRGI also produced guidance on state-owned enterprise (SOE) disclosures, supporting global norm strengthening and supporting efforts by organizations such as México Evalúa to assess NOC transparency.
Over time, we expanded beyond transparency to other core NOC governance questions, asking not just what is disclosed, but how institutions are run. This included delving into NOC reforms such as limiting political interference while ensuring sufficient oversight.
In the early 2010s, we also began engaging more deeply on NOCs in emerging “oil hotspots” such as Ghana, including partnering with Chatham House and the Commonwealth Secretariat to establish the New Petroleum Producers Group. Key to our approach was tailoring methods to different settings, balancing cross-country learning with the reality that not all countries are Norway.
Across these early engagements—from UNOC in Uganda to Pertamina in Indonesia—we started to combine cross-cutting NOC research with country-level engagement, and saw this as useful in achieving impact, including in difficult contexts such as Myanmar.
- Effective NOC governance is critical and multi-faceted. NOC transparency is necessary but not sufficient. It must be paired with other governance improvements, including stronger institutions and oversight.
- Real change is largely country driven. Global tools are most impactful when shaped by and applied through country programming, drawing on deep country knowledge and partnerships.
2. Leveraging a data-driven approach for impact (2017-2020)
The second phase focused on data to drive reform, inform decisions, shape incentives and encourage better performance of NOCs.
The Resource Governance Index (RGI) benchmarked more than 50 NOCs and showed widespread governance gaps. This data-driven approach helped lead to tangible changes.
In Qatar, engagement with Qatar Petroleum resulted in the company publishing annual and sustainability plans on its website for the first time. In Mexico, NRGI inputs to the National Hydrocarbons Commission (CNH), whose purview included regulation of Pemex, led to CNH pushing Pemex to improve transparency.
A major milestone in this phase was the launch of NRGI’s National Oil Company Database, the world’s largest open dataset on NOCs. It provides quantitative insights on fiscal transfer patterns and debt-sustainability risks, helping governments, private-sector analysts, journalists and international financial institutions better understand the profound public finance implications of NOC decisions.
NRGI also used this data-driven approach to support country-level analysis of opportunities and risks surrounding NOC oil sales, including in Ghana, and resource-backed loans, including in Nigeria.
Across these efforts, NRGI used data as a lever: to clarify fiscal stakes, justify accountability asks, and strengthen the connection between evidence and reform.
3. NOCs and the challenges of the energy transition (2021-2025)
In this third phase, the energy transition became central to our NOC work, which in turn helped shape broader understanding of the role NOCs play in shaping the global energy transition. In addition to controlling more than half of global oil and gas production, NOCs control around 40 percent of total sector investment and close to 60 percent of reserves.
NOCs play a critical role in how producing countries respond to declining long-term demand for fossil fuels. Their investment strategies can either reinforce continued dependence on oil extraction or support a more diversified energy future.
A major NRGI focus during this phase was on the “risky bets” of NOCs in a context of weakening global hydrocarbon demand. Importantly, NRGI framed these risks primarily as a public investment and fiscal management challenge. We highlighted the impact of NOC investment choices on national cash flows, debt sustainability and long-term fiscal space. This framing proved influential in Ghana where supported partners including the Alliance of Civil Society Organizations (CSOs) Working on Extractives, Anti-corruption and Good Governance to help prevent a risky prospective investment by GNPC.
Recognizing the value of peer benchmarking among NOCs, we assessed how 21 NOCs acknowledge, assess and mitigate transition risks. We complemented this comparative approach with analysis anchored in country challenges, linking transition to fiscal and institutional questions such as Pemex’s debt challenges in Mexico and Ecopetrol’s relationship with Colombia’s government.
During this phase our NOC work also expanded into climate-related areas, including the crucial role of NOCs in methane emissions in Mexico, Nigeria and the MENA region. We contributed to methane abatement finance guidance targeted at NOCs and worked to improve incorporation of NOCs in national climate commitments, including in Mexico. We also worked with partners such as IISD on more effective climate community engagement with NOCs, including at various COPs.
Across countries such as Colombia, Ghana, Mexico, Nigeria, Senegal and Uganda, our work responded to a growing demand among various stakeholders to deepen informed national debate on how NOCs navigate the energy transition.
4. The next phase: NOC transformation
NRGI’s NOC work is now entering an exciting next phase.
With the launch of our new report, we move beyond assessing risk to examining how governments, NOC leaders and other stakeholders can define credible NOC transformation pathways that manage transition risks while making the most of emerging opportunities.
The report sets out a framework to help countries:
- Define national-level policy goals,
- Assess capacities and constraints, including the NOC’s ability to pursue new business opportunities,
- Align mandates, governance structures and financial resources with the chosen transformation pathway, and
- pursue a just transition, given the important role NOCs play in shaping outcomes for workers and communities in oil producing regions.
This framework also reflects a key lesson from our in-country experience: there is no single model. What is feasible, and at what pace, will differ across countries and NOCs. Transformation pathways must reflect fiscal realities, political dynamics and institutional capacity.
Looking ahead
NRGI’s NOC work will continue to evolve to meet new challenges.
We will contribute to discussions on transition roadmaps, where the role of NOCs will be decisive, and to accountability on methane commitments, where NOCs lag significantly behind.
Our NOC transformation work will expand to better include energy security, the financing of transformation and links to national “green industrial policy” plans. We aim to update the NOC Database to integrate transition indicators, strengthening cross-country benchmarking and national advocacy.
This also means engaging more with narratives shaping NOC policies and strengthening strategic communications to better support national debate on NOC transformations.
As highlighted during the launch of NRGI’s 20th anniversary earlier this month, the lessons of the past provide a foundation to “confront the challenges that persist, and the new ones that emerge.” This is particularly true for NOCs. As NOCs take on a larger role in an uncertain energy future, the stakes are rising. The challenge is no longer just improving NOC governance, but defining the role of these companies in addressing the most pressing resource governance challenges—from navigating energy transition risks and opportunities to advancing both development and decarbonization.
NRGI looks forward to continuing this dynamic NOC work with partners to respond to these challenges.
Impact stories
For more on the real-world impacts of NRGI’s NOC work over the years, read the following impact stories:
Notes
-
1
With a gold price of USD 1,600 per ounce.
-
2
With a low-profit mine and a gold price of $1,600 per ounce.
-
3
Cecilia Jamasmie, “Petra Diamonds’ stake in Williamson to shrink as part of deal with Tanzania,” Mining.com, 13 December 2021, www.mining.com/petra-diamonds-stake-in-williamson-to-shrink-as-part-of-deal-with-tanzania.
-
4
Lifezone Metals, “Kabanga Nickel Signs Framework Agreement,” 19 January 2021, www.lifezonemetals.com/kabanga-nickel-signs-framework-agreement.
-
5
Thomas Scurfield and Silas Olan’g, “Magufuli Seeks the Right Balance for Tanzania’s Mining Fiscal Regime,” NRGI, 31 January 2019, www.resourcegovernance.org/blog/magufuli-seeks-right-balance-tanzania-mining-fiscal; Thomas Scurfield, “Tanzania Strikes a Better Balance with its Mining Fiscal Regime,” NRGI, 24 June 2020, www.resourcegovernance.org/blog/tanzania-strikes-better-balance-mining-fiscal-regime.
-
6
This framework agreement was published in a document setting out Barrick’s offer to buy the shares it did not already own in Acacia Mining, the previous owner of the Bulyanhulu, Buzwagi and North Mara mines in Tanzania. See Acacia Mining and Barrick Gold, Recommended Final Offer for Acacia Mining Plc by Barrick Gold Corporation, 2019, 66–79, s25.q4cdn.com/322814910/files/doc_downloads/acacia/Acacia-2.7-announcement.pdf.
-
7
The main revenue streams are import duty, skills development levy, royalty, corporate income tax, a share of dividends and shareholder loan repayments through state equity, and dividend withholding tax.
-
8
The earlier in time a shilling (Tanzania’s official currency unit) is received, the more it is worth. This is, first, because it can be used earlier; and second, because the future is uncertain, and no one can be sure they will receive that shilling in the future. To account for this time value of money, a “discount rate” is applied. In the sharing arrangement, this would mean that if the government received a shilling in year 1, the company would need to receive more than a shilling in year 2 for the benefits to be comparable. However, given cumulation here is based on actual cash flow, the company would need to receive only a shilling in year 2 for the benefits to be shared equally.
-
9
This provision for the company to earn its minimum return before sharing is triggered means Ecuador’s mechanism is similar to an R-based cash flow tax, commonly referred to as a Brown Tax. See, e.g., Robin Broadway and Michael Keen, “Theoretical perspectives on resource tax design,” in The Taxation of Petroleum and Minerals: Principles, Problems and Practice, edited by Philip Daniel, Michael Keen and Charles McPherson (Oxford: Routledge, 2010), 13–74.
-
10
Prices are taken from World Bank, “Commodities Price Data (The Pink Sheet),” www.worldbank.org/en/research/commodity-markets.
-
11
With a gold price of $1,600 per ounce.
-
12
With a gold price of $1,600 per ounce.
-
13
With a gold price of $1,600 per ounce.
-
14
The Fraser Institute survey estimates that, unless there are extremely harmful policies, around 60 percent of an investment decision tends to be based on a country’s geology. The other 40 percent comprises of several other factors, including political stability and policy predictability (given they affect the risk that investors will not be able to secure future returns generated by their investments), a conducive business environment and the tax level. See Julio Mejia and Elmira Aliakbari, Fraser Institute Annual Survey of Mining Companies 2022, (Fraser Institute, 2023), 8, www.fraserinstitute.org/studies/annual-survey-of-mining-companies-2022.
-
15
However, information gaps make it difficult for taxes to be structured to capture all excess profit. See Jean-Franҫois Wen, Progressive Taxation of Extractive Resources as Second-Best Optimal Policy (International Monetary Fund, 2018), www.imf.org/en/Publications/WP/Issues/2018/06/13/Progressive-Taxation-of-Extractive-Resources-as-Second-Best-Optimal-Policy-45923.
-
16
Recent research provides a sense of the potential revenue loss to governments from tax avoidance. The International Monetary Fund recently estimated that sub-Saharan African mining countries could be losing between $450 and $730 million in corporate income tax a year. See Sebastian Beer and Dan Devlin, Is There Money on the Table? Evidence on the Magnitude of Profit Shifting in the Extractive Industries (International Monetary Fund, 2021), www.imf.org/en/Publications/WP/Issues/2021/01/15/Is-There-Money-on-the-Table-Evidence-on-the-Magnitude-of-Profit-Shifting-in-the-Extractive-49983.
-
17
It is perhaps surprising that Tanzania’s 50-50 sharing arrangement generates an AETR greater than 50 percent (with a discount rate of 10 percent). This is despite AETR measuring government take as the share of pre-tax profits, which is larger than “economic benefits” (given economic benefits exclude interest payments). This outcome results from the 50-50 split being based on actual cash flow. The government receives revenue before the mining company through input and production taxes that do not depend on the mine making a profit. Because of these earlier revenues, the government receives a larger share on a discounted basis.
-
18
As reported in the S&P Global database. Legal risks are “expropriation, state contract alteration and contract enforcement risks.” Tax risks are “tax increase and tax inconsistency risks.” Control Risks scores these risks as still “very high” and “high” respectively (following Tanzania’s overhaul of extractives sector laws and other actions against existing mines in 2017) but reducing.
-
19
With a discount rate of 10 percent. While Ecuador’s sharing mechanism does not account for the labor profit share because none of it will go to the government from 2024 onwards, I have included it in the AETR because it is a tax on the project. The Democratic Republic of Congo regime has an excess profits tax that is triggered for a mine when the realized price is at least 25 percent higher than the price in its feasibility study. I assumed that the feasibility study has a price of $1,300 per ounce, so the excess profits tax is not triggered.
-
20
Total benefits in this case are a project’s revenues minus operating costs and replacement capital (but not minus exploration and development capital). This cash flow represents the money available to pay back the initial investment and provide a return. The government share of it is a common measure of progressivity.
-
21
Wen, Progressive Taxation of Extractive Resources as Second-Best Optimal Policy.
-
22
With a discount rate of 10 percent. The results for only some countries are shown to clearly depict each data point. The results for all the evaluated countries can be found in my model.
-
23
Scurfield, “Tanzania Strikes a Better Balance with its Mining Fiscal Regime.”
-
24
This feature is not fully reflected in Figure 5 given that “total benefits” use a slightly different definition of profits and are based on discounted cash flows.
-
25
With a gold price of $1,600 per ounce.
-
26
For example, an average 62 percent of respondents to the Fraser Institute surveys between 2017 and 2019 said the current implementation of Tanzania’s legal system would strongly discourage investment, and 73 percent said regulatory uncertainty would. See, e.g., Ashley Stedman, Jairo Yunis and Elmira Aliakbari, Fraser Institute Annual Survey of Mining Companies 2019 (Fraser Institute, 2020), www.fraserinstitute.org/studies/annual-survey-of-mining-companies-2019.
-
27
With a low-profit mine and a gold price of $1,600 per ounce.
-
28
Tax avoidance could extend the Philippines’ recovery period, and therefore delay the payment of some taxes including import duty and interest withholding tax, given the end of the recovery period depends on the reported profitability of a mine rather than an ex-ante assessment. However, the rule that the recovery period must end five years from the start of production regardless of whether pre-production expenses have been recouped limits the extent to which it can be extended.
-
29
The merits of these measures require further scrutiny. E.g., taking a share of loan repayments could result in lenders charging a higher interest rate to ensure they still recoup their loan and a minimum return. This would not only reduce taxable income but also make it harder for the government to assess whether an interest rate is reasonable, because it would not be comparable with industry benchmarks. It can also be difficult for a government to always determine whether a loan is from a related party or not. However, these considerations are outside the scope of this analysis.
-
30
Natural Resource Governance Institute (NRGI), Natural Resource Charter, 2nd edition, 2014, resourcegovernance.org/analysis-tools/publications/natural-resource-charter-2nd-ed.
-
31
Although its exclusion of several significant taxes from the government’s share of benefits means low-profit mines may still be impacted.
-
32
With a gold price of $1,600 per ounce.
-
33
Anna Fleming, Thomas Lassourd and David Manley, “Variable Royalties: an Answer to Volatile Mineral Prices?” in Handbook on the Future of Resource Taxation, African Tax Administration Forum and Intergovernmental Forum on Mining, Minerals, Metals and Sustainable Development (forthcoming), www.iisd.org/publications/brief/future-resource-taxation-roadmap.
-
34
Ensuring that interest rates used as a comparison apply to comparable assets with a similar risk profile is challenging, but rules of this nature have been successful in reducing profit shifting in other countries. See Beer and Devlin, Is There Money on the Table?.
-
35
The main taxes listed are VAT, royalty and corporate income tax. The regime also includes a share of pre-tax profits that is currently divided between the company’s workers and the government, with the portion received by the government included in its accumulated benefits. However, a recent court ruling means that all this labor profit share will go to workers from the start of 2024 and therefore none of it will be included in government benefits.
-
36
The discount rate used is specific to each mine and based on its weighted average cost of capital (WACC). I have assumed that WACC is around 7 percent in real terms. This is based on the typical discount rate for equity shareholders used by industry and government analysts of 8 percent in real terms, and the current average cost of debt for the mining sector as reported by Aswath Damodaran, Damodaran Online, www.pages.stern.nyu.edu/~adamodar.
-
37
Republic of the Philippines, Financial or Technical Assistance Agreement, mgb.gov.ph/attachments/article/79/PFC_FTAA.pdf.
-
38
For example, some terms in the original FTAA for an OceanaGold mine differed in some areas: Republic of the Philippines, Financial or Technical Assistance Agreement with Arimco Mining Corporation, 1994, www.resourcecontracts.org/contract/ocds-591adf-2792396017. I understand that the recently signed extension to this agreement has slightly different terms again.
-
39
A template of the FARI model and a user guide that explains all the concepts and workings of the model are available at International Monetary Fund, “Fiscal Analysis of Resource Industries,”www.imf.org/external/np/fad/fari.
Tax Avoidance, Tax Evasion and Trade Misinvoicing: Risks to Senegal’s Mining Sector
Key messages
- Tax evasion, trade misinvoicing and multinational companies’ efforts to avoid tax are important risks in the extractive sector, especially in developing countries, including Senegal.
- Senegal’s extractive governance arrangements are strong in many areas, including in preventing multinational tax avoidance.
- However, there is cause for concern in two areas: misinvoicing of international trade (imports and exports), and cost inflation by subcontractors in the extractive industries. Quantitative estimates of trade misinvoicing suggest the former is a significant problem for Senegal (including for its mining sector), which could lead to the loss of 1–3 percent of the country’s tax revenues.
- Authorities in Senegal should address such risks, especially before the country begins large-scale oil and gas production. Unless they do so, revenues from the country’s hydrocarbon resources will be at risk.
- Measures that Senegal can take include strengthening the tax authorities—especially customs and auditing accounts of extractive industry suppliers.
Senegal’s extractive industries could provide 8–9 percent of state revenues until 2030. But efforts by companies or individuals to dodge their tax responsibilities could reduce the contribution the sector makes. Evidence suggests, for example, that multinational corporations shift around 35 percent of their profits to tax havens. This problem affects the extractive industries and is particularly acute in developing countries such as Senegal.
Flows of funds linked to corruption, illegal mining, money laundering and avoiding regulations can also undermine resource-rich countries’ ability to benefit from their extractive industries. Senegal appears to be significantly affected by these problems. Various estimates have suggested that the country could be losing 1-12 percent of its tax revenues each year, due to various types of the illicit financial flows mentioned above, with a significant share linked to the mining sector. The wide range of estimates underscores the need for better data.
