The application of global tax reforms to the extractive industries

Cross-posted from the International Tax Review.

In this article I investigate whether IMF and OECD proposals on how to tax the digital economy, apply formulary apportionment, shift taxing rights and address profit shifting could apply to the natural resource sector. Alternatively, could improvements to source-based taxation hold the answer?

There are potentially large changes afoot in how the profits of multinationals are taxed. Motivated by the challenge of taxing the digital economy, the OECD is considering shifting some taxing rights into the market or destination country. The IMF is taking a broader view of corporate tax reform, with particular concern for the vulnerability of developing countries to profit shifting activities. 

However, neither proposal examines the options for taxing the extractive industries – mining, oil and gas.The OECD’s programme of work on digital tax reform suggests that commodities are likely to be left out of future developments (see OECD, ‘Programme of Work to Develop a Consensus Solution to the Tax Challenges Arising from the Digitalisation of the Economy’, May 2019). In an earlier paper the IMF acknowledged that most of the policy options for corporate tax reform, with some exceptions, may not be suitable for natural resources (see IMF, ‘Corporate Taxation in the Global Economy, March 2019). 

That is because source-based taxation is considered the most appropriate model of taxation for natural resources, which are location-specific. The intuition is right: resource-rich countries should get the lion’s share of the fiscal benefits from resource extraction. The challenge is that source-based taxation of natural resources is also vulnerable to profit shifting. 

Take for example, the court case in Australia against oil and gas company, Chevron. In Chevron Australia Holdings Pty Ltd v. Commissioner of Taxation, the Federal Court upheld the Australian Taxation Office’s (ATO’s) finding that Chevron had used a series of intra-company loans to shift profits offshore. The company was ordered to pay A$300 million ($209 million). 

Developing countries, on the other hand, simply cannot afford such a dispute, which cost the ATO A$10 million in legal fees, and is the equivalent of 45% of the Sierra Leonean tax authority’s total operating budget ($16.4 million) for 2016. Even common profit shifting challenges, such as intra-firm lending, loom especially large because of the scale of operations and the significant revenues at stake (see Figure 1). 

It is therefore necessary to determine whether the alternative models of taxation being considered, that are more robust against profit shifting, could also apply to natural resources. 

Figure 1. Resource revenue as a percentage of total revenues 

Source: Author’s estimates using the Government Revenue Dataset (ICTD) and EITI data

Source: Author’s estimates using the Government Revenue Dataset (ICTD) and EITI data


DBCFT and the ‘marketing intangibles proposal’


The proposed destination-based cash flow tax (DBCFT) is a levy on corporate income based on where goods end up, rather than where they were produced.Despite being more robust to profit shifting, resource-rich countries are unlikely to agree to let the profits from their natural resources be taxed in the customer’s location. 

The IMF paper, entitled ‘Corporate Taxation in the Global Economy’, estimated that resource-rich countries stand to lose 1% of GDP under such a tax. That is not a concern just for developing countries: In 2017,63% of Australia’s mineral and metal exports ($60.2 billion) went to China, whereas China accounted for only 24% of Australia’s imports.

Similar problems could arise if marketing intangibles are adopted for residual profit allocation, as proposed by the OECD. The dispute between the ATO and BHPregarding the company’s offshore marketing hub is a case in point. BHP was allocating a large portion of the profits from the sale of Australian iron ore to its hub in Singapore based on marketing intangibles. The ATO rejected the arrangement, leading to a large tax adjustment and to BHP increasing its ownership of the hub so that all profits in Singapore stemming from Australian mines are taxed in Australia. 


Formulary apportionment

Alternatively, formulary apportionment – or ‘fractional apportionment’, as the OECD is calling it – could apply to the extractive industries, but there are challenges. 

First, a formula that appropriately compensates the producing countries would need to be determined. Consider, for example, Rio Tinto and Anglo American’s global profits for 2017, shown in Figure 2 and 3. From the perspective of producing countries, the worst scenario is if profits are allocated to where the minerals are sold, for example, China, Europe and the US. A better model is if profits are apportioned equally between sales, assets, and employees. An even better result for Anglo American's mines is if the formula is modified to include assets and an ‘extraction factor’ (that is, total production, as in Alaska). The exception is South Africa and Chile, which receive a greater share of global profits if assets and employees are used.

Rio Tinto's mines, on the other hand, while better off under the 'Alaska model', receive the greatest share of global profits under assets and employees, although Australia is an exception. 

While it seems counterintuitive that producers would get more global profit under assets and employees than assets and production, a mine’s contribution to total production will depend on its stage of development. 

In Mongolia, for example, Rio Tinto’s Oyu Tolgoi copper mine is under construction and yet to ramp up production. Consequently, Mongolia’s share of total production is less than its share of total employees; thus, Mongolia receives less profit when production, rather than employees, are used to allocate profits. 

