This chapter describes suitable valuation methods of natural resources for the purposes of national accounts compilation. The preferred valuation method is using market prices or market-equivalent prices. When these are not available, it is recommended to apply the Net Present Value method of future resource rents, which is discussed in great detail, including the derivation of resource rent, the estimation of user costs of capital, the treatment of taxes and subsidies and discounting. The chapter also details how to measure and record depletion of natural resources and provides recommendations on the split-asset approach. Finally, guidance is provided on the treatment of negative resource rents.
Measuring Natural Resources in the National Accounts
A Compilation Guide
Abstract
105. This chapter describes suitable valuation methods for natural resources, as agreed in the context of the SNA update process and discussed by the EGNC, while Chapters 4 and 5 look at the practical applications in more detail for mineral and energy resources and for biological resources. The outline of this chapter is as follows. Section 3.2 will discuss suitable valuation methods for the purposes of national accounts compilation including a preference order. Section 3.3 discusses the Net Present Value method of resource rents. Section 3.4 discusses the treatment of taxes and subsidies in deriving resource rent. Section 3.5 details how to measure and record depletion, Section 3.6 covers the split-asset approach, Section 3.7 provides guidance on the treatment of negative resource rents. A text box with a summary of key recommendations concludes the chapter.1
106. The general principle when it comes to valuation is described by 2025 SNA (§3.60) as follows: “Transactions are valued at the actual price agreed upon by the transactors. Exchange values, or the observed market prices, are thus the basic reference for valuation in the integrated framework of the SNA.” Exchange value is defined as follows: “the current exchange value, often labelled as the “market price”, refers to the value at which goods, services, labour or assets are exchanged, or else could be exchanged, for cash (currency or transferable deposits). Exchange prices are the basis for valuation of transactions in the integrated framework of the SNA.” (2025 SNA §4.131)
107. Given the principal objective to measure exchange values, a range of methods can be applied for measurement. We will first discuss methods for valuing transactions and stocks including a preference order drawing from Chapter 4 of the 2025 SNA (especially its Annex). Then we will discuss valuation of natural resources in greater detail.
3.2.1. Methods for valuing transactions and stocks
108. Regarding transactions the preferred valuation method is based on observed exchange values (or observed market prices). Exchange prices for transactions are defined in the 2025 SNA (§4.142) as “amounts of money that willing buyers pay to acquire something from willing sellers; the exchanges are made between independent parties and on the basis of commercial considerations only, sometimes called “at arm’s length”.” The SNA places no restrictions on the type of market or the functioning of markets when observing prices to measure exchange values (e.g. markets need not be competitive), but notes that in some cases such as transfer pricing or concessional pricing, the accountant may deviate from observed market prices.
109. In the absence of observed market prices (2025 SNA §4A.3), it is recommended to apply prices for similar goods, services and assets, or market equivalent prices. When such markets do not exist, alternative valuation methods that approximate the main SNA valuation principles should be applied such as a sum-of-costs approach in case of non-market production. This approach is commonly used for the measurement of education and health services, and measures output by adding all costs that are incurred for providing these services, including intermediate consumption, compensation of employees, and the depreciation of assets (e.g. school buildings).2
110. There exist also various valuation methods commonly used in welfare economics based on measuring the Willingness-to-Pay or Willingness-to-Accept which are not appropriate for national accounts purposes, as they include the consumer surplus which should be excluded from exchange values (see Chapter 2).
111. The preference order for valuing assets (2025 SNA §4A.15-40) consists of observed market prices, market-equivalent prices, valuation based on past expenses (such as the written down replacement costs), nominal value, indirect valuation, and finally based on the present value of future economic benefits.3
Valuation of natural resources
112. The 2025 SNA mentions (§4A.30) that the present value method is typically used for valuing natural resources, in which case “the method comes down to estimating the present value of future economic benefits derived from these assets (i.e., the resource rents), which often need to be approximated by the so-called “residual value method”.” (§4A.32). It emphasises that the method requires making several assumptions and therefore is considered a last resort (§4A.33). Finally, a reference is made to the SEEA CF for more detailed guidance.
113. The SEEA CF (United Nations et al., 2014, p. 182 (§5.109)[1]) observes that for many non-produced natural resources (such as oil and gas deposits), observable market prices for the assets themselves in situ4 are not available (nor prices for which they were acquired), what is observable are the outputs derived from the natural resources through harvest (e.g. of timber) or extraction (oil or gas). Under these circumstances, the asset in situ can be valued based on the Net Present Value (NPV) method of future returns, where these returns are measured as resource rents.