This report includes estimates of Senegal’s tax revenue losses due to trade misinvoicing (false customs declarations): losses of USD57 million to $153 million (1–3 percent of tax revenues) depending on the method used, with a substantial share linked to the mining sector. If reclaimed, these amounts could be large enough to double the size of Senegal’s cash transfer support program for households.
How do such losses occur? Though Senegal’s tax system is strong in many ways and officials have high capacity for overseeing the taxes of mining companies, there are still weaknesses. For instance, Senegal could strengthen coordination and information sharing between the tax department and the Ministry of Mines. The country may also need to acquire more capacity to audit oil and gas companies for tax purposes, including through hiring geologists. In addition, Senegal may be losing tax revenues due to trade misinvoicing.
Gold exports to the United Arab Emirates appear to be particularly affected, although by themselves they would account for a minority of the tax revenue losses that Senegal experiences. According to the country’s national statistical agency, another major source of tax losses in the extractive industries is the under-declaration of the value of minerals, where these are processed in Senegal by the mining company (or a part of the same group). A further potential area of vulnerability is subcontractors in the extractive industries. When subcontractors based abroad supply extractive companies in Senegal from the same multinational group, the subcontractors can be used to inflate costs. Such subcontractors capture a substantial share of gross revenues from the extractive sector and can benefit from tax treaties that allow them to avoid taxes.
The Senegalese authorities should also remain vigilant to the risks of tax treaty abuse, given that several multinational companies operating in the extractive industries benefit from tax treaties that Senegal has signed with other countries (either directly, or indirectly via bilateral investment treaties). This report identifies a number of policies to address these problems. For example, it recommends that Senegal’s government and parliamentarians strengthen the country’s tax authority (Direction Générale des Impôts et des Domaines) with more personnel, and remain vigilant to the risks posed by tax treaties and bilateral investment treaties. To the Direction Générale des Impôts et des Domaines, the report recommends that officials improve co-ordination between the tax authority and the Ministry of Mines, and information sharing with customs authorities of Senegal’s trading partners, to help prevent trade misinvoicing.
Notes
-
1
With a gold price of USD 1,600 per ounce.
-
2
With a low-profit mine and a gold price of $1,600 per ounce.
-
3
Cecilia Jamasmie, “Petra Diamonds’ stake in Williamson to shrink as part of deal with Tanzania,” Mining.com, 13 December 2021, www.mining.com/petra-diamonds-stake-in-williamson-to-shrink-as-part-of-deal-with-tanzania.
-
4
Lifezone Metals, “Kabanga Nickel Signs Framework Agreement,” 19 January 2021, www.lifezonemetals.com/kabanga-nickel-signs-framework-agreement.
-
5
Thomas Scurfield and Silas Olan’g, “Magufuli Seeks the Right Balance for Tanzania’s Mining Fiscal Regime,” NRGI, 31 January 2019, www.resourcegovernance.org/blog/magufuli-seeks-right-balance-tanzania-mining-fiscal; Thomas Scurfield, “Tanzania Strikes a Better Balance with its Mining Fiscal Regime,” NRGI, 24 June 2020, www.resourcegovernance.org/blog/tanzania-strikes-better-balance-mining-fiscal-regime.
-
6
This framework agreement was published in a document setting out Barrick’s offer to buy the shares it did not already own in Acacia Mining, the previous owner of the Bulyanhulu, Buzwagi and North Mara mines in Tanzania. See Acacia Mining and Barrick Gold, Recommended Final Offer for Acacia Mining Plc by Barrick Gold Corporation, 2019, 66–79, s25.q4cdn.com/322814910/files/doc_downloads/acacia/Acacia-2.7-announcement.pdf.
-
7
The main revenue streams are import duty, skills development levy, royalty, corporate income tax, a share of dividends and shareholder loan repayments through state equity, and dividend withholding tax.
-
8
The earlier in time a shilling (Tanzania’s official currency unit) is received, the more it is worth. This is, first, because it can be used earlier; and second, because the future is uncertain, and no one can be sure they will receive that shilling in the future. To account for this time value of money, a “discount rate” is applied. In the sharing arrangement, this would mean that if the government received a shilling in year 1, the company would need to receive more than a shilling in year 2 for the benefits to be comparable. However, given cumulation here is based on actual cash flow, the company would need to receive only a shilling in year 2 for the benefits to be shared equally.
-
9
This provision for the company to earn its minimum return before sharing is triggered means Ecuador’s mechanism is similar to an R-based cash flow tax, commonly referred to as a Brown Tax. See, e.g., Robin Broadway and Michael Keen, “Theoretical perspectives on resource tax design,” in The Taxation of Petroleum and Minerals: Principles, Problems and Practice, edited by Philip Daniel, Michael Keen and Charles McPherson (Oxford: Routledge, 2010), 13–74.
-
10
Prices are taken from World Bank, “Commodities Price Data (The Pink Sheet),” www.worldbank.org/en/research/commodity-markets.
-
11
With a gold price of $1,600 per ounce.
-
12
With a gold price of $1,600 per ounce.
-
13
With a gold price of $1,600 per ounce.
-
14
The Fraser Institute survey estimates that, unless there are extremely harmful policies, around 60 percent of an investment decision tends to be based on a country’s geology. The other 40 percent comprises of several other factors, including political stability and policy predictability (given they affect the risk that investors will not be able to secure future returns generated by their investments), a conducive business environment and the tax level. See Julio Mejia and Elmira Aliakbari, Fraser Institute Annual Survey of Mining Companies 2022, (Fraser Institute, 2023), 8, www.fraserinstitute.org/studies/annual-survey-of-mining-companies-2022.
-
15
However, information gaps make it difficult for taxes to be structured to capture all excess profit. See Jean-Franҫois Wen, Progressive Taxation of Extractive Resources as Second-Best Optimal Policy (International Monetary Fund, 2018), www.imf.org/en/Publications/WP/Issues/2018/06/13/Progressive-Taxation-of-Extractive-Resources-as-Second-Best-Optimal-Policy-45923.
-
16
Recent research provides a sense of the potential revenue loss to governments from tax avoidance. The International Monetary Fund recently estimated that sub-Saharan African mining countries could be losing between $450 and $730 million in corporate income tax a year. See Sebastian Beer and Dan Devlin, Is There Money on the Table? Evidence on the Magnitude of Profit Shifting in the Extractive Industries (International Monetary Fund, 2021), www.imf.org/en/Publications/WP/Issues/2021/01/15/Is-There-Money-on-the-Table-Evidence-on-the-Magnitude-of-Profit-Shifting-in-the-Extractive-49983.
-
17
It is perhaps surprising that Tanzania’s 50-50 sharing arrangement generates an AETR greater than 50 percent (with a discount rate of 10 percent). This is despite AETR measuring government take as the share of pre-tax profits, which is larger than “economic benefits” (given economic benefits exclude interest payments). This outcome results from the 50-50 split being based on actual cash flow. The government receives revenue before the mining company through input and production taxes that do not depend on the mine making a profit. Because of these earlier revenues, the government receives a larger share on a discounted basis.
-
18
As reported in the S&P Global database. Legal risks are “expropriation, state contract alteration and contract enforcement risks.” Tax risks are “tax increase and tax inconsistency risks.” Control Risks scores these risks as still “very high” and “high” respectively (following Tanzania’s overhaul of extractives sector laws and other actions against existing mines in 2017) but reducing.
-
19
With a discount rate of 10 percent. While Ecuador’s sharing mechanism does not account for the labor profit share because none of it will go to the government from 2024 onwards, I have included it in the AETR because it is a tax on the project. The Democratic Republic of Congo regime has an excess profits tax that is triggered for a mine when the realized price is at least 25 percent higher than the price in its feasibility study. I assumed that the feasibility study has a price of $1,300 per ounce, so the excess profits tax is not triggered.
-
20
Total benefits in this case are a project’s revenues minus operating costs and replacement capital (but not minus exploration and development capital). This cash flow represents the money available to pay back the initial investment and provide a return. The government share of it is a common measure of progressivity.
-
21
Wen, Progressive Taxation of Extractive Resources as Second-Best Optimal Policy.
-
22
With a discount rate of 10 percent. The results for only some countries are shown to clearly depict each data point. The results for all the evaluated countries can be found in my model.
-
23
Scurfield, “Tanzania Strikes a Better Balance with its Mining Fiscal Regime.”
-
24
This feature is not fully reflected in Figure 5 given that “total benefits” use a slightly different definition of profits and are based on discounted cash flows.
-
25
With a gold price of $1,600 per ounce.
-
26
For example, an average 62 percent of respondents to the Fraser Institute surveys between 2017 and 2019 said the current implementation of Tanzania’s legal system would strongly discourage investment, and 73 percent said regulatory uncertainty would. See, e.g., Ashley Stedman, Jairo Yunis and Elmira Aliakbari, Fraser Institute Annual Survey of Mining Companies 2019 (Fraser Institute, 2020), www.fraserinstitute.org/studies/annual-survey-of-mining-companies-2019.
-
27
With a low-profit mine and a gold price of $1,600 per ounce.
-
28
Tax avoidance could extend the Philippines’ recovery period, and therefore delay the payment of some taxes including import duty and interest withholding tax, given the end of the recovery period depends on the reported profitability of a mine rather than an ex-ante assessment. However, the rule that the recovery period must end five years from the start of production regardless of whether pre-production expenses have been recouped limits the extent to which it can be extended.
-
29
The merits of these measures require further scrutiny. E.g., taking a share of loan repayments could result in lenders charging a higher interest rate to ensure they still recoup their loan and a minimum return. This would not only reduce taxable income but also make it harder for the government to assess whether an interest rate is reasonable, because it would not be comparable with industry benchmarks. It can also be difficult for a government to always determine whether a loan is from a related party or not. However, these considerations are outside the scope of this analysis.
-
30
Natural Resource Governance Institute (NRGI), Natural Resource Charter, 2nd edition, 2014, resourcegovernance.org/analysis-tools/publications/natural-resource-charter-2nd-ed.
-
31
Although its exclusion of several significant taxes from the government’s share of benefits means low-profit mines may still be impacted.
-
32
With a gold price of $1,600 per ounce.
-
33
Anna Fleming, Thomas Lassourd and David Manley, “Variable Royalties: an Answer to Volatile Mineral Prices?” in Handbook on the Future of Resource Taxation, African Tax Administration Forum and Intergovernmental Forum on Mining, Minerals, Metals and Sustainable Development (forthcoming), www.iisd.org/publications/brief/future-resource-taxation-roadmap.
-
34
Ensuring that interest rates used as a comparison apply to comparable assets with a similar risk profile is challenging, but rules of this nature have been successful in reducing profit shifting in other countries. See Beer and Devlin, Is There Money on the Table?.
-
35
The main taxes listed are VAT, royalty and corporate income tax. The regime also includes a share of pre-tax profits that is currently divided between the company’s workers and the government, with the portion received by the government included in its accumulated benefits. However, a recent court ruling means that all this labor profit share will go to workers from the start of 2024 and therefore none of it will be included in government benefits.
-
36
The discount rate used is specific to each mine and based on its weighted average cost of capital (WACC). I have assumed that WACC is around 7 percent in real terms. This is based on the typical discount rate for equity shareholders used by industry and government analysts of 8 percent in real terms, and the current average cost of debt for the mining sector as reported by Aswath Damodaran, Damodaran Online, www.pages.stern.nyu.edu/~adamodar.
-
37
Republic of the Philippines, Financial or Technical Assistance Agreement, mgb.gov.ph/attachments/article/79/PFC_FTAA.pdf.
-
38
For example, some terms in the original FTAA for an OceanaGold mine differed in some areas: Republic of the Philippines, Financial or Technical Assistance Agreement with Arimco Mining Corporation, 1994, www.resourcecontracts.org/contract/ocds-591adf-2792396017. I understand that the recently signed extension to this agreement has slightly different terms again.
-
39
A template of the FARI model and a user guide that explains all the concepts and workings of the model are available at International Monetary Fund, “Fiscal Analysis of Resource Industries,”www.imf.org/external/np/fad/fari.
Equitable Sharing of Mining Profits: The Best Deal for Tanzania?
-
Tanzania’s unique approach to mining taxation, described as an equitable sharing of economic benefits between government and mining company according to a negotiated split, could leave the country shortchanged.
-
Tanzanian citizens have access to the terms of only one profit-sharing deal, a 50-50 framework agreement for Barrick’s gold mines. The rest remain unpublished, despite laws that require the government to disclose deal terms.
-
The Barrick deal predicates government revenues more heavily on mine profitability than a more typical regime, making these revenues more uncertain. This structure also increases the risk that companies may seek to avoid tax payments.
-
Negotiating the split on a project-by-project basis, without any legal guardrails on the approach, increases the risks of corruption and unfavorable deal terms.
-
The Tanzanian government could eliminate these risks by borrowing provisions from other countries that also require a specific government share of mine profits but take a different approach.
-
However, unless the government has identified benefits from its current sharing mechanism that a more typical regime does not offer, it should not pursue a sharing mechanism and instead focus on improving the underlying regime, such as increasing its flexibility with regards to profits and reducing tax avoidance risks.
-
To gain more trust from citizens, Tanzania’s government should explain how the mechanism works in existing deals.
Summary
Tanzania has taken an innovative approach to mining taxation in recent years, based on an “equitable sharing principle.” It has negotiated deals by which the government and mining companies share “economic benefits” based on an agreed split. Public information about these deals is limited. My understanding of the sharing arrangement is based on: the framework agreement template in the Mining (State Participation) Regulations 2022, the framework agreement between the government and Barrick Gold that first detailed a sharing arrangement in 2019, and discussions with government and industry officials. The agreement with Barrick provides for a 50-50 split. I understand that the government has negotiated a larger share in at least some subsequent deals, but there is no public information about them.
“Economic benefits” are the sum of government revenues, return of shareholder capital, shareholder dividends and shareholder loan repayments. Therefore, by the end of a typical mine’s lifetime, its economic benefits should be broadly equivalent to its total pre-tax profit. Mining companies continue to pay the taxes set out in the legislated fiscal regime, such as royalty and corporate income tax. But discussions indicated that once a mining company has recouped its initial investment, sharing is triggered, and adjustments are made at the end of each year to achieve the agreed split. This sharing arrangement is a significant shift away from the fiscal regime set out in Tanzanian legislation and common across the world.
There are several other countries that require mining companies to pay a specified share of their profits to the government, such as Ecuador and the Philippines. However, Tanzania has taken a different approach. While Ecuador and the Philippines impose only a floor on the government share, the Tanzanian arrangement also imposes a ceiling. This ceiling—which in the Barrick example limits the government share of profit to 50 percent—has several consequences.
My modeling of the impact of the regime on both a gold mine and a nickel mine suggests that the 50-50 regime results in a lower government take than the country’s underlying regime set out in legislation. The underlying regime is likely to result in the government receiving a share of profits greater than 50 percent for much of a mine’s lifetime. Under the 50-50 arrangement, the government therefore must make a payment to the mining company or forgo future tax payments to rebalance.
This does not necessarily mean the 50-50 regime will generate less government revenue overall. The competitiveness of the underlying regime is uncertain, and companies could develop fewer mines under it. However, modeling suggests that the 50-50 regime also results in lower taxes than in many other countries in a comparative sample. A slightly larger government share, which I understand the government has negotiated in at least some other deals, could therefore represent a better balance.
Tanzania sharing of benefits across the lifetime of a gold mine making average profits1
The ceiling on government share has another consequence. It links all government revenues to mine profitability. The government can receive only a total amount capped at 50 percent of total benefits, irrespective of which tax initially provided its revenues. Low profits, which translate to minimal total benefit, could therefore impact even the payment of taxes not directly based on profit. This greater reliance on profitability introduces two risks for the Tanzanian state: it makes revenues less reliable and increases the risk that the mining company will seek to avoid tax payments.
Modeling suggests that there may be years in which the government receives no revenues from a mine that is producing but making low profits. The figure below shows how the government may not receive revenues for two years because of the build-up of benefits that the government receives prior to the point that sharing is triggered. With a mine of average profitability, total benefits are large enough that the government does not have to forgo all its tax payments in any year. However, with low profits, the mining company must retain all benefits for a period to rebalance. The example in the figure below shows the potential impact of the government’s accumulation of benefits prior to the sharing trigger, however periods of no or low profits at other points in the mine’s lifetime could also make revenues volatile.
Tanzania sharing of benefits across the lifetime of a gold mine making low profits, with all benefits shared
The sharing mechanism also significantly increases the government’s exposure to the risk of mining companies avoiding taxes. All government revenues become dependent on government capacity to effectively assess profit rather than just the government revenues from profit taxes. The figure below shows the potential impact of tax avoidance practices increasing costs and therefore reducing total benefits. With the underlying regime, only profit-based taxes such as corporate income tax would be impacted by such tax avoidance practices. These practices would not affect input and production taxes such as royalties. However, with the sharing mechanism, such practices could result in artificially low profits. This could trigger the cap for the government share of benefits and mean reductions in payments such as royalties, even though these payments are not usually dependent on profitability—which is one of the main reasons they are included in fiscal regimes.
Hypothetical exposure to tax avoidance risk with Tanzania's fiscal regimes2
The government may not have anticipated these risks because the sharing arrangement was initially developed for mines that were already producing, for which the risks are lower. The largest costs for producing mines have already been incurred, and so there is a lower risk of the sharing mechanism making revenues unreliable and there are fewer opportunities for tax avoidance. However, the government should consider whether this approach is optimal for new mines.
A larger government share of benefits reduces the risks of unreliable revenues and of tax avoidance. The higher the ceiling, the less likely government revenues are to hit it. Payments of input and production taxes such as royalties are therefore less likely to be reduced. However, a larger government share does not eliminate this risk.
The flurry of framework agreements that the Tanzanian government has signed since the Barrick deal suggests that investors like the sharing arrangement. It does not appear to be beneficial for Tanzania, however. This is particularly concerning given that the energy transition and resulting surge in demand for many of Tanzania’s minerals mean there is even more to lose.