It may also depend on the commodity as some are more labour intensive than others.For example, platinum mines are typically more labour intensive than other commodities, which is why Anglo American's employment data for South Africa is significantly higher than elsewhere. Both these examples demonstrate the need for careful analysis of apportionment factors over life-of-mine and for different commodities.


Figure 2. Apportionment of Anglo American’s global profits under four models of formulary apportionment (2017) (%)

Source: Anglo American’s annual report (2017) for accounting information. Data used is illustrative.

Source: Anglo American’s annual report (2017) for accounting information. Data used is illustrative.

Figure 3. Apportionment of Rio Tinto’s global profits under four models of formulary apportionment (2017) (%)

Source: Rio’s annual report (2017) for accounting information. Data used is illustrative.

Source: Rio’s annual report (2017) for accounting information. Data used is illustrative.

Another challenge for formulary apportionment involves intermediary mineral products. Many resource-rich developing countries export unfinished raw materials that are transferred to a related-party refinery for further processing before sale. 

Rio Tinto, for example, operates a mineral sands project in Madagascar. It produces ilmenite, which contains 60% titanium oxide. The raw material is shipped to Rio Tinto’s Fer et Titaneprocessing plant in Canada, where it is transformed into a 90% titanium dioxide chloride slag. A formula would need to apportion profits for the raw material extracted in Madagascar, as well as value added outside the country. While many diversified miners produce financial information by business unit (for example, iron ore, copper), further disaggregation of the relative costs of production and overheads would be required to set an appropriate formula for each commodity.

There is also a question of whether to use assets’ accounting value or fair market value. The difference may be quite substantial. For example, the accounting valuing of Rio’s Australian iron ore operating assets are $16 billion, whereas the market value is more than $50 billion. While the market value may be a more relevant basis for profit allocation – Australia’s tax take would be much higher, for example – the historical cost of purchase is more objective and easily verified.

Finally, resource-rich countries would need to be willing to accept global offsetting of corporate losses. Consider Anglo American’s operations in Brazil and Chile. If the Brazil mine incurs a loss, the company can offset that against the tax base of the mine in Chile, whose government thus gets less revenue from its minerals. 

From the taxpayer’s perspective, it is only fair that if income from Brazil can be included in Chile’s tax base, so too should a corollary loss. Chile, however, may disagree that the compensation it gets for its minerals should be impacted by activities in Brazil. Whether it is worthwhile for Chile and Brazil will depend on careful analysis of the risk to revenue under international tax rules and whether both countries would on average be better off under formulary apportionment.


Minimum tax on inbound investment

The OECD and IMF are also considering a minimum tax on inbound and outbound investment. A minimum tax on inbound investment should reduce cost-based profit shifting – a major concern for resource-rich developing countries that depend on foreign capital to develop their mineral deposits. It would also limit countries’ use of tax incentives, which could affect investment, although is a smaller concern in mining given investment is tied to a fixed resource and therefore less mobile.

While a minimum tax on inbound will help reduce cost-based profit shifting, extractive companies might still understate the value of mineral sales. However, for widely traded products (for example, precious metals and crude oil), that problem can be addressed through greater use of benchmark prices. Many Latin American countries already require taxpayers selling mineral products to offshore related parties to use the publicly quoted price to calculate corporate income tax (commonly known as the sixth method). For minerals that lack a market price, advance pricing agreements could be used to negotiate a pricing formula.


Source-based taxation for resource rents

Assuming that developing countries continue to use some source-based taxation for resource rents, the questions are what type of system and how resistant is it to profit shifting? 

There are several factors to consider when designing a resource tax system — efficiency and adaptability, for example. However, if vulnerability to profit shifting is taken in isolation, it is clear that some instruments are better than others. 

The main difference is between taxes on turnover and taxes on profits. Whereas a royalty is generally applied to the gross value of the mineral product, an income tax is applied to the taxpayer’s net profit. The base for corporate income tax will always be lower than the base for royalties. Corporate income tax is also more prone to tax avoidance because the base is net income. 

Another way for countries to capture the domestic value of their resources is by taking a share of the physical production. Theoretically, production sharing is less vulnerable to avoidance, provided there is a cap on how much of the production the investor can take to recover its costs.

To conclude, the international community will not be able to help developing countries meaningfully combat profit shifting without addressing the risks specific to the natural resources sector. 

Any proposal that aims to shift taxing rights to the market jurisdiction will not be appropriate for resource-rich countries and is likely to be met by significant resistance from governments. Formulary apportionment is more plausible, although further research is required. A minimum tax on inbound investment, combined with greater use of benchmark prices, and production-based taxes would go a long way in addressing profit shifting in the natural resource sector. No matter the system, coordination is key to prevent countries from increasing their exposure to tax-related competition and to avoid double taxation. Above all, it should be easier for developing countries to administer.  


This article is based on a presentation the author gave at the fifth annual World Bank Group/IMF Spring Meetings conference on taxation in Washington on April 14.

The data used is illustrative, taken from publicly available data that would need further breakdown to arrive at a more robust position. Rio and Anglo American were selected because of their advanced disclosure of data.