114. For some types of natural resources observable market prices may be available, such as transactions in (forest) land or transactions in standing timber. There are also instances where the value of a natural resource can be obtained indirectly through observed transactions in associated assets such as land or dwellings. In such cases, where observable market prices are available this would be the preferred valuation method.
115. The Eurostat and OECD compilation guide on land estimation (Eurostat and OECD, 2015, p. 65[2]) discusses several indirect approaches for valuing land: the residual approach, the land‑to‑structure ratio approach and the hedonic approach, that may also be relevant for valuing natural resources other than land. The residual approach estimates the value of land by deducting the value of structures (such as dwellings, which can be derived from a perpetual inventory model) from the combined asset value of land and structures (which is typically observed in market transactions). The same method can also be applied specifically for valuing forest land, starting from transactions in forest estates by deducting the value of standing timber, as further discussed in Section 5.2. In the case of renewable energy resources, one could compare the price of land without structures with the price of land with solar panels or wind turbines installed and use the price difference as a way to value the renewable energy resource.5
116. Hedonic approaches are discussed in the 2025 SNA primarily in relation to price measurement (§18.86-18.92) but may also be used for valuing natural resources such as standing timber or renewable energy resources (as further discussed in Chapters 4 and 5).
117. While markets for natural resources in situ may not exist, in many countries markets for rights to access natural resources exist (e.g. fish quota or rights to extract oil or gas). Using market prices for the value of these rights as a means to value the resource itself, is referred to as an access price method (United Nations et al., 2014, p. 154 (§5.128)[1]) in SEEA CF. Such values may provide a correct valuation of the natural resource, especially when the right (permit or license) was purchased during an open competition or when it is transferable i.e. it may be sold to other economic agents (2025 SNA, §11.210).
118. However, the value of the right will depend on the duration for which access is granted: if this period is shorter than the expected asset life of the resource, the value of the right may only provide a partial value of the natural resource. Moreover, “in practice, in many cases governments may give the access rights to extractors for free or do so at a price that is less than the true economic value. Further, trading of the rights may be restricted or prohibited. In these cases, there is no directly observable market valuation, and the present value of future resource rents should be used.” (2025 SNA, §11.211)
119. The 2025 SNA summarises (§4A.40) that “for non-financial assets, in the absence of observed market prices or market equivalent prices, two valuation methods are applied most frequently, either the written-down replacement cost method or the present value of future earnings. The former method is typically applied to fixed assets used in the production of goods and services, while the latter method is often the only alternative for arriving at an approximation of the value of natural resources.”
120. In the next section we will therefore discuss in greater detail the NPV of future resource rents method, which is the recommend method for valuing natural resources in the absence of market transactions that can be used for a direct or indirect valuation.
121. As markets for transactions in natural resources in situ6 are often thin or absent and the right (or quota) to access a resource oftentimes will only reveal a partial value of the resource, it is recommended to apply the Net Present Value (NPV) method to estimate asset values on the basis of future resource rents as a last resort option. The NPV method requires:
1. An estimation of resource rents in current and previous reference years as well as projection of future resource rents.
2. An estimation of the asset life (which can be infinite in case of renewable resources).
3. Discounting to convert future resource rents (flows) into a current period estimate of value (stock).
122. The NPV method relies on several assumptions and is therefore only to be used if market prices are not available. As a general principle, it is important to ensure comparability across countries by standardising the assumptions used to the extent possible. If this is not feasible, it is recommended to be transparent about the assumptions that are used in the metadata (accompanying documentation) and provide insights into the impact of any assumptions used.
123. We will start by discussing resource rent, and how it is recommended to be measured by the residual value method (RVM), including a discussion of key inputs such as the rate of return to capital. This will be followed by a discussion of the NPV method, including the choice of asset life of the resource, the role of expectations and the choice of the discount rate. The role of taxes and subsidies in deriving resource rent is discussed in Section 3.4.
3.3.1. Resource rent
124. First, it is important to clarify terminology. The 2025 SNA (§8.166) explains that “rent should be clearly distinguished from “resource rent” or “natural resource rent”. The latter represents the surplus value accruing to the extractor of an asset after all costs have been taken into account.”7 The resource rent concept is therefore equal to the concept of capital services for natural resources (2025 SNA §17.45). Capital services can be understood in this context as the benefits that are derived from using an asset in production, noting that the word “service” has a different meaning than usual, when it is understood as the output of a production process.