Negotiating the split on a project basis, without any guardrails in legislation on the approach to this split, increases the risk of a bad deal for the country. The Mining (State Participation) Regulations allow the government to negotiate the sharing of economic benefits. However, legislation provides only a limited methodology for calculating economic benefits and does not provide a range within which the split will be agreed. Given that the split essentially overrides the underlying fiscal regime set out in legislation, this lack of guardrails is concerning. It increases the impact of uneven capacity between the government and company negotiating teams, the likelihood of costly mistakes, and the risk of corruption. An uneven fiscal regime is also more difficult to administer.
The government could look to Ecuador and the Philippines’ regimes for ideas on how to refine its approach. Those countries have avoided these risks by setting only a floor on the government share. By allowing the mining company to earn its minimum return before sharing is triggered, Ecuador’s sharing mechanism performs more like a windfall profit tax. It therefore reduces the risk of making marginal mines unviable.
However, removing the ceiling from Tanzania’s sharing mechanism will likely result in most mines paying the tax level in the underlying regime. This is because the relatively high taxes of Tanzania’s underlying regime will make the floor on the government share redundant. The government could therefore decide not to pursue a sharing mechanism of any form in new deals, since it becomes an unnecessary complication.
Unless the government has identified benefits to the sharing mechanism that a more typical regime does not offer, it is better for the country that the government focuses on improving the underlying regime. Tanzania’s taxes are less responsive to mine profitability than most other regimes, which risks making some of Tanzania’s mines unviable. The government could therefore consider a royalty rate that varies with prices. It could also review its approach to tackling tax avoidance risks from inflated financing costs. For example, rather than taking a share of shareholder loan repayments, which may not have the intended effect, it could prohibit the interest rate on any loan from exceeding the lowest market rate available for such loans.
Even if the government decides not to incorporate a sharing mechanism in any new deals, explaining to the public how the mechanism works in existing deals is critical. The lack of public information and understanding of the mechanism risks undermining public trust in the government’s management of the sector. Part of this public sensitization should include disclosure of the contracts that contain the “benefit sharing” approach, as already required by law.
Introduction
In 2019, Tanzania and Barrick Gold renegotiated the mineral development agreements (MDAs) for three of the country’s gold mines. The renegotiated deals provide for a 50-50 share of “economic benefits.” Since then, the government has entered a 55-45 sharing arrangement with Petra Diamonds for Tanzania’s only large-scale diamond mine, with the government receiving the larger share.3 These mines are already producing, but the government has also applied this sharing approach to deals for at least some new projects. The largest mine in the pipeline—Kabanga, which will produce nickel and cobalt—has signed a framework agreement with the government that provides for economic benefits to be “equitably shared.”4
This sharing arrangement is a significant shift away from the fiscal regime set out in Tanzanian legislation and common across the world. The typical regime is based on royalties and corporate income tax, without any adjustment mechanism to achieve a specified government share of mine profits overall. I analyzed this regime previously, benchmarking it against a range of other mining countries.5 I now analyze the Tanzanian sharing arrangement to establish whether it represents a better deal for the country. I then explore how the government could refine it as it negotiates deals for new projects in the pipeline. Lessons are drawn from the approach of Ecuador and the Philippines, which also target a specified share of a mine’s profits for the government but in different ways.
The economic model and data that I used for this analysis are available on the Natural Resource Governance Institute’s website here. I provide further information on my modeling approach in subsequent sections and the appendix.
1. Key difference between benefit-sharing approaches: Tanzania’s ceiling on the government share
Public information on the structure of Tanzania’s sharing arrangements is limited. While the Mining (State Participation) Regulations 2022 provide for the government to negotiate the sharing of economic benefits on the basis of an “equitable sharing principle,” they do not fully set out its structure. The Barrick MDAs and agreements for other projects have not been disclosed. The framework agreement between the government and Barrick that informed the MDAs provides the most comprehensive outline of the structure that is publicly available.6 I therefore focus on this 50-50 arrangement, but also discuss the implications of a different split, such as the 55-45 agreed with Petra Diamonds, in my analysis. The three key elements of the sharing arrangement are set out below.
Definition of economic benefits. “Economic benefits” could be interpreted to mean different things. The broadest definition would entail not only direct financial benefits from the mine in the form of government revenues and company profits, but also benefits such as the creation of jobs and demand for goods and services. However, the Mining (State Participation) Regulations state that “economic benefits” comprise of government revenues directly payable by the company to the government,7 shareholder dividends and shareholder loan repayments. Therefore, by the end of a typical mine’s lifetime, its economic benefits should be broadly equivalent to its total pre-tax profit. These benefits are calculated on a cumulative basis from the start of a mine’s operations (that is, from the start of the license period). I understand that cumulation in this case is based on actual not discounted cash flow, and therefore does not account for the time value of money.8
Trigger for sharing mechanism. Sharing is triggered once cumulative, post-tax cash flow is positive, meaning that all exploration and development costs have been recouped. My understanding is that the timing of this trigger is set in advance using the mine life plan rather than based on actual performance.
Sharing mechanism. The mining company pays the taxes set out in the fiscal regime, such as royalty and corporate income tax, in line with the typical approach. However, after the sharing trigger, if one party’s cumulative share of economic benefits is greater than 50 percent at the end of a year, it must pay the other party the amount required to rebalance. This payment could be made at that point in time or, if the company has overpaid, treated as an advance payment of future taxes.
Several other countries require mines to pay a specified government share
The Ecuador and Philippines regimes also require a government share of profits of 50 percent. Each country’s definition of profits differs slightly. However, a more significant difference is that, unlike under the Tanzanian approach, if the Ecuador or Philippines government receives a share of more than 50 percent, the government does not have to compensate the mining company. In other words, their mechanisms act as a floor but not a ceiling on the government’s share.
Another key difference between the regimes is that, while the Tanzania and Ecuador regimes calculate the benefits accrued by the government or mining company on a cumulative basis, the Philippines regime does not. However, while cumulation is based on actual cash flow in the Tanzania regime, it is based on discounted cash flow in the Ecuador regime. This means that sharing is triggered in Ecuador only once the mining company has earned its minimum rate of return—that is, the return the company needs to be willing to develop the mine initially.9
Further details of the Ecuador and Philippines regimes are set out in the appendix.
Modeling the benefit-sharing arrangement for gold and nickel mines
I modeled how the Tanzania, Ecuador and Philippines approaches may work for a gold mine and a nickel mine of average profitability. I assumed a gold price of USD 1,600 per ounce and a nickel price of $17,000 per tonne, which are around the 10-year average between 2013 and 2022.10 In the figures below, “initial government benefits” are the revenues the government would have received without the sharing mechanism and “final government benefits” the revenues it ultimately receives after any rebalancing. I show only the results for the gold mine below, but the results for the nickel mine were similar.
Tanzania. As Figure 1 shows, the government receives benefits from the mine prior to the sharing trigger from input taxes, royalty and some corporate income tax payments. Sharing is triggered in year 7 of the project’s life. The government will have a cumulative share that is significantly greater than 50 percent at this point. It therefore must make a payment to the mining company or forgo future tax payments to rebalance. I modeled the latter, given it will probably be politically difficult for the government to make a direct payment to the company. Following this initial rebalancing, the fiscal regime continues to generate a larger initial share of the benefits for the government than for the company. The government therefore continues to forgo a portion of future tax payments to rebalance. At no point in this scenario does the sharing mechanism result in an increase in government revenue.
Figure 1. Tanzania sharing of benefits across the lifetime of a gold mine making average profits11
Ecuador. As Figure 2 shows, sharing is triggered in year 8 of the project’s life. This is a year later than with the Tanzania approach because, unlike with the Tanzania approach, the mining company is permitted to earn its minimum return before sharing. From this point until year 12, annual company returns are significantly greater than government revenues. Therefore, despite the government having received revenues in previous years, its share of cumulative benefits falls below 50 percent, and the company makes an additional payment. However, these payments do not result in the government and company receiving the same monetary amount each year. Sharing is based on discounted cash flow. The revenue the government received in the early years of the mine is worth more than the same monetary amount received by the company later. This reduces the size of the additional payments the company needs to make for cumulative benefits to be shared equally.
Figure 2. Ecuador sharing of benefits across the lifetime of average gold mine12
Philippines. As Figure 3 shows, the recovery period ends after three years of production, in year 7 of the project’s life. Because the basic government share is less than 50 percent of net mining revenue at this point until year 16, the mining company pays an additional government share. For the following few years, the basic government share is at least 50 percent of net mining revenue. This means the company does not make any additional payments. However, unlike with the Tanzania approach, the government does not have to compensate the company for receiving more than 50 percent of net mining revenue.
Figure 3. Philippines sharing of benefits across the lifetime of average gold mine13
2. Tanzania’s benefit-sharing arrangement could leave the country shortchanged
I reviewed the Tanzanian approach against the typical objectives of a fiscal regime: maximizing government revenue without deterring investment; reliable revenues; flexibility as profits change; and simplicity to limit tax avoidance risks.
Typical objectives of a fiscal regime
Maximizing government revenue without deterring investment. A fiscal regime should set the highest taxes that a mine can bear without discouraging investment. What this tax level is depends on the country’s wider investment climate. This is because investors often place greater weight in their investment decision on several non-fiscal factors, such as a country’s geology and policy predictability, than on the tax level.14
Reliable revenues. Many governments want a regime that generates some revenue for their budget each year irrespective of whether a mine is making low or high profits. Reliability is particularly important for countries that depend heavily on the mining sector to finance the budget. However, high taxes when a mine is making no or low profits can prevent low-profit mines from being developed or surviving downturns. A regime must therefore balance reliability with ensuring a wide range of mines are viable across different scenarios.
Flexibility as profits change. A regime should aim to capture any profit above the investor’s minimum return, with flexible (or “progressive”) taxes significantly increasing government take once this threshold has been reached.15
Simplicity. Governments are often concerned about tax avoidance.16 Some taxes (namely those based on profitability) are more difficult to measure than others, which makes them more susceptible to company manipulation. Governments therefore tend to want the regime to contain at least some simple taxes (such as royalties), depending on their tax administration capacity.
I set out below how the 50-50 regime performs against these criteria in relation to a new gold mine. I compared the performance of the 50-50 regime with the underlying regime. I also compared it with the Ecuador and Philippines regimes—with and without their sharing mechanisms to isolate their effect and draw lessons for Tanzania—as well as with some of the world’s other gold producers. Tanzania is competing with these countries for investment, and therefore how they compare is important. I also analyzed the impact of these regimes on a nickel mine and found similar results.
The benefit-sharing arrangement reduces government take
Tanzania’s 50-50 regime has a lower government take than the country’s underlying regime set out in legislation. This is because of the ceiling it imposes on the government’s share. This does not necessarily mean it will generate less government revenue overall, however. More mines could potentially be developed under the 50-50 regime if investors prefer it, thereby resulting in higher aggregate government revenues.
Figure 4 sets out how the government take differs between Tanzania’s two regimes, and how they compare to the regimes of Ecuador, the Philippines and several other gold producers. I used the average effective tax rate (AETR) to make this comparison.
The 50-50 regime appears to strike a reasonable balance between generating government revenue and competitiveness for a mine with average profitability. As Figure 4 shows, I estimate it has an AETR of around 54 percent.17 However, this is lower than for several other countries in my sample. Tanzania therefore probably has scope for slightly higher taxes than the 50-50 regime entails without reducing the country’s competitiveness. This is particularly the case given government efforts to improve the wider investment climate appear to be paying off. Control Risks considers Tanzania’s legal and tax risks to be on a downward trend.18 Indeed, I understand the government has negotiated a larger share in at least some other deals.
In contrast to the Tanzanian mechanism, Ecuador’s and the Philippines’ mechanisms both increase government take from a mine with average profitability, compared to the regimes that would otherwise be applicable. However, the Ecuador mechanism has a larger impact. This is because it does not count several significant taxes—such as import duties and withholding taxes on interest and dividends—toward the government’s share of benefits. The exclusion of these taxes increases the additional payment the mining company must make for the government to receive 50 percent of cumulative benefits.
Figure 4. Average effective tax rate for average gold mine with gold price of USD 1,600 per ounce19
The benefit-sharing arrangement both limits windfall profits capture and makes revenues unreliable
Figure 5 shows government take at different profit levels—measured in terms of government share of total benefits.20 A regime should set government take from low-profit mines to generate reliable revenues while still providing the minimum return to investors. A flexible regime would then have much higher government take from higher-profit mines. The AETR for these mines would be flat, with flexible instruments to capture windfall profits balanced out by reliable instruments such as royalties. This translates to government share of total benefits slightly increasing with profit.21
Figure 5. Government share of total benefits at different profit levels22
The instruments in Tanzania’s underlying regime make it relatively inflexible. Its corporate income tax and minority state equity mean that it should capture windfall profits reasonably well, if effectively administered. However, the regime also places high taxes on low-profit mines, largely because of a high royalty. As I noted in my previous analysis, the government could consider adjusting these instruments to make the underlying regime more flexible.23
The sharing mechanism also limits the flexibility of the regime since it is designed to preserve the share of government in total benefits, regardless of economic conditions. The mechanism results in the regime capturing a similar share of profits whether a mine generates low or high profits.24 This means the government must decide, when determining the split between government and mining company, whether to impose a large government take on low-profit mines and capture a large share of any windfall profits; or to provide more relief to low-profit mines but in doing so to capture a smaller share of windfall profits. The 50-50 split takes the latter approach, making low-profit mines more likely to be viable than with the underlying regime. However, the sharing mechanism means that this must come at the cost of a smaller government share of windfall profits.
The sharing mechanism also makes government revenues unreliable. This is because it makes all government revenues dependent on the size of total benefits. The government can receive only 50 percent of total benefits irrespective of which tax they were initially derived from. Low profits, which in turn mean total benefits are small, could therefore impact even the payment of taxes not directly based on profit.
As Figure 6 shows, my modeling suggests there may be years in which the government does not receive any revenues from a mine with low profits. In this scenario, these two years of no government revenues are caused by the build-up of benefits that the government receives prior to the sharing trigger. Once sharing is triggered, the government must forgo tax payments to rebalance and for 50-50 sharing to be achieved. With a mine of average profitability, total benefits are large enough that the government does not have to forgo all its tax payments in any year, as shown in Figure 1. However, with low profits, the mining company must retain all benefits for a period to rebalance.
While this example shows the potential impact of the government’s accumulation of benefits prior to the sharing trigger, periods of no or low profits at other points in the mine’s lifetime could also make revenues volatile. Therefore, while the sharing mechanism may still generate reasonable government revenues over the project’s lifetime, it appears to undermine the reliability of revenues from year to year. As a result, it appears to create a regime that neither captures a significant share of windfall profits compared to many of the countries in our sample nor provides reliable revenues when profits are low.
Figure 6. Tanzania sharing of benefits across the lifetime of a gold mine making low profits, with all benefits shared25
This lack of reliability could have significant repercussions for public trust. As noted above, I do not think it will be politically feasible for the government to make a direct payment to the mining company when the government has captured a greater share of the benefits than the agreed split. It is more likely that the government will forgo future tax payments. However, a scenario in which the government will need to forgo all tax payments in a given year will also be politically difficult. As a result, the sharing arrangement could create situations in which the government feels forced into entering supplementary arrangements to ensure it receives some revenues each year—for example, an arrangement that spreads the amount required to rebalance over several years.
Such arrangements, whether fully disclosed to the public or not, could lead to the sharing approach reducing public trust. This loss of trust would put pressure on the government to change the fiscal regime and even the wider legal and regulatory framework. If acted on, this could shake confidence in policy predictability and deter investment, as the country experienced following the sweeping reforms in 2017.26
In contrast to the Tanzania mechanism, the Ecuador mechanism helps increase flexibility. The mining company being able to earn its minimum return before sharing, and sharing being based on cumulative benefits, means the mechanism has a limited impact on low-profit mines. For higher-profit mines, it increases government take. In this way, it operates much like a windfall profit tax. Nonetheless, its flexibility could be improved further. The exclusion of some taxes from the government’s share of benefits means that even low-profit mines may need to make an additional payment. This is particularly impactful because the rest of the Ecuador regime imposes a relatively high burden on mines even when profits are low.
The Philippines’ mechanism improves the government’s ability to capture windfall profits. However, it risks increasing the burden on low-profit mines. This is because, while the limited taxes during the recovery period provide some relief, sharing is triggered before the mining company earns its minimum return and is based on annual not cumulative benefits. This also limits the scope to capture an even greater share of windfall profits through the sharing mechanism, as a larger government share would risk a higher burden on low-profit mines. A separate windfall profit tax would therefore be needed for this purpose instead.
The benefit-sharing arrangement increases tax avoidance risks for Tanzania
The Tanzania mechanism significantly increases the government’s exposure to tax avoidance risks. As noted above, all government revenues become dependent on profit. They therefore all become dependent on government capacity to effectively assess profit rather than just the government revenues from profit taxes.
Figure 7 shows the potential impact of tax avoidance practices increasing costs and therefore reducing total benefits. The increase in costs does not affect the payment of input and production taxes with the underlying regime. However, it does with the sharing mechanism as it means 50 percent of total benefits includes a lower value of the input and production taxes that would have been due to the government.
Figure 7. Hypothetical exposure to tax avoidance risk with Tanzania's fiscal regimes27
This feature of the mechanism exists regardless of the size of the split. A larger government share reduces the risk that tax avoidance will lower total benefits to the extent that the government’s share is less than the value of input and production taxes. However, it does not eliminate that risk.
The Ecuador and Philippines mechanisms, in contrast, do not significantly affect government exposure to tax avoidance risks. Determining the need for additional payments to government depends on government capacity to effectively measure profits. But because these regimes do not place a ceiling on the government’s share, tax avoidance that reduces profits would not affect the payment of input and production taxes.28
The Tanzania regime does contain measures intended to reduce tax avoidance risks related to financing costs. I understand that the Barrick MDAs treat related party loans as being interest free in the calculation of total benefits. The Mining (State Participation) Regulations also unusually provide for mandatory state equity in projects to give the government a share of any shareholder loan repayments. This provision is intended not only to generate larger government revenues directly, but also to reduce the incentive to shift profits through inflated financing costs.29
These protective measures are not tied to the structure of the sharing mechanism, however. They could be included in any regime. Indeed, the sharing mechanism appears to neutralize the impact of taking a share of loan repayments. The government can receive only 50 percent of economic benefits regardless of whether they are generated from a share of loan repayments or other revenue streams.
3. Tanzania may do better by not including the benefit-sharing mechanism in new deals
Setting not just a floor but also a ceiling on the government share has introduced three main risks into Tanzania’s regime: it potentially sets an unnecessarily low ceiling on government revenues (depending on the agreed split); it makes those revenues less reliable; and it increases their exposure to tax avoidance.
The government may not have anticipated some of these consequences because it initially developed the sharing arrangement for mines that were already producing. The largest costs for these mines have already been incurred and they are now profitable. This reduces the risk of the sharing mechanism making revenues unreliable and increasing the potential for tax avoidance. However, it would be advisable for the government to consider whether this approach is optimal for new mines. Particularly given the energy transition and resulting surge in demand for many of Tanzania’s minerals, there is even more to gain from a good deal and even more to lose from a bad deal.
A larger government share of benefits reduces the risk of unreliable revenues and tax avoidance. However, it does not eliminate this risk. Negotiating the split on a project basis, without any guardrails in legislation, increases the risk of a bad deal. Legislation provides only a limited methodology for calculating economic benefits and does not provide a range within which the split will be agreed. Given that the split essentially overrides the underlying fiscal regime set out in legislation, this lack of guardrails is concerning. It increases the impact of uneven capacity between the government and company negotiating teams, the likelihood of costly mistakes and the opportunity for corruption. An uneven fiscal regime is also more difficult to administer.30
The government could look to Ecuador and the Philippines’ regimes for ideas on how to refine its approach. By setting only a floor on the government share, their sharing mechanisms do not exacerbate the risk of unreliable revenues and tax avoidance. By allowing the mining company to earn its minimum return before sharing is triggered, Ecuador’s sharing mechanism performs more like a windfall profit tax. It therefore reduces the risk of making marginal mines unviable.31
Removing the ceiling from Tanzania’s sharing mechanism is complicated by the relatively high taxes of Tanzania’s underlying regime. A floor of 50 percent of profits results in Ecuador and the Philippines receiving additional payments because government take from other taxes in their regimes does not reach that threshold. This is not the case for Tanzania as government take already exceeds that floor.
Figure 8 shows my modeling of a gold mine of average profitability. The mechanism—without a ceiling and the calculation of benefits based on discounted cash flow to allow mining companies to earn their minimum return before sharing is triggered—does not have any effect on the share of benefits between the government and company. My modeling suggests the mechanism would also not have any effect on this share for mines of higher and lower profitability.
Figure 8. Tanzania sharing of benefits across the lifetime of a gold mine making average profits with no ceiling32
Removing the ceiling from the sharing mechanism will therefore likely result in most mines paying the tax level in the underlying regime. The government could of course increase the floor so that the mechanism does result in mines making additional payments. My modeling suggests a floor of close to 60 percent would trigger additional payments from a range of mines. However, this would not be advisable given the underlying regime already has higher taxes than most other countries in my sample. Increasing government take would further challenge Tanzania’s competitiveness, even if these additional payments were payable only once the mining company had earned its minimum return.
The adjustments I suggest would therefore make the sharing mechanism superfluous for most mines. The government could therefore decide not to pursue a sharing mechanism of any form in new deals given it becomes an unnecessary complication. Indeed, the government is better focusing on improvements to the underlying regime. As set out above, Tanzania’s taxes are less responsive to mine profitability than most other regimes I analyzed. The government could therefore consider a royalty rate that varies with prices.33 This would provide some relief to mines when prices are low, and capture more revenue for government when prices are high. The government could also review its approach to tackling tax avoidance risks from inflated financing costs. For example, rather than taking a share of any loan repayments, it could consider adopting a rule for the mining sector that is currently only in the Petroleum Act: that the interest rate on a loan should not exceed the lowest market rate available for such loans.34
Conclusion
The Tanzanian government should be commended for its willingness to pursue an innovative approach to mining taxation after years of frustration at the deals it had previously struck. However, without reassessing this approach, it risks leaving itself shortchanged again—particularly given the lack of guardrails in legislation. Considering the challenges with the arrangement I have identified, there may be a sense of déjà vu in the longer term with a public backlash against both mining companies and the government.
The government could adjust the mechanism before finalizing deals for new mines by borrowing elements from the Ecuador and Philippines regimes. However, unless the government has identified benefits from the sharing mechanism that a more typical regime does not offer, the best option is not to pursue a sharing mechanism of any form and instead to focus on improving the underlying regime.
Even if the government decides not to incorporate a sharing mechanism in any new deals, explaining to the public how it works in existing deals is critical. The current lack of public information and understanding of the mechanism risks undermining public trust in it. Part of this public sensitization should include disclosure of the contracts that contain the mechanism, as already required by the Tanzania Extractive Industries (Transparency and Accountability) Act 2015. The government could also consider publicly reporting the expected or observed economic benefits from the mines to which a sharing arrangement applies. This should also help increase public trust that the country is receiving the amount that it is due.
Appendix: Ecuador and Philippines regimes
Ecuador’s approach
The Ecuador regime requires mines to provide a government share of at least 50 percent of “accumulated benefits.” This requirement is established in the constitution of the country. It is primarily set out in the general regulations of the mining law.
Definition of accumulated benefits. “Accumulated benefits” are the sum of the government revenues specified in the regulations and any free cash flow available to the mining company.35 These benefits are calculated on a cumulative basis from the start of a mine’s operations (though the applicable government revenue streams are payable only from the start of production). This cumulation accounts for the time value of money. Cash flows are discounted to reflect that, the earlier they occur, the more they are worth to either party.36
Trigger for sharing mechanism. Sharing is triggered once cumulative, discounted free cash flow is positive. At this point, the mining company has recouped all exploration and development costs. Because the trigger is based on discounted cash flow, it will also have earned its minimum rate of return.
Sharing mechanism. The mining company pays the taxes set out in the fiscal regime in line with the taxes and royalties defined in legal and contractual provisions. After the sharing trigger, if cumulative government revenues are less than 50 percent of accumulated benefits at the end of the year, the company must pay a “sovereign adjustment” to increase the government share to 50 percent. However, if government revenues are more than 50 percent of accumulated benefits, the government does not have to compensate the company.
Philippines’ approach
The Philippines regime requires mines that operate under a financial or technical assistance agreement (FTAA) to provide a government share of at least 50 percent of annual “net mining revenue” after a cost recovery period. Any mine licensed to a foreign company must have a FTAA.37 This arrangement is primarily set out in a publicly available model FTAA. The financial or technical assistance agreement for a given project is subject to negotiations and therefore may differ slightly from others.38 I focus on the regime set out in the model FTAA.
Definition of “net mining revenue.” Net mining revenue is sales revenue (net of transport and processing charges) minus deductible expenses in a given year. Deductible expenses include, among other things, development costs after the start of production, operating costs, interest payments and royalties.
Trigger for sharing mechanism. The sharing mechanism is triggered at the end of the “recovery period.” The recovery period ends the earlier of five years from the start of production or the point at which all pre-production expenses have been recouped. The timing of this trigger is based on the actual performance of a mine rather than set in advance.
Sharing mechanism. The mining company pays the “basic government share” throughout the project’s lifetime. However, the composition of this basic government share up to the end of the recovery period and after the recovery period differs. A comprehensive list of the taxes included in the government share during these two periods can be found in the model FTAA and applicable legislation. The main components up to the end of the recovery period include excise tax, royalties and local business tax. The basic government share after the recovery period includes these taxes as well as import duties, corporate income tax, and withholding taxes on interest and dividends.
After the end of the recovery period, if the basic government share is less than 50 percent of net mining revenue in a given year, the mining company must pay an “additional government share” to increase the total government share to 50 percent of net mining revenue. However, if the basic government share is more than 50 percent of net mining revenue, the government does not have to compensate the company.
Modeling approach
I used an adapted version of the IMF’s Fiscal Analysis of Resource Industries (FARI) economic model published in 2016 for this analysis.39
I based my analysis on two mines: one produces gold doré and the other produces primarily nickel sulfide but also some cobalt. As the effects of a fiscal regime can differ depending on the specific cost, production and ultimate profitability of a mine, I created a range of mine profiles. I based the gold mine profiles on the characteristics of actual mines in Tanzania, using S&P Global Market Intelligence’s Capital IQ Pro database. The S&P database does not yet contain detailed economic data for any of Tanzania’s upcoming nickel mines, and therefore I developed rough approximations based on their potential size and the cost of other nickel mines across the world. I also made assumptions about the financing structure of these mines. I based these assumptions on the values used as common practice by industry and government analysts.
The assumptions I used for the two gold mine profiles referenced in this analysis are set out in Table 1. These assumptions result in a pre-tax internal rate of return of 42 percent for the average-profit mine and 17 percent for the low-profit mine.
Table 1. Key assumptions for modeled gold mine profiles
|
Parameter |
Average-profit mine |
Low-profit mine |
|
Peak production of gold |
350,000 ounces |
100,000 ounces |
|
Production life |
24 years |
16 years |
|
Development and expansion capital |
$300 million |
$450 million |
|
Replacement capital per year |
$40 million |
$8 million |
|
Operating cost |
$700 per ounce |
$800 per ounce |
|
Transport, treatment and refining charges (TC/RC) |
$3 per ounce |
$6 per ounce |
|
Debt-to-equity ratio |
50% |
50% |
|
Real discount rate of equity shareholders |
8% |
8% |
|
Real interest rate on debt |
6% |
6% |
|
Inflation rate |
2% |
2% |
The assumptions I used for the nickel mine profiles in this analysis are set out in Table 2. These assumptions result in a pre-tax internal rate of return of 35 percent for the average-profit mine and 16 percent for the low-profit mine.
Table 2. Key assumptions for modeled nickel mine profiles
|
Parameter |
Average-profit mine |
Low-profit mine |
|
Peak production of nickel |
40,000 tonnes |
15,000 tonnes |
|
Associated cobalt (percent of gross sales from nickel) |
15% |
5% |
|
Production life |
30 years |
10 years |
|
Development and expansion capital |
$1,600 million |
$450 million |
|
Replacement capital per year |
$20 million |
$6 million |
|
Operating cost |
$6,000 per tonne |
$8,000 per tonne |
|
Transport, treatment and refining charges (TC/RC) |
$2,600 per tonne |
$2,600 per tonne |
|
Debt-to-equity ratio |
50% |
50% |
|
Real discount rate of equity shareholders |
8% |
8% |
|
Real interest rate on debt |
6% |
6% |
|
Inflation rate |
2% |
2% |
I applied the fiscal regimes in my sample to these mine profiles. To keep the model and my analysis as simple as possible, I did not include taxes that are likely to comprise a relatively small proportion of total government revenue for a large-scale mine, such as annual license fees. I also did not model employment taxes that are ultimately paid by the mine employees rather than the companies. I also simplified some elements of the regimes. For example, while import duties vary according to trade agreements between countries (e.g., East Africa Community trade rules) and according to the goods and services imported, I assumed one rate for all imported inputs.
Acknowledgements
The author wishes to thank Thomas Lassourd, senior policy advisor, and Viola Tarus, policy advisor for tax and extractives, at the Intergovernmental Forum on Mining, Minerals, Metals and Sustainable Development, and Amir Shafaie, Moses Kulaba, Silas Olan’g, Sophia Rwegellera and William Davis of the Natural Resource Governance Institute, for reviewing this brief. He also wishes to thank Thomas Baunsgaard, deputy division chief of the Tax Policy Division at the International Monetary Fund, and Daniel Mulé, policy lead for extractive industries, tax and transparency at Oxfam, for their input.
Notes
-
1
With a gold price of USD 1,600 per ounce.
-
2
With a low-profit mine and a gold price of $1,600 per ounce.
-
3
Cecilia Jamasmie, “Petra Diamonds’ stake in Williamson to shrink as part of deal with Tanzania,” Mining.com, 13 December 2021, www.mining.com/petra-diamonds-stake-in-williamson-to-shrink-as-part-of-deal-with-tanzania.
-
4
Lifezone Metals, “Kabanga Nickel Signs Framework Agreement,” 19 January 2021, www.lifezonemetals.com/kabanga-nickel-signs-framework-agreement.
-
5
Thomas Scurfield and Silas Olan’g, “Magufuli Seeks the Right Balance for Tanzania’s Mining Fiscal Regime,” NRGI, 31 January 2019, www.resourcegovernance.org/blog/magufuli-seeks-right-balance-tanzania-mining-fiscal; Thomas Scurfield, “Tanzania Strikes a Better Balance with its Mining Fiscal Regime,” NRGI, 24 June 2020, www.resourcegovernance.org/blog/tanzania-strikes-better-balance-mining-fiscal-regime.
-
6
This framework agreement was published in a document setting out Barrick’s offer to buy the shares it did not already own in Acacia Mining, the previous owner of the Bulyanhulu, Buzwagi and North Mara mines in Tanzania. See Acacia Mining and Barrick Gold, Recommended Final Offer for Acacia Mining Plc by Barrick Gold Corporation, 2019, 66–79, s25.q4cdn.com/322814910/files/doc_downloads/acacia/Acacia-2.7-announcement.pdf.
-
7
The main revenue streams are import duty, skills development levy, royalty, corporate income tax, a share of dividends and shareholder loan repayments through state equity, and dividend withholding tax.
-
8
The earlier in time a shilling (Tanzania’s official currency unit) is received, the more it is worth. This is, first, because it can be used earlier; and second, because the future is uncertain, and no one can be sure they will receive that shilling in the future. To account for this time value of money, a “discount rate” is applied. In the sharing arrangement, this would mean that if the government received a shilling in year 1, the company would need to receive more than a shilling in year 2 for the benefits to be comparable. However, given cumulation here is based on actual cash flow, the company would need to receive only a shilling in year 2 for the benefits to be shared equally.
-
9
This provision for the company to earn its minimum return before sharing is triggered means Ecuador’s mechanism is similar to an R-based cash flow tax, commonly referred to as a Brown Tax. See, e.g., Robin Broadway and Michael Keen, “Theoretical perspectives on resource tax design,” in The Taxation of Petroleum and Minerals: Principles, Problems and Practice, edited by Philip Daniel, Michael Keen and Charles McPherson (Oxford: Routledge, 2010), 13–74.
-
10
Prices are taken from World Bank, “Commodities Price Data (The Pink Sheet),” www.worldbank.org/en/research/commodity-markets.
-
11
With a gold price of $1,600 per ounce.
-
12
With a gold price of $1,600 per ounce.
-
13
With a gold price of $1,600 per ounce.
-
14
The Fraser Institute survey estimates that, unless there are extremely harmful policies, around 60 percent of an investment decision tends to be based on a country’s geology. The other 40 percent comprises of several other factors, including political stability and policy predictability (given they affect the risk that investors will not be able to secure future returns generated by their investments), a conducive business environment and the tax level. See Julio Mejia and Elmira Aliakbari, Fraser Institute Annual Survey of Mining Companies 2022, (Fraser Institute, 2023), 8, www.fraserinstitute.org/studies/annual-survey-of-mining-companies-2022.
-
15
However, information gaps make it difficult for taxes to be structured to capture all excess profit. See Jean-Franҫois Wen, Progressive Taxation of Extractive Resources as Second-Best Optimal Policy (International Monetary Fund, 2018), www.imf.org/en/Publications/WP/Issues/2018/06/13/Progressive-Taxation-of-Extractive-Resources-as-Second-Best-Optimal-Policy-45923.
-
16
Recent research provides a sense of the potential revenue loss to governments from tax avoidance. The International Monetary Fund recently estimated that sub-Saharan African mining countries could be losing between $450 and $730 million in corporate income tax a year. See Sebastian Beer and Dan Devlin, Is There Money on the Table? Evidence on the Magnitude of Profit Shifting in the Extractive Industries (International Monetary Fund, 2021), www.imf.org/en/Publications/WP/Issues/2021/01/15/Is-There-Money-on-the-Table-Evidence-on-the-Magnitude-of-Profit-Shifting-in-the-Extractive-49983.
-
17
It is perhaps surprising that Tanzania’s 50-50 sharing arrangement generates an AETR greater than 50 percent (with a discount rate of 10 percent). This is despite AETR measuring government take as the share of pre-tax profits, which is larger than “economic benefits” (given economic benefits exclude interest payments). This outcome results from the 50-50 split being based on actual cash flow. The government receives revenue before the mining company through input and production taxes that do not depend on the mine making a profit. Because of these earlier revenues, the government receives a larger share on a discounted basis.
-
18
As reported in the S&P Global database. Legal risks are “expropriation, state contract alteration and contract enforcement risks.” Tax risks are “tax increase and tax inconsistency risks.” Control Risks scores these risks as still “very high” and “high” respectively (following Tanzania’s overhaul of extractives sector laws and other actions against existing mines in 2017) but reducing.
-
19
With a discount rate of 10 percent. While Ecuador’s sharing mechanism does not account for the labor profit share because none of it will go to the government from 2024 onwards, I have included it in the AETR because it is a tax on the project. The Democratic Republic of Congo regime has an excess profits tax that is triggered for a mine when the realized price is at least 25 percent higher than the price in its feasibility study. I assumed that the feasibility study has a price of $1,300 per ounce, so the excess profits tax is not triggered.