125. The resource rent should be distinguished from rent on natural resources (D45) as defined in national accounts (2025 SNA, §8.115) as “the income receivable by the owner of a non-produced natural resource or another non-produced non-financial asset (the lessor or landlord) for putting the asset at the disposal of another institutional unit (a lessee or tenant) for use in production.” As mentioned previously, when rights to exploit a natural resource have been allocated by an open competition process or when rights are transferable, the value paid by the extractor may provide a proper valuation of the resource (i.e. of the part covered by the right).8 However, as stated in SEEA CF (§5.217), “in many instances fees, taxes and royalties actually collected tend to understate total resource rent, as the rates may be set with other priorities in mind, for example, encouraging investment and employment.” 9
126. It is therefore recommended to calculate and compare the resource rent with rent payments (when the contract is annual) and apply a threshold: if rent payments are on average more than 90% of the resource rent, it may be assumed that economic ownership remains with the legal owner and there is no need to split the asset between legal owner and extractor. A threshold of 90% is chosen to reflect that the calculation of resource rent also depends on various assumption and is subject to uncertainty; if less than 90% is appropriated it is reasonable to assume that part of the asset value is with the extractor. This check would need to be performed when the contract is entered into (or the mechanism established) and during benchmark revisions. In between these years, it may be assumed that current value of the rights provides a proper valuation. When the contract is for multiple years, a comparison should be made between the value of the contract and NPV of resource rent for the same number of years, with the same 90% threshold applied.10
127. The recommended method for estimating resource rent is the residual value method (RVM) (2025 SNA §4A.32). The reason behind this method is that we can understand value added (for instance in oil extraction) as a return to the use of labour and a combination of capital assets in production: produced assets (such as machinery), non-produced non-financial assets and natural resources. The return to labour is measured through remuneration of employees;11 the return to capital used in production (excl. natural resources) is estimated based on information about the relevant assets and depreciation; the return to natural resources can then be obtained as a residual (hence the name of the method). The return to natural resources is called resource rent and equals the net return to natural resources and depletion.
Residual value method
128. It is useful to make a distinction between a top-down method and a bottom-up method for applying the RVM. Both methods are examples of an “activity-based” approach. Table 3-1 explains the derivation of resource rent in greater detail according to the top-down method. The bottom-up method is discussed below.
129. The usual point of departure for the compilation of the top-down method is information readily available in the national accounts in the form of Gross Operating Surplus (GOS) by an economic activity (classified by the International Standard Industrial Classification of All Economic Activities (ISIC)). GOS is obtained by deducting intermediate consumption, remuneration of employees and other taxes on production from output at basic prices and adding other subsidies on production.12 The GOS is more or less the same thing as business operating profits, and measures what a business retains as a surplus after paying certain production-based taxes13 and receiving subsidies. If we wish to value coal and lignite resources, for example, we would use the GOS of ISIC Division 05 (Mining of coal and lignite). The precise delineation of economic activities when applying the RVM will be further discussed for individual resources in Chapters 4 and 5.
130. Gross mixed income (GMI) is recorded for unincorporated enterprises14 when it is not possible to separate remuneration of employees from return to capital. In ISIC Section B Mining and quarrying, GMI is expected to be insignificant which is why the SEEA CF (which focused on the valuation of subsoil assets) labels the row GOS (with an accompanying footnote about GMI). However, according to the 2025 SNA (§7.154 and §7.155) household production of electricity and heat will be classified as output by an unincorporated enterprise owned by households.
131. When measuring the value of renewable energy resources, the inclusion of GMI will therefore be relevant. Also in case of the valuation of biological resources, mixed income may be relevant for instance in case of fishing activities or timber harvesting activities that go beyond own final consumption (i.e. there has to be a commercial purpose) to value the natural resource (2025 SNA §13.24). However, it is really only the GOS element of mixed income (equivalent to the profits of the unincorporated enterprise in the households sector) that we would wish to use as the starting point.15
132. Second, we need to deduct Specific subsidies related to production (D31 and D39) and add Specific taxes related to production (D21 and D29) yielding what Table 3-1 calls GOS and GMI for the derivation of resource rent. The reason behind is that if the resource is owned by government, and we want to assess what the resource is worth to society, we should neutralise the specific taxes and subsidies as they constitute mere distributions. Specific taxes and subsidies will be further discussed in Section 3.4.
133. The final step to obtain resource rent in Table 3-1 is to subtract the user costs of capital (excl. natural resources), consisting of two elements:
depreciation; and
a return to capital used in production.
134. An alternative is to derive resource rent using the bottom-up method towards RVM. This may be useful in cases where the national accounts and economic activity/industrial classification breakdowns do not provide a 1-1 link16 to the asset that the compiler is trying to estimate, and detailed source data is available which would produce a better result. For example, for renewable energy resources, the ISIC Revision 4 classification does not distinguish between renewable and non-renewable resources, let alone provide a breakdown between different types of renewable resources such as solar, wind or hydro power. Therefore, GOS or GMI (the starting point for the top-down approach) are unlikely to be available from the national accounts at the required level of detail.
135. This will change in future as countries implement ISIC Revision 5, which shows electric power generation activities from renewable sources in class 3512 separately from non-renewable sources (class 3511). If the figures at ISIC class level feed through into the national accounts, it may become feasible to use the top-down method for renewables in future. However, breakdowns by type of renewable energy (solar, wind, hydro etc) will still not be available, so some compilers may still prefer the bottom-up approach for estimating resource rent from renewables.
136. The bottom-up method follows the same recipe as shown in Table 3-1 but uses other data sources (e.g. business surveys of industries engaged in a specific economic activity) to derive GOS/GMI that is specific to the activity in question. The same steps are then followed to derive resource rent as for the top-down method, namely deducting specific subsidies on production and adding specific taxes on production; and then subtracting the user costs of capital excluding the extracted resource (depreciation and return to capital). In this method, these items may also be calculated specifically for the activity in question, rather than using information in the national accounts.
137. A challenge of a bottom-up method may lie in proxying national accounts concepts as best as possible. For instance, business survey data may provide figures for depreciation, but these may be based on historic costs and may need to be adjusted to better reflect exchange value/written-down replacement cost.
User costs of capital
138. Whichever approach is followed to apply the RVM (top-down or bottom-up), we need to estimate the user costs of capital in deriving resource rent. These user costs consist of two elements: depreciation and a return to capital (excl. the extracted resource)17 used in production. One may ask: why do we include the return to capital used in production in the user costs (see Table 3-1), rather than just depreciation costs? The generally agreed principle across all standards and guidelines is that a return to capital used in production also needs to be included in the user costs, as there are costs involved with investing in and/or owning assets either in the form of financing costs (e.g. interest paid on loans), or, in case of assets already in possession, because one could have made a return by investing the capital elsewhere (opportunity costs).
139. What is the scope of the assets for which the return should be calculated? Capital used in production may consist of AN1 produced non-financial assets (excluding natural resources) and AN2 non-produced non-financial assets (excluding natural resources) such as purchased goodwill and marketing assets as these assets also play a role in generating the operating surplus. It should be noted here that the 2025 SNA only recognises marketing assets under specific circumstances (2025 SNA §11.20) “when a unit is purchased in its entirety, or an identifiable marketing asset is sold to another unit.” While for the exploitation of some natural resources such as subsoil assets brand names (e.g. of oil companies) may be relevant, in many cases, the use of AN2 non-produced non-financial assets (excluding natural resources) will be immaterial (and/or difficult to measure), so in practice the return to capital will predominantly consist of produced assets.
140. Produced assets cover: AN11 fixed assets (excluding produced natural resource) referred to in this Guide also as fixed capital; AN12 inventories (excluding produced natural resource); and AN13 valuables. Valuables represent a store of value and therefore do not contribute to production. Inventories may consist of unsold output or goods stored before use in production.18 The former are the result of production, not a means of production, and can be excluded on conceptual grounds. It is proposed to exclude also the latter on practical grounds as it is challenging to distinguish inventories of goods before use from unsold output and the amounts are expected to be relatively small.
141. Therefore, in this compilation guide, the rate of return to capital used in production (excl. natural resources) will in practical terms rather consist of the “rate of return to fixed capital” (although it is recommended to include AN2, specifically AN23, when the use of these assets is expected to be material for the relevant economic activity).