-
20
Total benefits in this case are a project’s revenues minus operating costs and replacement capital (but not minus exploration and development capital). This cash flow represents the money available to pay back the initial investment and provide a return. The government share of it is a common measure of progressivity.
-
21
Wen, Progressive Taxation of Extractive Resources as Second-Best Optimal Policy.
-
22
With a discount rate of 10 percent. The results for only some countries are shown to clearly depict each data point. The results for all the evaluated countries can be found in my model.
-
23
Scurfield, “Tanzania Strikes a Better Balance with its Mining Fiscal Regime.”
-
24
This feature is not fully reflected in Figure 5 given that “total benefits” use a slightly different definition of profits and are based on discounted cash flows.
-
25
With a gold price of $1,600 per ounce.
-
26
For example, an average 62 percent of respondents to the Fraser Institute surveys between 2017 and 2019 said the current implementation of Tanzania’s legal system would strongly discourage investment, and 73 percent said regulatory uncertainty would. See, e.g., Ashley Stedman, Jairo Yunis and Elmira Aliakbari, Fraser Institute Annual Survey of Mining Companies 2019 (Fraser Institute, 2020), www.fraserinstitute.org/studies/annual-survey-of-mining-companies-2019.
-
27
With a low-profit mine and a gold price of $1,600 per ounce.
-
28
Tax avoidance could extend the Philippines’ recovery period, and therefore delay the payment of some taxes including import duty and interest withholding tax, given the end of the recovery period depends on the reported profitability of a mine rather than an ex-ante assessment. However, the rule that the recovery period must end five years from the start of production regardless of whether pre-production expenses have been recouped limits the extent to which it can be extended.
-
29
The merits of these measures require further scrutiny. E.g., taking a share of loan repayments could result in lenders charging a higher interest rate to ensure they still recoup their loan and a minimum return. This would not only reduce taxable income but also make it harder for the government to assess whether an interest rate is reasonable, because it would not be comparable with industry benchmarks. It can also be difficult for a government to always determine whether a loan is from a related party or not. However, these considerations are outside the scope of this analysis.
-
30
Natural Resource Governance Institute (NRGI), Natural Resource Charter, 2nd edition, 2014, resourcegovernance.org/analysis-tools/publications/natural-resource-charter-2nd-ed.
-
31
Although its exclusion of several significant taxes from the government’s share of benefits means low-profit mines may still be impacted.
-
32
With a gold price of $1,600 per ounce.
-
33
Anna Fleming, Thomas Lassourd and David Manley, “Variable Royalties: an Answer to Volatile Mineral Prices?” in Handbook on the Future of Resource Taxation, African Tax Administration Forum and Intergovernmental Forum on Mining, Minerals, Metals and Sustainable Development (forthcoming), www.iisd.org/publications/brief/future-resource-taxation-roadmap.
-
34
Ensuring that interest rates used as a comparison apply to comparable assets with a similar risk profile is challenging, but rules of this nature have been successful in reducing profit shifting in other countries. See Beer and Devlin, Is There Money on the Table?.
-
35
The main taxes listed are VAT, royalty and corporate income tax. The regime also includes a share of pre-tax profits that is currently divided between the company’s workers and the government, with the portion received by the government included in its accumulated benefits. However, a recent court ruling means that all this labor profit share will go to workers from the start of 2024 and therefore none of it will be included in government benefits.
-
36
The discount rate used is specific to each mine and based on its weighted average cost of capital (WACC). I have assumed that WACC is around 7 percent in real terms. This is based on the typical discount rate for equity shareholders used by industry and government analysts of 8 percent in real terms, and the current average cost of debt for the mining sector as reported by Aswath Damodaran, Damodaran Online, www.pages.stern.nyu.edu/~adamodar.
-
37
Republic of the Philippines, Financial or Technical Assistance Agreement, mgb.gov.ph/attachments/article/79/PFC_FTAA.pdf.
-
38
For example, some terms in the original FTAA for an OceanaGold mine differed in some areas: Republic of the Philippines, Financial or Technical Assistance Agreement with Arimco Mining Corporation, 1994, www.resourcecontracts.org/contract/ocds-591adf-2792396017. I understand that the recently signed extension to this agreement has slightly different terms again.
-
39
A template of the FARI model and a user guide that explains all the concepts and workings of the model are available at International Monetary Fund, “Fiscal Analysis of Resource Industries,”www.imf.org/external/np/fad/fari.
Together at Last: Reflections on In-Person Learning in the Wake of the Coronavirus Pandemic
The 2022 edition of the “Executive Course on Oil, Gas and Mining Governance” program took place at the University of Oxford in September, after a two-year hiatus caused by the coronavirus pandemic. NRGI and Oxford’s Blavatnik School of Government welcomed 44 participants from 21 countries, representing a wide range of stakeholders from the public, private and non-governmental sectors.
This year’s program built on the Natural Resource Charter as a guiding framework and addressed how the extractive sector has changed in the wake of the pandemic. Throughout the week, course instructors explored both the technical policy aspects of resource management and the complex interplay between governments, companies and citizens. Participants reflected on the elements of a national natural resource strategy and what it would take to develop such a strategy in their respective contexts.
For many this was the first in-person program for since the pandemic, so we were sure to include interactive elements that allowed participants to actively learn from each other about key challenges and best resource governance practices. To make the course as engaging as possible, we asked participants to each identify one personal challenge related to the topics covered in the course and discuss it with others so they could gain a broader perspective on these issues. Participants noted this engendered meaningful conversations that facilitated collaborative information sharing and knowledge building for all. The facilitation of these types of interactive, peer-to-peer activities proved key to the overall impact of the course, underscoring the importance of in-person delivery of capacity development programs and the need to continue to offer a mix of physical and virtual trainings in the future.
During the course, I sat down with participants to learn more about their experiences and thoughts on the program. Many were excited to be able to come together to discuss recent developments in the extractive sector. One participant noted that gathering in Oxford “provided an invaluable opportunity to interact, share and learn from different players in the oil, gas and mining sectors from around the world.”
One participant discussed regulatory and policy gaps in their country and the role informed citizenry and community participation play in shaping law-making. In this respect, the course was instrumental in hearing the perspectives of other countries and lessons learned.
Another participant echoed this sentiment, saying, “the interactions and space for team building during the course allowed for establishing lasting bonds with other participants and faculty. The conversations and discussions outside of class helped to learn more about other countries’ perspectives on natural resource management and governance. That might not have been possible in an online setting, where daily tasks take over and do not allow one to fully immerse in the learning experience.”
This year’s program provided a unique opportunity to meet with colleagues from different geographic locations and specializations that one may not normally meet in their professional circles.
A third participant shared that “this year’s program provided a unique opportunity to meet with colleagues from different geographic locations and specializations that one may not normally meet in their professional circles. The course provided a platform for rich discussions and knowledge sharing. It facilitated networking, which would go a long way to sustaining built relationships.” Another learner highlighted license allocation and the negotiation process as a challenge in their country and noted that the course provided an enriching platform to discuss this with faculty and industry experts.
A fifth participant highlighted the uniqueness of the course, which provided “the space to address recent developments in the oil and gas industry with world-class experts and the opportunity to think about tangible policy and non-policy solutions to bring back home.”
From my discussions with participants, it became clear that the program is important for fostering relationships. “The practical nature of the program exceeded my expectations, particularly how well it addressed the issues we face in our everyday roles,” noted one. “I have established and exchanged contacts with several fellow participants to enhance future collaborations and gained information for future engagements in promoting transparency and good governance in the natural resource governance within my areas of operations and with other partners outside the network.”
Another participant said they were able to meet with participants from neighboring countries and discuss in-depth common challenges in extractive sector investments. It facilitated a conversation they both wanted to continue at home and a sense of working together for better management and governance of oil, gas and mineral resources.
Hear more participant perspectives from Argentina, Ghana, Kenya, Libya, Sierra Leone, Tunisia and more, in a series of video interviews.
Photos by John Cairns for NRGI
Executive Course on Oil, Gas and Mining Governance
This course is for senior policy-makers to manage and govern their nation’s oil, gas or mineral resources for a better future.
Notes
-
1
With a gold price of USD 1,600 per ounce.
-
2
With a low-profit mine and a gold price of $1,600 per ounce.
-
3
Cecilia Jamasmie, “Petra Diamonds’ stake in Williamson to shrink as part of deal with Tanzania,” Mining.com, 13 December 2021, www.mining.com/petra-diamonds-stake-in-williamson-to-shrink-as-part-of-deal-with-tanzania.
-
4
Lifezone Metals, “Kabanga Nickel Signs Framework Agreement,” 19 January 2021, www.lifezonemetals.com/kabanga-nickel-signs-framework-agreement.
-
5
Thomas Scurfield and Silas Olan’g, “Magufuli Seeks the Right Balance for Tanzania’s Mining Fiscal Regime,” NRGI, 31 January 2019, www.resourcegovernance.org/blog/magufuli-seeks-right-balance-tanzania-mining-fiscal; Thomas Scurfield, “Tanzania Strikes a Better Balance with its Mining Fiscal Regime,” NRGI, 24 June 2020, www.resourcegovernance.org/blog/tanzania-strikes-better-balance-mining-fiscal-regime.
-
6
This framework agreement was published in a document setting out Barrick’s offer to buy the shares it did not already own in Acacia Mining, the previous owner of the Bulyanhulu, Buzwagi and North Mara mines in Tanzania. See Acacia Mining and Barrick Gold, Recommended Final Offer for Acacia Mining Plc by Barrick Gold Corporation, 2019, 66–79, s25.q4cdn.com/322814910/files/doc_downloads/acacia/Acacia-2.7-announcement.pdf.
-
7
The main revenue streams are import duty, skills development levy, royalty, corporate income tax, a share of dividends and shareholder loan repayments through state equity, and dividend withholding tax.
-
8
The earlier in time a shilling (Tanzania’s official currency unit) is received, the more it is worth. This is, first, because it can be used earlier; and second, because the future is uncertain, and no one can be sure they will receive that shilling in the future. To account for this time value of money, a “discount rate” is applied. In the sharing arrangement, this would mean that if the government received a shilling in year 1, the company would need to receive more than a shilling in year 2 for the benefits to be comparable. However, given cumulation here is based on actual cash flow, the company would need to receive only a shilling in year 2 for the benefits to be shared equally.
-
9
This provision for the company to earn its minimum return before sharing is triggered means Ecuador’s mechanism is similar to an R-based cash flow tax, commonly referred to as a Brown Tax. See, e.g., Robin Broadway and Michael Keen, “Theoretical perspectives on resource tax design,” in The Taxation of Petroleum and Minerals: Principles, Problems and Practice, edited by Philip Daniel, Michael Keen and Charles McPherson (Oxford: Routledge, 2010), 13–74.
-
10
Prices are taken from World Bank, “Commodities Price Data (The Pink Sheet),” www.worldbank.org/en/research/commodity-markets.
-
11
With a gold price of $1,600 per ounce.
-
12
With a gold price of $1,600 per ounce.
-
13
With a gold price of $1,600 per ounce.
-
14
The Fraser Institute survey estimates that, unless there are extremely harmful policies, around 60 percent of an investment decision tends to be based on a country’s geology. The other 40 percent comprises of several other factors, including political stability and policy predictability (given they affect the risk that investors will not be able to secure future returns generated by their investments), a conducive business environment and the tax level. See Julio Mejia and Elmira Aliakbari, Fraser Institute Annual Survey of Mining Companies 2022, (Fraser Institute, 2023), 8, www.fraserinstitute.org/studies/annual-survey-of-mining-companies-2022.
-
15
However, information gaps make it difficult for taxes to be structured to capture all excess profit. See Jean-Franҫois Wen, Progressive Taxation of Extractive Resources as Second-Best Optimal Policy (International Monetary Fund, 2018), www.imf.org/en/Publications/WP/Issues/2018/06/13/Progressive-Taxation-of-Extractive-Resources-as-Second-Best-Optimal-Policy-45923.
-
16
Recent research provides a sense of the potential revenue loss to governments from tax avoidance. The International Monetary Fund recently estimated that sub-Saharan African mining countries could be losing between $450 and $730 million in corporate income tax a year. See Sebastian Beer and Dan Devlin, Is There Money on the Table? Evidence on the Magnitude of Profit Shifting in the Extractive Industries (International Monetary Fund, 2021), www.imf.org/en/Publications/WP/Issues/2021/01/15/Is-There-Money-on-the-Table-Evidence-on-the-Magnitude-of-Profit-Shifting-in-the-Extractive-49983.
-
17
It is perhaps surprising that Tanzania’s 50-50 sharing arrangement generates an AETR greater than 50 percent (with a discount rate of 10 percent). This is despite AETR measuring government take as the share of pre-tax profits, which is larger than “economic benefits” (given economic benefits exclude interest payments). This outcome results from the 50-50 split being based on actual cash flow. The government receives revenue before the mining company through input and production taxes that do not depend on the mine making a profit. Because of these earlier revenues, the government receives a larger share on a discounted basis.
-
18
As reported in the S&P Global database. Legal risks are “expropriation, state contract alteration and contract enforcement risks.” Tax risks are “tax increase and tax inconsistency risks.” Control Risks scores these risks as still “very high” and “high” respectively (following Tanzania’s overhaul of extractives sector laws and other actions against existing mines in 2017) but reducing.
-
19
With a discount rate of 10 percent. While Ecuador’s sharing mechanism does not account for the labor profit share because none of it will go to the government from 2024 onwards, I have included it in the AETR because it is a tax on the project. The Democratic Republic of Congo regime has an excess profits tax that is triggered for a mine when the realized price is at least 25 percent higher than the price in its feasibility study. I assumed that the feasibility study has a price of $1,300 per ounce, so the excess profits tax is not triggered.
-
20
Total benefits in this case are a project’s revenues minus operating costs and replacement capital (but not minus exploration and development capital). This cash flow represents the money available to pay back the initial investment and provide a return. The government share of it is a common measure of progressivity.
-
21
Wen, Progressive Taxation of Extractive Resources as Second-Best Optimal Policy.
-
22
With a discount rate of 10 percent. The results for only some countries are shown to clearly depict each data point. The results for all the evaluated countries can be found in my model.
-
23
Scurfield, “Tanzania Strikes a Better Balance with its Mining Fiscal Regime.”
-
24
This feature is not fully reflected in Figure 5 given that “total benefits” use a slightly different definition of profits and are based on discounted cash flows.
-
25
With a gold price of $1,600 per ounce.
-
26
For example, an average 62 percent of respondents to the Fraser Institute surveys between 2017 and 2019 said the current implementation of Tanzania’s legal system would strongly discourage investment, and 73 percent said regulatory uncertainty would. See, e.g., Ashley Stedman, Jairo Yunis and Elmira Aliakbari, Fraser Institute Annual Survey of Mining Companies 2019 (Fraser Institute, 2020), www.fraserinstitute.org/studies/annual-survey-of-mining-companies-2019.
-
27
With a low-profit mine and a gold price of $1,600 per ounce.
-
28
Tax avoidance could extend the Philippines’ recovery period, and therefore delay the payment of some taxes including import duty and interest withholding tax, given the end of the recovery period depends on the reported profitability of a mine rather than an ex-ante assessment. However, the rule that the recovery period must end five years from the start of production regardless of whether pre-production expenses have been recouped limits the extent to which it can be extended.
-
29
The merits of these measures require further scrutiny. E.g., taking a share of loan repayments could result in lenders charging a higher interest rate to ensure they still recoup their loan and a minimum return. This would not only reduce taxable income but also make it harder for the government to assess whether an interest rate is reasonable, because it would not be comparable with industry benchmarks. It can also be difficult for a government to always determine whether a loan is from a related party or not. However, these considerations are outside the scope of this analysis.
-
30
Natural Resource Governance Institute (NRGI), Natural Resource Charter, 2nd edition, 2014, resourcegovernance.org/analysis-tools/publications/natural-resource-charter-2nd-ed.
-
31
Although its exclusion of several significant taxes from the government’s share of benefits means low-profit mines may still be impacted.
-
32
With a gold price of $1,600 per ounce.
-
33
Anna Fleming, Thomas Lassourd and David Manley, “Variable Royalties: an Answer to Volatile Mineral Prices?” in Handbook on the Future of Resource Taxation, African Tax Administration Forum and Intergovernmental Forum on Mining, Minerals, Metals and Sustainable Development (forthcoming), www.iisd.org/publications/brief/future-resource-taxation-roadmap.
-
34
Ensuring that interest rates used as a comparison apply to comparable assets with a similar risk profile is challenging, but rules of this nature have been successful in reducing profit shifting in other countries. See Beer and Devlin, Is There Money on the Table?.
-
35
The main taxes listed are VAT, royalty and corporate income tax. The regime also includes a share of pre-tax profits that is currently divided between the company’s workers and the government, with the portion received by the government included in its accumulated benefits. However, a recent court ruling means that all this labor profit share will go to workers from the start of 2024 and therefore none of it will be included in government benefits.
-
36
The discount rate used is specific to each mine and based on its weighted average cost of capital (WACC). I have assumed that WACC is around 7 percent in real terms. This is based on the typical discount rate for equity shareholders used by industry and government analysts of 8 percent in real terms, and the current average cost of debt for the mining sector as reported by Aswath Damodaran, Damodaran Online, www.pages.stern.nyu.edu/~adamodar.
-
37
Republic of the Philippines, Financial or Technical Assistance Agreement, mgb.gov.ph/attachments/article/79/PFC_FTAA.pdf.
-
38
For example, some terms in the original FTAA for an OceanaGold mine differed in some areas: Republic of the Philippines, Financial or Technical Assistance Agreement with Arimco Mining Corporation, 1994, www.resourcecontracts.org/contract/ocds-591adf-2792396017. I understand that the recently signed extension to this agreement has slightly different terms again.