142. Depreciation (2025 SNA, §7.267) is “the decline, during the course of the accounting period, in the current value of the stock of fixed assets, including (cultivated) biological resources yielding repeat products, owned and used by a producer as a result of physical deterioration, normal obsolescence or normal accidental damage”. Depreciation is commonly measured through the application of a perpetual inventory model (PIM). A PIM (OECD, 2009[3]) requires as inputs the investments made in various types of fixed assets, their service lives (i.e. number of years), and depreciation profiles (e.g. linear or geometric), as well as appropriate price indices for the various types of assets.19
143. In order to value natural resources, it is therefore important to first estimate depreciation for a sufficient number of fixed assets used in production. In terms of sequencing, it is therefore generally recommended to first implement capital measurement using a PIM, see OECD (2009[3]) for further guidance.20
144. The return to capital is usually described as measuring the opportunity cost of money tied up in these assets. The SEEA CF §6.105 notes: “The estimation of capital costs should include both the consumption of fixed capital and the opportunity cost of investing in the assets which is equivalent to estimating a rate of return on the assets.” To calculate the return to capital (excluding natural resources), we need the net stock of fixed assets (i.e. the gross stock minus depreciation) from national accounts by an economic activity (ISIC) and choose a suitable rate of return. The chosen rate of return is then multiplied by the net stock of fixed assets. Because of differing risk premia in different countries, it is not feasible to recommend a common rate of return for capital to be used across countries for market producers.21 Nevertheless, in the interests of enhancing consistency in the compilation of the national accounts, this guide provides below discussion and recommendations on selecting an appropriate rate of return.
145. It is important to emphasise that the rate of return to fixed capital is different from the rate used in the discounting of future resource rents (“the discount rate”). The discount rate should ideally be a risk-free and stable rate (see Section 3.3.2), and a common rate across countries may be chosen in the interests of international comparability. By contrast, the rate of return to fixed capital reflects risk, which differs according to countries’ specific financial and economic circumstances.
146. According to SEEA Central Framework (United Nations et al., 2014[1]) and OECD (2009, pp. 66 (Section 8.3 Rates of return - conceptual considerations)[3]), two approaches can be taken to estimating rates of return to fixed assets for a specific economic activity: an endogenous approach and an exogenous approach. In the endogenous approach the rate would be equal to the net operating surplus (of the activity in focus) divided by the value of the net stock of produced assets (of the activity in focus). This approach implicitly assumes that there is no return attributable to non-produced assets, including natural resources, and hence it is not recommended. It should, however, form an upper bound of the estimated rate of return to fixed assets. The exogenous approach assumes that the expected rate of return to fixed assets is equal to an exogenous rate of return, which is recommended in the SEEA and also recommended here. How should we set the exogenous rate?
147. Different options for calculating an exogenous rate of return to fixed capital exist:
1. Activity-specific rate of return. Such a rate implies the return that an investor actually gets when investing. As remarked in SEEA CF (§5.143), while an activity-specific return may be preferred in theory, it is unlikely that direct measures (e.g. in financial markets) of this risk premium can be found in many countries.22
2. Economy-wide rate of return. The main reasoning behind selecting an economy wide return (i.e. all ISIC sections or total economy S1) is based on opportunity costs: one would look for the return one could have obtained by investing elsewhere in the economy. Using an economy-wide return can also be understood as an expression of a normal or average rate of return (as mentioned in SEEA CF §5.144). A disadvantage of the economy-wide approach is that the net operating surplus of S1 includes also returns to natural resources, which we are trying to measure, and the approach is therefore conceptually unsuitable.
3. “Everything but” the natural resource extraction activities rate of return. This approach suggested by Liu (2023[4]; 2016[5]) can be seen as an improved way of estimating an economy-wide return which avoids the above-mentioned inconsistency. It estimates an economy wide return but excludes the natural resource extraction/harvesting activities in doing so. It should be remarked that for countries that do not have a large extractive sector, the difference between the economy wide and the “everything but” approach will be minimal.
4. A rate based on cost of financing. This method looks not at the actual cost of financing (which is captured in national accounts), but at what it would cost to finance such an investment (e.g. based on average cost of borrowing). However, the cost of financing would provide only a lower bound (if cost of financing were higher than the rate of return, a project would not get off the ground). Also, it may be volatile, which could make the calculations more difficult.
148. The economy-wide return and “Everything but” approach resemble the endogenous approach in deriving a rate of return implicitly by dividing the operating surplus by the stock of produced assets, however they are considered exogeneous methods as they use information external to the activity in question. While all methods have pros and cons, based on general criteria of conceptual soundness and practicality to implement, it is recommended to apply the “Everything but” approach. The activity–specific rate of return method is considered as a Tier 3 approach.