-
39
A template of the FARI model and a user guide that explains all the concepts and workings of the model are available at International Monetary Fund, “Fiscal Analysis of Resource Industries,”www.imf.org/external/np/fad/fari.
Politics Are Holding Back Implementation of Extractive Sector Transparency Rules
“Our country has good rules, it’s the implementation that fails.” This is a common refrain we hear from partners working on natural resource governance, ranging from civil society actors pushing for contract transparency in the Democratic Republic of Congo (DRC) to those advocating for improved state-owned enterprise (SOE) governance in Mongolia.
We know good laws and regulations aren’t enough to achieve good governance; the ability and willingness to follow through with them is key. So why do so many resource-rich countries fail to adequately implement their “good rules”?
Two decades of experience from the transparency, participation and accountability (TPA) field has led to a growing appreciation of the need to better integrate politics. As detailed in a recent discussion paper for the Executive Session on the Politics of Extractive Industries, this applies throughout the extractives “transparency lifecycle.” This resonates with our experience that political considerations can make or break the implementation of extractive sector transparency commitments.
At the Natural Resource Governance Institute (NRGI) we’re increasingly considering the “implementation gap.” We define this gap as the difference between adopted rules and their implementation in practice. The 2017 Resource Governance Index (RGI) started measuring this gap, finding that the worse a country’s overall sector governance, the wider its implementation gap. The gap was also a focus of a subsequent sub-Saharan Africa analysis. That report found that of 22 countries that had revised their sector laws between 2000 and 2016, all but two exhibited an implementation gap, keeping them from realizing benefits from the investment in legal reforms.
Identifying and measuring implementation gaps remains central to the RGI, with early findings from the 2021 edition showing both progress in addressing gaps (e.g., in Guinea) and areas where gaps are particularly pronounced (e.g., around local impacts in Peru).
So, what are the political causes of—and potential responses to—implementation gaps? Here we draw on NRGI’s country experience to answer these questions and consider the implications for NRGI and others committed to sector policy reform.
Political causes of implementation gaps
Our experience and a review of relevant literature points to two broad interrelated categories of causes for gaps: (i) political factors, and (ii) administrative constraints and shortcomings in rulemaking processes. Even for the second category, the underlying causes may be primarily political. For example, choices by parliament or ministers or competition between ministries and SOEs can result in a lack of resources or capacity within institutions tasked with implementation or oversight, or insufficient coordination between policymakers and implementers.
Within the political factors category, we have identified a few specific causes behind implementation gaps. A common cause is resistance by different governmental actors, to protect political or economic interests (including potentially private or individual interests), often through parallel governance structures. In many countries, there are unclear or overlapping governance structures both horizontally—between parallel agencies such as ministries and SOEs —and vertically—between national and local agencies. In these situations, even if the law assigns implementation responsibility to a specific agency, this may not always correspond to political realities and power dynamics, with other actors potentially blocking implementation.
In the extractive sector, powerful SOEs can play an important role in this dynamic. For example, in the DRC, implementation of the country’s contract transparency rules has been particularly challenging for contracts agreed by SOEs. The DRC’s mining SOEs play a critical de facto role in licensing, and international actors have investigated transactions to which they are party, potentially contributing to an aversion to transparency among SOE decisionmakers. While the law charges a technical unit of the Ministry of Mines with implementing contract disclosure, in practice this unit has at times struggled to access documents from SOEs, which answer to a different ministry.
Another common political factor behind implementation gaps is a decline in prominence on the political agenda of an issue, which disincentivizes political leaders (and the individuals and institutions who answer to them) from implementing related rules. Once a law is passed, momentum behind reforms often fades, as political gains in many instances derive from the passage of laws rather than implementation. This aligns with a broader dynamic identified in the discussion paper: where political leaders’ chief motivation for adopting a policy is reputational gain, preparing for elections or responding to scandal or political pressure, implementation can be weak or fleeting. Such dynamics have led to countries signing up to transparency commitments such as the Extractive Industries Transparency Initiative (EITI) for optics reasons, but then demonstrating limited appetite to significantly implement the commitments.
This echoes what we have heard from civil society activists and journalists in Tanzania, namely that, under the previous government, the ruling party promised during elections to advance transparency to appease citizen pressure and to respond to international transparency norms. While this led to passage of the Tanzania Extractive Industries (Transparency and Accountability) Act in 2015, endorsing several transparency requirements, the government viewed transparency as an agenda driven by the opposition, clashing with its focus on infrastructure development and industrialization. After the 2015 elections, authorities showed little interest in implementing the law.
Competing sources of norms also contribute to implementation gaps. International transparency norms, generally championed by international and national civil society organizations, sometimes fail to connect with the approaches and priorities of community-based organizations, which tend to prioritize more direct impacts (e.g., jobs, environmental damage, use of land) or of regional bodies (e.g., African Union), which tend to focus on industrialization and structural transformation.
These disconnects can weaken support for and oversight of implementation. For example, in Tanzania, national civil society actors felt that the officials in the former government portrayed international transparency norms as externally driven, costly and not a high national priority. These dynamics may be partly why some donors’ approaches are shifting to focus on prioritizing local work, and ensuring that that work informs global norms instead of working from the global to the local level.
Mind (or better yet close) the gap
What does this mean for an organization like NRGI and others working on extractive sector governance? We set out below some initial guiding thoughts:
Consider implementation early and often. As a first step we must factor implementation—and the risk of gaps and their political causes—into various parts of our work, including when selecting where and how we work, who we work with, and the content of our analysis and recommendations.
Ultimately, governments are responsible for implementation, but other actors such as international development agencies, donors, technical advisors, regional bodies, and civil society organizations can play a role in avoiding implementation gaps in the first place as well. This includes by identifying (and pushing for rectification of) shortcomings in policy and legal designs that can hinder implementation. The same actors can also look to support potential responses to implementation blockages, including by strengthening oversight and demand for implementation.
International development agencies, donors, technical advisors, regional bodies, and civil society organizations play a key part in shaping laws aimed at realizing extractive sector transparency. But too often these actors are driven by—and rewarded for—securing the establishment and formal adoption of norms. Civil society organizations must take care to not declare victory when a law is passed and not move on too quickly to campaigning for the next norm. Similarly, development agencies, donors, and governments should ensure that when linking reforms to incentives (e.g., financing), they prioritize effective implementation and continued compliance, and not just initial commitment.
Development partners and civil society organizations should embed implementation considerations in their technical assistance and advocacy, in both a political and a technical sense. On the technical side, they should push for legislation that clearly specifies the actor responsible for implementation, includes a budget and timeframe for implementation, and avoids conflict between various laws or excessive fragmentation between different implementing actors. More broadly, both national NGOs and international norm setters should advocate for practical and context-appropriate rules, sensitive to the trade-off between “best practice” and “implementability.”
Understanding the incentives for and against implementation. International development actors, including NRGI, should consult political agendas and key constituencies such as implementing agencies early on to gauge their incentives for (or disincentives against) implementation. Actors should also look to align their advocacy with existing and upcoming official technical assistance programs endorsed by the government where possible. In Mongolia for example, advocacy by NRGI and Mongolian civil society actors for implementing the greater SOE transparency set out in the Glass Accounts Law was supported by oversight actors such as auditors and the political agendas of certain social movements and political parties. Coordinating this advocacy with official technical assistance to SOEs by the Asian Development Bank further helped to address the implementation gap.
Boost demand for implementation. As with the Mongolia example, oversight actors demanding implementation play an important role, especially where government incentives for implementation are weak. The influential 2012 Global Integrity and CIPE report on implementation gaps put it well; the “implementation gap will be eradicated only when the supply of sound laws is met by corresponding demand on the ground to implement them.” NRGI aims to complement our legal reform and advocacy work by partnering with organizations that actively monitor implementation after laws are passed. In the DRC, we partnered with local CSOs to produce lists of non-disclosed contracts and submit them to government and international partners. In Ghana, NRGI followed its technical assistance and advocacy around the country’s petroleum law with the convening of a civil society group to monitor the country’s first oil bid round, with a focus on implementation of transparency rules. Civic space is vital to such efforts. Without freedom of speech, citizens are reluctant to speak out against lack of implementation. Hence the importance of initiatives to preserve this space, such as that of EITI.
Different actors have different roles to play around the implementation gap. Larger international institutions can contribute technical and financial resources in more politically informed ways to promote implementation, while national actors are better placed to influence underlying political incentives. At NRGI we aim both to lead on closing the implementation gap in natural resource governance and to further sensitize other actors, including our partners in civil society and government, to this critical issue.
Amir Shafaie is the legal and economic programs director at the Natural Resource Governance Institute (NRGI). Moses Kulaba is the East Africa manager at NRGI. Kaisa Toroskainen is a former Africa program officer at NRGI.
This post also appeared on the CCSI blog.
Notes
-
1
With a gold price of USD 1,600 per ounce.
-
2
With a low-profit mine and a gold price of $1,600 per ounce.
-
3
Cecilia Jamasmie, “Petra Diamonds’ stake in Williamson to shrink as part of deal with Tanzania,” Mining.com, 13 December 2021, www.mining.com/petra-diamonds-stake-in-williamson-to-shrink-as-part-of-deal-with-tanzania.
-
4
Lifezone Metals, “Kabanga Nickel Signs Framework Agreement,” 19 January 2021, www.lifezonemetals.com/kabanga-nickel-signs-framework-agreement.
-
5
Thomas Scurfield and Silas Olan’g, “Magufuli Seeks the Right Balance for Tanzania’s Mining Fiscal Regime,” NRGI, 31 January 2019, www.resourcegovernance.org/blog/magufuli-seeks-right-balance-tanzania-mining-fiscal; Thomas Scurfield, “Tanzania Strikes a Better Balance with its Mining Fiscal Regime,” NRGI, 24 June 2020, www.resourcegovernance.org/blog/tanzania-strikes-better-balance-mining-fiscal-regime.
-
6
This framework agreement was published in a document setting out Barrick’s offer to buy the shares it did not already own in Acacia Mining, the previous owner of the Bulyanhulu, Buzwagi and North Mara mines in Tanzania. See Acacia Mining and Barrick Gold, Recommended Final Offer for Acacia Mining Plc by Barrick Gold Corporation, 2019, 66–79, s25.q4cdn.com/322814910/files/doc_downloads/acacia/Acacia-2.7-announcement.pdf.
-
7
The main revenue streams are import duty, skills development levy, royalty, corporate income tax, a share of dividends and shareholder loan repayments through state equity, and dividend withholding tax.
-
8
The earlier in time a shilling (Tanzania’s official currency unit) is received, the more it is worth. This is, first, because it can be used earlier; and second, because the future is uncertain, and no one can be sure they will receive that shilling in the future. To account for this time value of money, a “discount rate” is applied. In the sharing arrangement, this would mean that if the government received a shilling in year 1, the company would need to receive more than a shilling in year 2 for the benefits to be comparable. However, given cumulation here is based on actual cash flow, the company would need to receive only a shilling in year 2 for the benefits to be shared equally.
-
9
This provision for the company to earn its minimum return before sharing is triggered means Ecuador’s mechanism is similar to an R-based cash flow tax, commonly referred to as a Brown Tax. See, e.g., Robin Broadway and Michael Keen, “Theoretical perspectives on resource tax design,” in The Taxation of Petroleum and Minerals: Principles, Problems and Practice, edited by Philip Daniel, Michael Keen and Charles McPherson (Oxford: Routledge, 2010), 13–74.
-
10
Prices are taken from World Bank, “Commodities Price Data (The Pink Sheet),” www.worldbank.org/en/research/commodity-markets.
-
11
With a gold price of $1,600 per ounce.
-
12
With a gold price of $1,600 per ounce.
-
13
With a gold price of $1,600 per ounce.
-
14
The Fraser Institute survey estimates that, unless there are extremely harmful policies, around 60 percent of an investment decision tends to be based on a country’s geology. The other 40 percent comprises of several other factors, including political stability and policy predictability (given they affect the risk that investors will not be able to secure future returns generated by their investments), a conducive business environment and the tax level. See Julio Mejia and Elmira Aliakbari, Fraser Institute Annual Survey of Mining Companies 2022, (Fraser Institute, 2023), 8, www.fraserinstitute.org/studies/annual-survey-of-mining-companies-2022.
-
15
However, information gaps make it difficult for taxes to be structured to capture all excess profit. See Jean-Franҫois Wen, Progressive Taxation of Extractive Resources as Second-Best Optimal Policy (International Monetary Fund, 2018), www.imf.org/en/Publications/WP/Issues/2018/06/13/Progressive-Taxation-of-Extractive-Resources-as-Second-Best-Optimal-Policy-45923.
-
16
Recent research provides a sense of the potential revenue loss to governments from tax avoidance. The International Monetary Fund recently estimated that sub-Saharan African mining countries could be losing between $450 and $730 million in corporate income tax a year. See Sebastian Beer and Dan Devlin, Is There Money on the Table? Evidence on the Magnitude of Profit Shifting in the Extractive Industries (International Monetary Fund, 2021), www.imf.org/en/Publications/WP/Issues/2021/01/15/Is-There-Money-on-the-Table-Evidence-on-the-Magnitude-of-Profit-Shifting-in-the-Extractive-49983.
-
17
It is perhaps surprising that Tanzania’s 50-50 sharing arrangement generates an AETR greater than 50 percent (with a discount rate of 10 percent). This is despite AETR measuring government take as the share of pre-tax profits, which is larger than “economic benefits” (given economic benefits exclude interest payments). This outcome results from the 50-50 split being based on actual cash flow. The government receives revenue before the mining company through input and production taxes that do not depend on the mine making a profit. Because of these earlier revenues, the government receives a larger share on a discounted basis.
-
18
As reported in the S&P Global database. Legal risks are “expropriation, state contract alteration and contract enforcement risks.” Tax risks are “tax increase and tax inconsistency risks.” Control Risks scores these risks as still “very high” and “high” respectively (following Tanzania’s overhaul of extractives sector laws and other actions against existing mines in 2017) but reducing.
-
19
With a discount rate of 10 percent. While Ecuador’s sharing mechanism does not account for the labor profit share because none of it will go to the government from 2024 onwards, I have included it in the AETR because it is a tax on the project. The Democratic Republic of Congo regime has an excess profits tax that is triggered for a mine when the realized price is at least 25 percent higher than the price in its feasibility study. I assumed that the feasibility study has a price of $1,300 per ounce, so the excess profits tax is not triggered.
-
20
Total benefits in this case are a project’s revenues minus operating costs and replacement capital (but not minus exploration and development capital). This cash flow represents the money available to pay back the initial investment and provide a return. The government share of it is a common measure of progressivity.
-
21
Wen, Progressive Taxation of Extractive Resources as Second-Best Optimal Policy.
-
22
With a discount rate of 10 percent. The results for only some countries are shown to clearly depict each data point. The results for all the evaluated countries can be found in my model.
-
23
Scurfield, “Tanzania Strikes a Better Balance with its Mining Fiscal Regime.”
-
24
This feature is not fully reflected in Figure 5 given that “total benefits” use a slightly different definition of profits and are based on discounted cash flows.
-
25
With a gold price of $1,600 per ounce.
-
26
For example, an average 62 percent of respondents to the Fraser Institute surveys between 2017 and 2019 said the current implementation of Tanzania’s legal system would strongly discourage investment, and 73 percent said regulatory uncertainty would. See, e.g., Ashley Stedman, Jairo Yunis and Elmira Aliakbari, Fraser Institute Annual Survey of Mining Companies 2019 (Fraser Institute, 2020), www.fraserinstitute.org/studies/annual-survey-of-mining-companies-2019.
-
27
With a low-profit mine and a gold price of $1,600 per ounce.
-
28
Tax avoidance could extend the Philippines’ recovery period, and therefore delay the payment of some taxes including import duty and interest withholding tax, given the end of the recovery period depends on the reported profitability of a mine rather than an ex-ante assessment. However, the rule that the recovery period must end five years from the start of production regardless of whether pre-production expenses have been recouped limits the extent to which it can be extended.
-
29
The merits of these measures require further scrutiny. E.g., taking a share of loan repayments could result in lenders charging a higher interest rate to ensure they still recoup their loan and a minimum return. This would not only reduce taxable income but also make it harder for the government to assess whether an interest rate is reasonable, because it would not be comparable with industry benchmarks. It can also be difficult for a government to always determine whether a loan is from a related party or not. However, these considerations are outside the scope of this analysis.
-
30
Natural Resource Governance Institute (NRGI), Natural Resource Charter, 2nd edition, 2014, resourcegovernance.org/analysis-tools/publications/natural-resource-charter-2nd-ed.
-
31
Although its exclusion of several significant taxes from the government’s share of benefits means low-profit mines may still be impacted.
-
32
With a gold price of $1,600 per ounce.
-
33
Anna Fleming, Thomas Lassourd and David Manley, “Variable Royalties: an Answer to Volatile Mineral Prices?” in Handbook on the Future of Resource Taxation, African Tax Administration Forum and Intergovernmental Forum on Mining, Minerals, Metals and Sustainable Development (forthcoming), www.iisd.org/publications/brief/future-resource-taxation-roadmap.
-
34
Ensuring that interest rates used as a comparison apply to comparable assets with a similar risk profile is challenging, but rules of this nature have been successful in reducing profit shifting in other countries. See Beer and Devlin, Is There Money on the Table?.
-
35
The main taxes listed are VAT, royalty and corporate income tax. The regime also includes a share of pre-tax profits that is currently divided between the company’s workers and the government, with the portion received by the government included in its accumulated benefits. However, a recent court ruling means that all this labor profit share will go to workers from the start of 2024 and therefore none of it will be included in government benefits.
-
36
The discount rate used is specific to each mine and based on its weighted average cost of capital (WACC). I have assumed that WACC is around 7 percent in real terms. This is based on the typical discount rate for equity shareholders used by industry and government analysts of 8 percent in real terms, and the current average cost of debt for the mining sector as reported by Aswath Damodaran, Damodaran Online, www.pages.stern.nyu.edu/~adamodar.
-
37
Republic of the Philippines, Financial or Technical Assistance Agreement, mgb.gov.ph/attachments/article/79/PFC_FTAA.pdf.
-
38
For example, some terms in the original FTAA for an OceanaGold mine differed in some areas: Republic of the Philippines, Financial or Technical Assistance Agreement with Arimco Mining Corporation, 1994, www.resourcecontracts.org/contract/ocds-591adf-2792396017. I understand that the recently signed extension to this agreement has slightly different terms again.
-
39
A template of the FARI model and a user guide that explains all the concepts and workings of the model are available at International Monetary Fund, “Fiscal Analysis of Resource Industries,”www.imf.org/external/np/fad/fari.
What Ghana’s 2021 Budget Reveals About Its Oil Fund Management
Ghana’s 2021 budget, titled Economic Revitalization through Completion, Consolidation and Continuity, was widely anticipated. Ghanaians hoped that the budget would include steps to address both the immense challenges brought on by the coronavirus pandemic and low revenues resulting from a downturn in crude oil prices. Revenue shortfall amounted to GH₵13.6 billion while expenditures rose by an estimated GH₵11.7 billion in 2020; this has led to more uncertainty about the future of Ghana’s recovery plans.
In addition, as in many other African economies, debt also threatens the projected economic recovery in Ghana, something the government has acknowledged in the 2021 budget.
Commodities are expected to play an important role in addressing Ghana’s debt challenges. Petroleum exports will likely decline in the medium term , from a peak of 71 million barrels in 2019 to about 59 million barrels in 2024, and so will revenues. Though crude oil prices have recovered from the depths of the second quarter of 2020, there is no guarantee that these will be sustained, thus limiting the role of petroleum revenues in addressing Ghana’s debt burden. Nonetheless, petroleum is still an important source of revenue and a likely part of the solution to Ghana’s public financial challenges.
Unresolved governance issues
Despite the 2021 budget’s attempt to address concerns around petroleum revenue spending in Ghana, it fails to address key governance questions around what will happen to unspent revenue and what constitutes public investment, among other issues.
Allocation for public investment
Public investment is generally understood to be spending on assets or infrastructure with long lifespans. But the lack of a clear explanation of Ghana’s public investment expenditures has over time led to various approaches by the Ministry of Finance on where to allocate the funds, leading to disagreements among different stakeholders. For example, between 2011 and 2016, 70 percent of oil revenue designated for budget support (the Annual Budget Funding Amount, or ABFA) was allocated to capital spending, which includes roads, agriculture, education and health infrastructure, against 30 percent allocated to recurrent expenditure, which typically covers payments for services and supplies. The Petroleum Revenue Management Act (PRMA) Section 21(4) states: "For any financial year, a minimum of seventy percent of the Annual Budget Funding Amount shall be used for public investment expenditures consistent with the long-term national development plan or with subsection (3)."
In 2017, the government prioritized spending of oil revenues on “physical infrastructure and service delivery in education,” in line with the PRMA. Civil society organizations raised concerns about whether this spending was consistent with the Ministry of Finance’s previous interpretation, which had restricted public investment expenditure to capital spending Between 2017 and 2019, recurrent expenditure on the Free Senior High School Program, which included the payment of fees and supplies, represented an average of 52 percent of ABFA. The Public Interest and Accountability Committee (PIAC) called on the government on numerous occasions to adhere to the Ministry of Finance’s previous interpretation of the PRMA but the government argued that investment in education qualifies as public investment expenditure and that is not entirely synonymous with capital expenditure, with an explanation included in the 2021 budget.
This changing interpretation of “public investment expenditure” has led to recurrent expenditures for which volatile petroleum revenues are ill-suited. Previous NRGI research has emphasized the challenges with spending volatile resource revenues on recurrent expenditure. There is broad consensus that public investment in key sectors can enhance growth. Ghana’s government should clarify the definition of public investment expenditure to refer to only non-recurrent expenditure in the PRMA Regulations (L.I. 2381) or in a PRMA amendment. This would also encourage consistent implementation of the rule by future ministers of finance.
Unspent petroleum revenues at the end of the year
The use of unspent petroleum revenues at the end of the budget year is another unsolved issue. Successive governments have managed unspent ABFA differently since this first came up in 2014. That year a total of USD 222.93 million (GH₵666.06 million) of ABFA was unspent. The amount, according to the Ministry of Finance, was returned to the Bank of Ghana at the end of that year as the bank assessed the government’s financial position. However, in 2015, the Ministry of Finance did not disclose of how it applied a balance of GH₵36.41 million. The following year’s Ministry of Finance reconciliation report made no mention of the unused balance, as it should have. Similarly, the Ministry did not account for the unused balance of GH₵77.73 million at the end of 2016 in its report to the parliament.
Treatment of Unused ABFA, (2014-2020) (GH₵ millions)
Click to enlarge
Source: Ghana Ministry of Finance (Reconciliation Reports, 2014-2016; 2021 Budget Statement)
In 2020, the Ministry of Finance brought forward and spent unused ABFA from 2017-2019 (amounting to GH₵1.48 billion; see figure above) alongside allocations for 2020. Since 2018, the PIAC and other civil society actors have called upon Ghana’s government to provide details of the unspent revenues; these calls only elicited a public response in the heat of the campaign leading up to the December 2020 elections. According to the ministry, the combined 2017 and 2018 amount of GH₵652.29 million was transferred to the Road Fund while 2019’s GH₵827.60 million was used to meet the 2020 ABFA shortfall due to the pandemic. It is unclear which provisions, if any, in the PRMA support the transfer of unspent ABFA into the Road Fund or to meet ABFA shortfalls in subsequent years. Also unclear is which year’s appropriation (budget)—which provides the legal authorization to spend based on agreed activities—covered the spending of these unused revenues.
Additional challenges to the financial management of Ghana’s public sector include delays in the contracting process for projects, disbursement of funds and contract execution. Issues related to unspent ABFA will continue unless these challenges are resolved. Dialogue and consensus building among stakeholders on how to treat unspent funds in the future would avoid various conflicting interpretations arising between provisions in the Public Financial Management Act and the PRMA.
While the PRMA does not prescribe how unused ABFA at the end of the year should be treated, Section 21(1) provides that it must be part of the national budget and subject to same budgetary processes, bringing the ABFA under the public financial management rules in terms of management and accounting. And Section 26 of the Public Financial Management Act states clearly that such appropriations lapse by the end of the first month of the following year, meaning that the government needs a new appropriation from the parliament to use these funds.
Section 26 of the Public Financial Management Act, 2016

Inability to spend within the appropriation period, for whatever reason, creates public financial management complications in terms of unspent balances, re-budgeting and project execution. With petroleum funds, unspent funds make tracking and accounting for revenues within the framework of the PRMA provisions difficult.
Ghana’s government must urgently tackle the underlying issues of public financial management and delays in projects on which the money is spent. In addition, stakeholders (the Ministry of Finance and ministries that receive ABFA; oversight bodies like PIAC and Ghana EITI; and civil society actors) must engage in order to answer the following questions and possibly factor them into future amendments of the PRMA:
- What should happen to unused ABFA and where should these unused funds be saved?
- Should new appropriation in accordance with the Public Financial Management Act allow for a change in projects and programs earmarked in a previous appropriation?
- Does government ringfence ABFA funds, and if not, should they?
Discretionary capping of the Ghana Stabilization Fund
The Petroleum Revenue Management Regulations (L.I. 2381), passed in 2019, sought to address the situation where the minister of finance sets a discretionary cap on the Ghana Stabilization Fund by defining a formula for setting the cap. Regulation 8(1) states “In pursuance of Section 23 of the Act, the Minister shall, in recommending the maximum amount of accumulated resources in the Ghana Stabilisation Fund, ensure that the amount is not less than the average Annual Budget Funding Amount over a three year period.” While this provides guidance for setting a maximum balance on the fund, no minimum amount is specified; this could leave little money in the fund for its stated purpose of addressing revenue volatilities or debt. Though the 2021 budget did not propose a cap, the implementation of L.I. 2381 will mean that the Ghana Stabilization Fund would likely not grow beyond USD 400 million in the medium term (based on petroleum revenue projections in the 2021 budget). While a lower ceiling on the Ghana Stabilization Fund would make funds available for real-time debt servicing, it reduces the opportunity for government to rely on the fund to address shocks to the economy, fund future budgets and service future debt.
Ways forward
Resource revenues and good public financial management have the potential to help address Ghana’s financing and borrowing needs, and its economic recovery plans. To ensure this potential is realized:
- The government should adhere to the generally accepted interpretation of public investment as referring to non-recurrent spending, and clarify this in the Petroleum Revenue Management Regulations (L.I. 2381) to avoid using volatile revenue streams for recurrent expenditures.
- The Ministry of Finance should initiate consultations on an amendment of the PRM regulations that would provide clarity on how government should manage unspent revenues.
- The government should undertake a review of the capping mechanism of the stabilization fund.
Denis Gyeyir is an Accra-based Africa program officer with the Natural Resource Governance Institute (NRGI).
Notes
-
1
With a gold price of USD 1,600 per ounce.
-
2
With a low-profit mine and a gold price of $1,600 per ounce.
-
3
Cecilia Jamasmie, “Petra Diamonds’ stake in Williamson to shrink as part of deal with Tanzania,” Mining.com, 13 December 2021, www.mining.com/petra-diamonds-stake-in-williamson-to-shrink-as-part-of-deal-with-tanzania.
-
4
Lifezone Metals, “Kabanga Nickel Signs Framework Agreement,” 19 January 2021, www.lifezonemetals.com/kabanga-nickel-signs-framework-agreement.
-
5
Thomas Scurfield and Silas Olan’g, “Magufuli Seeks the Right Balance for Tanzania’s Mining Fiscal Regime,” NRGI, 31 January 2019, www.resourcegovernance.org/blog/magufuli-seeks-right-balance-tanzania-mining-fiscal; Thomas Scurfield, “Tanzania Strikes a Better Balance with its Mining Fiscal Regime,” NRGI, 24 June 2020, www.resourcegovernance.org/blog/tanzania-strikes-better-balance-mining-fiscal-regime.
-
6
This framework agreement was published in a document setting out Barrick’s offer to buy the shares it did not already own in Acacia Mining, the previous owner of the Bulyanhulu, Buzwagi and North Mara mines in Tanzania. See Acacia Mining and Barrick Gold, Recommended Final Offer for Acacia Mining Plc by Barrick Gold Corporation, 2019, 66–79, s25.q4cdn.com/322814910/files/doc_downloads/acacia/Acacia-2.7-announcement.pdf.
-
7
The main revenue streams are import duty, skills development levy, royalty, corporate income tax, a share of dividends and shareholder loan repayments through state equity, and dividend withholding tax.
-
8
The earlier in time a shilling (Tanzania’s official currency unit) is received, the more it is worth. This is, first, because it can be used earlier; and second, because the future is uncertain, and no one can be sure they will receive that shilling in the future. To account for this time value of money, a “discount rate” is applied. In the sharing arrangement, this would mean that if the government received a shilling in year 1, the company would need to receive more than a shilling in year 2 for the benefits to be comparable. However, given cumulation here is based on actual cash flow, the company would need to receive only a shilling in year 2 for the benefits to be shared equally.
-
9
This provision for the company to earn its minimum return before sharing is triggered means Ecuador’s mechanism is similar to an R-based cash flow tax, commonly referred to as a Brown Tax. See, e.g., Robin Broadway and Michael Keen, “Theoretical perspectives on resource tax design,” in The Taxation of Petroleum and Minerals: Principles, Problems and Practice, edited by Philip Daniel, Michael Keen and Charles McPherson (Oxford: Routledge, 2010), 13–74.
-
10
Prices are taken from World Bank, “Commodities Price Data (The Pink Sheet),” www.worldbank.org/en/research/commodity-markets.
-
11
With a gold price of $1,600 per ounce.
-
12
With a gold price of $1,600 per ounce.
-
13
With a gold price of $1,600 per ounce.
-
14
The Fraser Institute survey estimates that, unless there are extremely harmful policies, around 60 percent of an investment decision tends to be based on a country’s geology. The other 40 percent comprises of several other factors, including political stability and policy predictability (given they affect the risk that investors will not be able to secure future returns generated by their investments), a conducive business environment and the tax level. See Julio Mejia and Elmira Aliakbari, Fraser Institute Annual Survey of Mining Companies 2022, (Fraser Institute, 2023), 8, www.fraserinstitute.org/studies/annual-survey-of-mining-companies-2022.
-
15
However, information gaps make it difficult for taxes to be structured to capture all excess profit. See Jean-Franҫois Wen, Progressive Taxation of Extractive Resources as Second-Best Optimal Policy (International Monetary Fund, 2018), www.imf.org/en/Publications/WP/Issues/2018/06/13/Progressive-Taxation-of-Extractive-Resources-as-Second-Best-Optimal-Policy-45923.
-
16
Recent research provides a sense of the potential revenue loss to governments from tax avoidance. The International Monetary Fund recently estimated that sub-Saharan African mining countries could be losing between $450 and $730 million in corporate income tax a year. See Sebastian Beer and Dan Devlin, Is There Money on the Table? Evidence on the Magnitude of Profit Shifting in the Extractive Industries (International Monetary Fund, 2021), www.imf.org/en/Publications/WP/Issues/2021/01/15/Is-There-Money-on-the-Table-Evidence-on-the-Magnitude-of-Profit-Shifting-in-the-Extractive-49983.
-
17
It is perhaps surprising that Tanzania’s 50-50 sharing arrangement generates an AETR greater than 50 percent (with a discount rate of 10 percent). This is despite AETR measuring government take as the share of pre-tax profits, which is larger than “economic benefits” (given economic benefits exclude interest payments). This outcome results from the 50-50 split being based on actual cash flow. The government receives revenue before the mining company through input and production taxes that do not depend on the mine making a profit. Because of these earlier revenues, the government receives a larger share on a discounted basis.
-
18
As reported in the S&P Global database. Legal risks are “expropriation, state contract alteration and contract enforcement risks.” Tax risks are “tax increase and tax inconsistency risks.” Control Risks scores these risks as still “very high” and “high” respectively (following Tanzania’s overhaul of extractives sector laws and other actions against existing mines in 2017) but reducing.
-
19
With a discount rate of 10 percent. While Ecuador’s sharing mechanism does not account for the labor profit share because none of it will go to the government from 2024 onwards, I have included it in the AETR because it is a tax on the project. The Democratic Republic of Congo regime has an excess profits tax that is triggered for a mine when the realized price is at least 25 percent higher than the price in its feasibility study. I assumed that the feasibility study has a price of $1,300 per ounce, so the excess profits tax is not triggered.
-
20
Total benefits in this case are a project’s revenues minus operating costs and replacement capital (but not minus exploration and development capital). This cash flow represents the money available to pay back the initial investment and provide a return. The government share of it is a common measure of progressivity.
-
21
Wen, Progressive Taxation of Extractive Resources as Second-Best Optimal Policy.
-
22
With a discount rate of 10 percent. The results for only some countries are shown to clearly depict each data point. The results for all the evaluated countries can be found in my model.
-
23
Scurfield, “Tanzania Strikes a Better Balance with its Mining Fiscal Regime.”
-
24
This feature is not fully reflected in Figure 5 given that “total benefits” use a slightly different definition of profits and are based on discounted cash flows.
-
25
With a gold price of $1,600 per ounce.
-
26
For example, an average 62 percent of respondents to the Fraser Institute surveys between 2017 and 2019 said the current implementation of Tanzania’s legal system would strongly discourage investment, and 73 percent said regulatory uncertainty would. See, e.g., Ashley Stedman, Jairo Yunis and Elmira Aliakbari, Fraser Institute Annual Survey of Mining Companies 2019 (Fraser Institute, 2020), www.fraserinstitute.org/studies/annual-survey-of-mining-companies-2019.
-
27
With a low-profit mine and a gold price of $1,600 per ounce.
-
28
Tax avoidance could extend the Philippines’ recovery period, and therefore delay the payment of some taxes including import duty and interest withholding tax, given the end of the recovery period depends on the reported profitability of a mine rather than an ex-ante assessment. However, the rule that the recovery period must end five years from the start of production regardless of whether pre-production expenses have been recouped limits the extent to which it can be extended.
-
29
The merits of these measures require further scrutiny. E.g., taking a share of loan repayments could result in lenders charging a higher interest rate to ensure they still recoup their loan and a minimum return. This would not only reduce taxable income but also make it harder for the government to assess whether an interest rate is reasonable, because it would not be comparable with industry benchmarks. It can also be difficult for a government to always determine whether a loan is from a related party or not. However, these considerations are outside the scope of this analysis.
-
30
Natural Resource Governance Institute (NRGI), Natural Resource Charter, 2nd edition, 2014, resourcegovernance.org/analysis-tools/publications/natural-resource-charter-2nd-ed.
-
31
Although its exclusion of several significant taxes from the government’s share of benefits means low-profit mines may still be impacted.
-
32
With a gold price of $1,600 per ounce.
-
33
Anna Fleming, Thomas Lassourd and David Manley, “Variable Royalties: an Answer to Volatile Mineral Prices?” in Handbook on the Future of Resource Taxation, African Tax Administration Forum and Intergovernmental Forum on Mining, Minerals, Metals and Sustainable Development (forthcoming), www.iisd.org/publications/brief/future-resource-taxation-roadmap.