149. When applying the “Everything but” approach, the rate of return is obtained by dividing the net operating surplus of all economic activities minus the activities of the natural resources industry and minus the net operating surplus of non-market production, by the net fixed assets of the same set of activities and sectors. With the 2025 SNA, the net operating surplus for non-market production will no longer be zero, as a net return to capital will be included when estimating output (using the sum of costs approach). However, as the rate of return applied to assets used for non-market production is a risk-free rate, including these activities / sectors in the calculations would lead to an underestimate. It is therefore recommended in practice to use the following scope of activities (in ISIC Revision 5) and sectors: Total economy minus ISIC A (Agriculture, Forestry and Fishing), minus B (Mining and Quarrying), minus D3512 (Electric power generation activities from renewable sources), minus S121 (Central bank), S13 (General government) and S15 (Non-profit institutions serving households).23
150. Based on the various elements described in this section it is possible to derive resource rents for the current and previous accounting periods. As discussed in the next section, these resource rents are averaged/smoothed for the last 3-10 years and used to estimate future resource rents. The projected future resource rents are then used as input for the NPV method of calculating asset values.
3.3.2. Net Present Value method
151. The asset value of the natural resource can be measured by applying the NPV method which converts future resource rents (flows) into a current period estimate of value (stock) by discounting:
Equation 3.1
with the opening stock value at time t; the resource rent generated during year t; r the discount rate, and N the asset life of the natural resource.
The main theory (2025 SNA, Chapter 17; OECD (2009[3])) behind using NPV is that the value of an asset should be equal to the benefits (capital services, described here as resource rents) that we expect to derive from its use over its expected lifetime.
152. While resource-specific recommendations are discussed in Chapters 4 and 5, a number of general recommendations for applying the NPV method are made in this section. We start by looking at the physical asset accounts that underpin the calculation. We then examine each part of the NPV calculation: the estimation of the asset life, projections of future resource rents, and discounting to convert future resource rents (flows) into a current period estimate of value (stock) using the discount rate. Finally, we include a brief discussion of the impact of any new information or changes in assumptions on the NPV estimates when compiling a time series of asset values.
The physical asset accounts
153. It is recommended to underpin the estimates by first compiling a physical asset account for the natural resource. Having a physical foundation is important as it allows deriving separate projections for the levels of future extraction, removals, gross catch24 or harvest (depending on the type of resource, see Table 3‑2) and for the development of prices as reflected in the unit resource rent (see Subsection Projections of resource rent).
154. This guide recommends compilation of physical asset accounts according to the SEEA. As the SEEA applies a broader asset boundary than the SNA, it encompasses the required physical information for national accounts compilation (as a subset). The physical asset account should provide, at a minimum, information about opening and closing stocks and the level of extraction (or removal/harvest, catch/landings), and preferably also further details of changes during the accounting period due to discoveries (if applicable), reappraisals, reclassifications and catastrophic losses (see Table 3‑2). The precise definitions and interpretations of each of these terms will be discussed in Chapters 4 and 5.
Note: “na” means not applicable. * An asterisk indicates that this entry is usually not significant for the resource or is typically not separately identified in the source data. In practice, not all cells that reflect the possibility of an entry here should be shown separately in the published accounts for each type of resource.
Source: Adapted from 2025 SNA Table 35.1, by splitting mineral and energy resources into renewable and non-renewable and excluding water resources.
The asset life
155. Regarding the setting of the asset life of the resource (“N” in the NPV equation), in case of non-renewable energy and mineral resources, the asset life of the resource is obtained indirectly based on opening stocks and the projected levels of extraction (see Section 4.2.2). A key restriction is that the total projected extraction over the asset life of the resource at the start of the accounting period should be equal to the total physical opening stock. It is not common to allow for projections of future discoveries and/or reclassifications.
156. In the case of renewable energy resources, intuitively it might be thought that an infinite asset life is appropriate. However, due to the uncertainty caused by technological developments, climate change and political uncertainty, this is not recommended and default asset lives are proposed for different types of resources (see Section 4.3.2).
157. In the case of forest land, the asset life will depend on the rate of removals compared to (net) growth. Here also a default asset life is suggested (see Section 5.2.2).
158. In the case of fish resources (non-cultivated resources yielding once-only products) (see Section 5.3.2), the assumption about asset life of the resource is derived from an assessment of fish stocks and whether or not they are sustainably managed.
Projections of resource rent
159. To calculate resource rent for each year of the asset’s life in the future, the recommended method is first to calculate a “unit resource rent”. This is done by dividing the resource rent for each past accounting period (year), which is the result of the steps described in Section 3.3.1, by the physical amount of the resource that has been produced or extracted.
160. It is recommended in the standard approach to use an average of actual unit resource rents (called smoothing), as resource rents can fluctuate due to volatility in commodity prices (embodied in output), as a result of which the current unit resource rent will likely not be a good predictor of future resource rents.25 The number of years used for the average will depend on the type of resource, but typically would range from three to ten years.26 It should be noted that although this smoothing is done using past resource rents, it is to be used for projecting future resource rents. As resource rent is not reported in the integrated framework of national accounts, smoothing does not impact on the already reported national accounts data for previous years.
161. In order to apply smoothing of unit resource rents, we need to first bring the unit resource rent of the previous years to the same price level as the current accounting period by applying a price deflator.27
162. Projections of future resource rent are then calculated as the average unit resource rent multiplied by the physical extraction/(net) growth expected in each future year. It is recommended:
To assume that the unit resource rent remains constant in future unless specific policies have been implemented which would allow to estimate a specific unit resource rent path.
To assume a constant level of extraction/(net) growth equal to the last period’s, unless a specific extraction profile is available.28
Discounting future flows of resource rent
163. The last part of the NPV calculation involves discounting to convert future resource rents (flows) into a current period estimate of value (stock) (see Table 3‑3). The choice of a particular discount rate has a major impact on estimates of natural resources value and some impact on estimates of depletion of (relevant) natural resources, which in turn – in the 2025 SNA – feed into net estimates of production and income. As regards options for a discount rate, a set of recommendations is set out in this section, after a brief discussion of discounting.
164. The role of discounting in creating a current period estimate of asset values reflects the finding (e.g. from behavioural economics) that people put a lower value on income they expect to receive in the future than on income they receive now. The further into the future the income will be received, the lower its perceived value in the current period. However, peoples’ perceptions of the value of something in the future differ depending on their views of the importance of the asset being valued as well as factors such as whether they are naturally risk averse or risk takers, their consumption patterns and inflation.
165. A high discount rate will imply that resource rents in future years will not contribute much to the current asset value; therefore the use of higher discount rates will produce lower estimates of the asset value.
166. A choice always needs to be made about whether to express the discount rate in real or nominal terms. The real rate is the underlying rate expressing the value over time in a world of constant prices, while the nominal rate includes inflation and is equal to the real discount rate plus inflation. This is more a practical than a conceptual issue: if resource rent is recorded in current prices, then the discount rate used must be nominal (including inflation); whereas if the resource rent is expressed in constant prices, the discount rate must be “real” (excluding inflation).29
167. Other important choices concern whether the rate used is stable (understood to mean that it is fixed at an agreed rate which stays the same over several reference periods) or varying, for example to reflect changes in markets; and whether the rate is a single rate for the whole of the forecast period, or “declining” (for example, the rates used by the United Kingdom are 3.5% for the first 30 years of the forecast period; 3.0% for 31-75 years ahead and 2.5% for 76-125 years ahead).
168. A survey was conducted of EGNC members, to which ten countries that produce valuations of natural resources responded. The survey found that most use real discount rates (although two countries use nominal rates); and most use stable discount rates in the form of a single rate for the whole forecast period, while one country (the UK) uses a stable and declining discount rate. For those countries using real discount rates, the rates used were around 3.5% to 4%, except in the case of the United States which uses 7% real.
169. The rationales for the choice of discount rate vary between countries: several countries have selected rates based on long-term government bond yields, while the UK refers to a social rate of time preference (SRTP), see Box 3-1 Both of these approaches adopt the view that the discount rate for valuing natural resources should be risk-free. On the other hand, the US uses a discount rate based on the rate of return to capital in the private sector and some countries in Latin America use rates based on the cost of borrowing in (private) markets. These approaches are market-based, including the risks faced by private companies extracting or exploiting the natural resources.
A common starting point in economics when discussing discount rates is the so-called Ramsey formula for the SRTP expressed as consisting of several components (OECD, 2009[3]):
Equation 3.2
with g the trend of the growth in real per capita household consumption; e the elasticity of marginal utility of consumption, i.e. it indicates the percentage change in utility from an additional percent of consumption; P is the survival probability of an individual – it captures the risk that an individual in society is not able to benefit from future returns on an investment – which conceptually is supposed to capture a “rate of pure time-discounting”; w a coefficient that reflects the degree of “selfishness” of present generations vis-à-vis future generations.
Often a simpler version of the equation is applied: STRP = d + e*g although different specifications of these terms can be found. For instance, in the UK’s Green Book (HM Treasury, 2023[6]) d is interpreted as the rate of time preference consisting itself of the pure rate of time preference and an allowance for unpredictable risk together estimated at 1.5%. The e*g part is called the wealth effect and estimated as 2% leading to a STRP of 3.5 %.
OECD (2018, p. 201[7]) refers to the right-hand side of the equation as the welfare-preserving rate of return to consumption, with d interpreted as the utility discount rate.