-
34
Ensuring that interest rates used as a comparison apply to comparable assets with a similar risk profile is challenging, but rules of this nature have been successful in reducing profit shifting in other countries. See Beer and Devlin, Is There Money on the Table?.
-
35
The main taxes listed are VAT, royalty and corporate income tax. The regime also includes a share of pre-tax profits that is currently divided between the company’s workers and the government, with the portion received by the government included in its accumulated benefits. However, a recent court ruling means that all this labor profit share will go to workers from the start of 2024 and therefore none of it will be included in government benefits.
-
36
The discount rate used is specific to each mine and based on its weighted average cost of capital (WACC). I have assumed that WACC is around 7 percent in real terms. This is based on the typical discount rate for equity shareholders used by industry and government analysts of 8 percent in real terms, and the current average cost of debt for the mining sector as reported by Aswath Damodaran, Damodaran Online, www.pages.stern.nyu.edu/~adamodar.
-
37
Republic of the Philippines, Financial or Technical Assistance Agreement, mgb.gov.ph/attachments/article/79/PFC_FTAA.pdf.
-
38
For example, some terms in the original FTAA for an OceanaGold mine differed in some areas: Republic of the Philippines, Financial or Technical Assistance Agreement with Arimco Mining Corporation, 1994, www.resourcecontracts.org/contract/ocds-591adf-2792396017. I understand that the recently signed extension to this agreement has slightly different terms again.
-
39
A template of the FARI model and a user guide that explains all the concepts and workings of the model are available at International Monetary Fund, “Fiscal Analysis of Resource Industries,”www.imf.org/external/np/fad/fari.
Disclosures Pave the Way for Transparency in Resource-Backed Loans
During the commodity boom of the last two decades, many governments and state-owned enterprises in developing countries signed significant “resource-backed loans” (RBLs). These loans provide governments access to finance in exchange for, or collateralized by, future streams of income from natural resource wealth, such as oil and minerals; Chinese entities are the dominant lenders in this space. In NRGI’s 2020 study on RBLs, which explored the experience of RBL borrowers in sub-Saharan Africa and Latin America between 2004 and 2018, we found that Chinese entities account for $152 billion (more than 90 percent) of the total amount lent through identified RBLs.
NRGI and others have frequently criticized RBL deals for their opacity. Information about agreed terms and conditions is scant. At the time of our study, we lacked basic information, such as the interest rate, for 33 out of the 52 cases surveyed, and just one loan document was publicly accessible: the Sicomines loan that the Democratic Republic of the Congo (DRC) signed in 2011. But since the publication of our report a year ago, the transparency landscape has already showed signs of change.
Last week, AidData released a trove of loan agreements between Chinese entities and sovereign borrowers in developing countries. The researchers scoured debt information management systems, official registers and gazettes, and parliamentary websites to find 100 publicly available Chinese loan agreements in 24 countries. Their database, and accompanying “How China Lends” report comparing terms with other loan agreements from a benchmark sample, both confirms and expands our understanding of sovereign debt agreements involving Chinese state lenders.
One key finding of the report is the widespread use of far-reaching confidentiality clauses in the Chinese loan agreements meant to restrict borrowers from disclosing loan information. This confirms worries about opacity and the need to rein in confidentiality provisions in sovereign debt agreements, as previously highlighted. However, the fact that a significant number of Chinese loan agreements were publicly accessible despite these confidentiality provisions may point to a pathway to greater transparency: leveraging and building on borrower country laws and procedures requiring parliamentary ratification and/or disclosure of sovereign debt agreements. Carve-outs from confidentiality, where applicable law requires disclosure, are relatively common and is likely how many of the Chinese loan agreements ended up publicly disclosed and in the database notwithstanding the confidentiality clauses.
In addition to the Sicomines loan, the AidData database includes contracts for five more RBLs: three with the government of Ecuador, one with the government of Ghana and one with Venezuela’s state-owned enterprises (SOEs) PDVSA and BANDES. We featured these deals in the 2020 NRGI report, and the key terms we had reported in relation to loan amounts, interest and maturities were the same or very similar to those included in the contracts. (We plan to update our report’s dataset with insights from the AidData work, adding new details such as missing grace period information).
Access to the contracts allows us to better understand certain critical details about these loans. For example, while we and others have reported Ghana’s 2018 loan from Sinohydro, a Chinese SOE, as a $2 billion loan, the published agreement clarifies that the Chinese lenders offered $1.7 billion and that the Ghanaian government sought an additional $300 million elsewhere. The contracts also shed light on how the government has used Ghana’s bauxite revenues as collateral and how this approach fits within broader patterns of collateralization in Chinese lending described in the AidData report, namely the requirement to maintain a funded bank account with the lender that could be seized in case of non-payment. In Ghana’s case, the agreement specifies that bauxite proceeds be kept in an offshore escrow account and that the government entered into an escrow agreement with creditor Sinohydro.
Disclosed loan agreements can also be useful for countries, like Guinea, with similar Chinese loans that are not yet disclosed. In 2017 the Guinean government signed an agreement with China for a $20 billion credit line, from which loans backed by bauxite receipts would be drawn with a group of Chinese state-owned companies and lenders. Information on the credit line and loans signed in 2018 was limited, but we now have the headline parameters through Guinea’s latest Extractive Industries Transparency Initiative (EITI) report (broadly aligned with EITI’s recent guidance on barter deals). Guinean civil society organizations and EITI Guinea have been instrumental in making this disclosure happen. The remaining challenge is access to the actual loan and infrastructure contracts that would unveil the overall commitments between the government and its Chinese partners. This would enable oversight actors to undertake a comprehensive risk assessment, both of the government’s ability to repay existing loans and whether it is advisable for the government to draw down further loans on the existing line of credit given the uncertain future economic context.
The need for transparency around loans and financing arrangements involving natural resources is not limited to those involving Chinese lenders. RBLs from commodity trading firms are often even more opaque than those from Chinese lenders and have destabilized economies in borrower countries. Transparency is also essential as governments use increasingly creative ways to raise finances leveraging their natural resources, as recent developments in the DRC and Ghana illustrate. The DRC’s commitment to disclose mining contracts, and civil society monitoring of the commitment, resulted in the recent publication of one such contract. In 2017, the country’s giant state-owned mineral company Gécamines sold its future mining royalty revenues from the Metalkol cobalt and copper project to Multree Limited, a company associated with Dan Gertler, who has faced U.S. sanctions on the basis of alleged corruption. While certain details of the deal remain unclear, it appears that Gécamines exchanged future royalties for an upfront payment of $55 million and Multree’s assumption of a $28 million debt owed by Gécamines to Metalkol. In Ghana, the government recently confirmed that it will resume its own controversial plan to sell an interest in future gold royalties to private investors through a stock exchange listing.
As many developing countries struggle with debt, increased transparency around Chinese lending and resource-based financing is critically important. Access to information about past deals will enable borrower countries seeking fresh financing or contemplating debt renegotiation to better structure and negotiate. And other lenders can use the information to lead by example in providing finance on more sustainable terms. Perhaps most importantly, sovereign debt transparency is also critical for civil society organizations and other oversight actors. According to the authors of the “How China Lends” report:
Disclosing all debt contracts, however difficult politically, should become the norm rather than the exception. This would give citizens the ability to hold their governments accountable for the debt contracts signed in their name. Public debt should be public.
Disclosure of sovereign loan agreements is a natural progression from the publication of oil, gas and mining contracts, which has improved markedly in recent years. In 2009, governments and companies expressed skepticism around the publication of these documents, but just over 10 years later at least 49 countries have officially disclosed contracts and over 20 companies publicly support the practice. Extractive sector contract transparency is now an established global norm endorsed by the IMF and required by EITI.
Given the similar public interest questions surrounding the management of public debt and the governance of natural resource wealth, those working for greater transparency of sovereign debt should build on these gains. RBLs are of particular importance—they are the intersection of extractives transparency and debt transparency. Momentum around greater sovereign debt transparency is building and this can only benefit the governance of natural resources.
Notes
-
1
With a gold price of USD 1,600 per ounce.
-
2
With a low-profit mine and a gold price of $1,600 per ounce.
-
3
Cecilia Jamasmie, “Petra Diamonds’ stake in Williamson to shrink as part of deal with Tanzania,” Mining.com, 13 December 2021, www.mining.com/petra-diamonds-stake-in-williamson-to-shrink-as-part-of-deal-with-tanzania.
-
4
Lifezone Metals, “Kabanga Nickel Signs Framework Agreement,” 19 January 2021, www.lifezonemetals.com/kabanga-nickel-signs-framework-agreement.
-
5
Thomas Scurfield and Silas Olan’g, “Magufuli Seeks the Right Balance for Tanzania’s Mining Fiscal Regime,” NRGI, 31 January 2019, www.resourcegovernance.org/blog/magufuli-seeks-right-balance-tanzania-mining-fiscal; Thomas Scurfield, “Tanzania Strikes a Better Balance with its Mining Fiscal Regime,” NRGI, 24 June 2020, www.resourcegovernance.org/blog/tanzania-strikes-better-balance-mining-fiscal-regime.
-
6
This framework agreement was published in a document setting out Barrick’s offer to buy the shares it did not already own in Acacia Mining, the previous owner of the Bulyanhulu, Buzwagi and North Mara mines in Tanzania. See Acacia Mining and Barrick Gold, Recommended Final Offer for Acacia Mining Plc by Barrick Gold Corporation, 2019, 66–79, s25.q4cdn.com/322814910/files/doc_downloads/acacia/Acacia-2.7-announcement.pdf.
-
7
The main revenue streams are import duty, skills development levy, royalty, corporate income tax, a share of dividends and shareholder loan repayments through state equity, and dividend withholding tax.
-
8
The earlier in time a shilling (Tanzania’s official currency unit) is received, the more it is worth. This is, first, because it can be used earlier; and second, because the future is uncertain, and no one can be sure they will receive that shilling in the future. To account for this time value of money, a “discount rate” is applied. In the sharing arrangement, this would mean that if the government received a shilling in year 1, the company would need to receive more than a shilling in year 2 for the benefits to be comparable. However, given cumulation here is based on actual cash flow, the company would need to receive only a shilling in year 2 for the benefits to be shared equally.
-
9
This provision for the company to earn its minimum return before sharing is triggered means Ecuador’s mechanism is similar to an R-based cash flow tax, commonly referred to as a Brown Tax. See, e.g., Robin Broadway and Michael Keen, “Theoretical perspectives on resource tax design,” in The Taxation of Petroleum and Minerals: Principles, Problems and Practice, edited by Philip Daniel, Michael Keen and Charles McPherson (Oxford: Routledge, 2010), 13–74.
-
10
Prices are taken from World Bank, “Commodities Price Data (The Pink Sheet),” www.worldbank.org/en/research/commodity-markets.
-
11
With a gold price of $1,600 per ounce.
-
12
With a gold price of $1,600 per ounce.
-
13
With a gold price of $1,600 per ounce.
-
14
The Fraser Institute survey estimates that, unless there are extremely harmful policies, around 60 percent of an investment decision tends to be based on a country’s geology. The other 40 percent comprises of several other factors, including political stability and policy predictability (given they affect the risk that investors will not be able to secure future returns generated by their investments), a conducive business environment and the tax level. See Julio Mejia and Elmira Aliakbari, Fraser Institute Annual Survey of Mining Companies 2022, (Fraser Institute, 2023), 8, www.fraserinstitute.org/studies/annual-survey-of-mining-companies-2022.
-
15
However, information gaps make it difficult for taxes to be structured to capture all excess profit. See Jean-Franҫois Wen, Progressive Taxation of Extractive Resources as Second-Best Optimal Policy (International Monetary Fund, 2018), www.imf.org/en/Publications/WP/Issues/2018/06/13/Progressive-Taxation-of-Extractive-Resources-as-Second-Best-Optimal-Policy-45923.
-
16
Recent research provides a sense of the potential revenue loss to governments from tax avoidance. The International Monetary Fund recently estimated that sub-Saharan African mining countries could be losing between $450 and $730 million in corporate income tax a year. See Sebastian Beer and Dan Devlin, Is There Money on the Table? Evidence on the Magnitude of Profit Shifting in the Extractive Industries (International Monetary Fund, 2021), www.imf.org/en/Publications/WP/Issues/2021/01/15/Is-There-Money-on-the-Table-Evidence-on-the-Magnitude-of-Profit-Shifting-in-the-Extractive-49983.
-
17
It is perhaps surprising that Tanzania’s 50-50 sharing arrangement generates an AETR greater than 50 percent (with a discount rate of 10 percent). This is despite AETR measuring government take as the share of pre-tax profits, which is larger than “economic benefits” (given economic benefits exclude interest payments). This outcome results from the 50-50 split being based on actual cash flow. The government receives revenue before the mining company through input and production taxes that do not depend on the mine making a profit. Because of these earlier revenues, the government receives a larger share on a discounted basis.
-
18
As reported in the S&P Global database. Legal risks are “expropriation, state contract alteration and contract enforcement risks.” Tax risks are “tax increase and tax inconsistency risks.” Control Risks scores these risks as still “very high” and “high” respectively (following Tanzania’s overhaul of extractives sector laws and other actions against existing mines in 2017) but reducing.
-
19
With a discount rate of 10 percent. While Ecuador’s sharing mechanism does not account for the labor profit share because none of it will go to the government from 2024 onwards, I have included it in the AETR because it is a tax on the project. The Democratic Republic of Congo regime has an excess profits tax that is triggered for a mine when the realized price is at least 25 percent higher than the price in its feasibility study. I assumed that the feasibility study has a price of $1,300 per ounce, so the excess profits tax is not triggered.
-
20
Total benefits in this case are a project’s revenues minus operating costs and replacement capital (but not minus exploration and development capital). This cash flow represents the money available to pay back the initial investment and provide a return. The government share of it is a common measure of progressivity.
-
21
Wen, Progressive Taxation of Extractive Resources as Second-Best Optimal Policy.
-
22
With a discount rate of 10 percent. The results for only some countries are shown to clearly depict each data point. The results for all the evaluated countries can be found in my model.
-
23
Scurfield, “Tanzania Strikes a Better Balance with its Mining Fiscal Regime.”
-
24
This feature is not fully reflected in Figure 5 given that “total benefits” use a slightly different definition of profits and are based on discounted cash flows.
-
25
With a gold price of $1,600 per ounce.
-
26
For example, an average 62 percent of respondents to the Fraser Institute surveys between 2017 and 2019 said the current implementation of Tanzania’s legal system would strongly discourage investment, and 73 percent said regulatory uncertainty would. See, e.g., Ashley Stedman, Jairo Yunis and Elmira Aliakbari, Fraser Institute Annual Survey of Mining Companies 2019 (Fraser Institute, 2020), www.fraserinstitute.org/studies/annual-survey-of-mining-companies-2019.
-
27
With a low-profit mine and a gold price of $1,600 per ounce.
-
28
Tax avoidance could extend the Philippines’ recovery period, and therefore delay the payment of some taxes including import duty and interest withholding tax, given the end of the recovery period depends on the reported profitability of a mine rather than an ex-ante assessment. However, the rule that the recovery period must end five years from the start of production regardless of whether pre-production expenses have been recouped limits the extent to which it can be extended.
-
29
The merits of these measures require further scrutiny. E.g., taking a share of loan repayments could result in lenders charging a higher interest rate to ensure they still recoup their loan and a minimum return. This would not only reduce taxable income but also make it harder for the government to assess whether an interest rate is reasonable, because it would not be comparable with industry benchmarks. It can also be difficult for a government to always determine whether a loan is from a related party or not. However, these considerations are outside the scope of this analysis.
-
30
Natural Resource Governance Institute (NRGI), Natural Resource Charter, 2nd edition, 2014, resourcegovernance.org/analysis-tools/publications/natural-resource-charter-2nd-ed.
-
31
Although its exclusion of several significant taxes from the government’s share of benefits means low-profit mines may still be impacted.
-
32
With a gold price of $1,600 per ounce.
-
33
Anna Fleming, Thomas Lassourd and David Manley, “Variable Royalties: an Answer to Volatile Mineral Prices?” in Handbook on the Future of Resource Taxation, African Tax Administration Forum and Intergovernmental Forum on Mining, Minerals, Metals and Sustainable Development (forthcoming), www.iisd.org/publications/brief/future-resource-taxation-roadmap.
-
34
Ensuring that interest rates used as a comparison apply to comparable assets with a similar risk profile is challenging, but rules of this nature have been successful in reducing profit shifting in other countries. See Beer and Devlin, Is There Money on the Table?.
-
35
The main taxes listed are VAT, royalty and corporate income tax. The regime also includes a share of pre-tax profits that is currently divided between the company’s workers and the government, with the portion received by the government included in its accumulated benefits. However, a recent court ruling means that all this labor profit share will go to workers from the start of 2024 and therefore none of it will be included in government benefits.
-
36
The discount rate used is specific to each mine and based on its weighted average cost of capital (WACC). I have assumed that WACC is around 7 percent in real terms. This is based on the typical discount rate for equity shareholders used by industry and government analysts of 8 percent in real terms, and the current average cost of debt for the mining sector as reported by Aswath Damodaran, Damodaran Online, www.pages.stern.nyu.edu/~adamodar.
-
37
Republic of the Philippines, Financial or Technical Assistance Agreement, mgb.gov.ph/attachments/article/79/PFC_FTAA.pdf.
-
38
For example, some terms in the original FTAA for an OceanaGold mine differed in some areas: Republic of the Philippines, Financial or Technical Assistance Agreement with Arimco Mining Corporation, 1994, www.resourcecontracts.org/contract/ocds-591adf-2792396017. I understand that the recently signed extension to this agreement has slightly different terms again.
-
39
A template of the FARI model and a user guide that explains all the concepts and workings of the model are available at International Monetary Fund, “Fiscal Analysis of Resource Industries,”www.imf.org/external/np/fad/fari.