170. A key difference in rationale behind the choice of the discount rate is whether the chosen rate for valuing natural resource should be a “risk-free” rate or should include a measure of market-related risk of the activity. It is recommended that market-related risk should be captured by the rate of return to fixed capital (see Section 3.3.1) rather than by the discount rate used for valuing natural resource assets.
171. A consequence of the NPV method is that if different discount rates are used by compilers, the results will not be comparable. A compromise approach is therefore suggested consisting of two elements:
A common, stable discount rate agreed by the EGNC which would serve as the central rate against which countries would do sensitivity analysis (and could also be applied by countries that wish to apply this discount rate); and
For countries that wish to select their own rate(s), recommendations on acceptable methods for selecting rates.
172. For the common stable real discount rate for calculating natural resource assets and depletion, it is proposed to use 2%. The main rationale being that the rate should be a risk-free rate based on the yields of long-term government debt securities. The rate may seem low compared with current country practices, but the yields of long-term government debt securities (over 20 years) have on average decreased significantly over the past decades.
173. A rate of 2% for the common stable real discount rate to be used in calculating natural resource assets and depletion is also consistent with the work of Drupp et al. (2015[8]) on social discount rates; and it would be consistent with the rate agreed in the European Ageing Working Group that is used to estimate pension entitlements in the national accounts. It is proposed to review the common rate every five years under the auspices of the ISWGNA, but it should be noted that the review would not necessarily lead to a change in the rate.
174. As the 2025 SNA recommends the split asset approach (see Section 3.6), it would be possible to apply different discount rates for the extractor (producer) (usually in the private sector) and for the legal owner (usually government). This would suppress the relative value of the asset of the producer compared with its share in rents captured. However, applying different discount rates for different units when valuing the same asset is hard to justify and would make the balance sheets more volatile. It is therefore not recommended.
175. In case countries prefer to choose their own rates, the following recommendations are made:
If a government-prescribed discount rate for valuing natural resources exists (such as the United Kingdom's Green Book (HM Treasury, 2023[6])), apply this rate.
Otherwise, the rate used should be a stable rate.
The preferred option is an average yield of government debt securities (real), representing a risk-free rate; in light of the long asset lives of natural resources, the securities should have a maturity of at least ten years; the average should be taken over the last ten years (minimum).
Another alternative would be to apply a SRTP; here it is recommended to apply the so-called Ramsey formula as expressed in OECD (2009[3]) – see Box 3-1.
176. Countries would be free to set their own discount rates as long as they also include a valuation using the common agreed rate as part of the sensitivity analysis. It is recommended that countries also provide an estimate of the cost of depletion using the common agreed rate. The common agreed rate can also be considered as a default (or Tier 2) approach for countries that wish to use it.
177. In all cases, countries should strive for transparency and provide metadata (including the specific rate applied; real or nominal; stable; single rate for the whole forecast period (non-declining) or declining; and rationale for the choice of the rate) when disclosing their results.
Monetary asset accounts
178. The previous section explained how an asset value for a specific year is derived. For national accounts compilation it is also important to describe how to obtain a time series of asset values. For this purpose, it is recommended to compile what SEEA CF calls a monetary asset account for individual natural resources. Recommendations on how to compile such accounts for energy and mineral resources, timber and forest land and fish resources are included in Chapters 4 and 5.
179. The monetary asset account that is used in the calculations of NPV of future resource rents method (Table 3‑4)30 depicts how asset values change during the accounting period as a result of any discoveries, depletion or regeneration of the asset, catastrophic losses, reclassifications and reappraisals. Another possible cause for changes in asset values consists in changes in expected prices, a revision of the discount rate, or due to changes in the extraction path: this is to be recorded as a revaluation during the year in question, so that the balance at the end of the year is affected but not the opening balance.
180. The various elements of the monetary asset account are subsequently included in various parts of the sequence of economic accounts. The monetary asset account forms also the basis for the estimation of depletion, as will be further discussed in Section 3.5 and in the relevant sections of Chapters 4 and 5.
181. In order to compile a monetary asset account, the NPV is applied twice: first to obtain the value of the opening stocks and (a year later) to obtain the value of opening stocks of the next year (which is by definition equal to the closing stock of the period in question).
182. The NPV estimate is to be considered a best estimate at the time it is made and is always forward looking. For instance, if the monetary asset account is compiled as of November 2021 for the calendar year 2020 (opening stock 1 January 2020, closing stock 31 December 2020), then the closing stock estimate would be based on all information available as of November 2021. It would include the actual extraction (or removals, gross catch or harvest) that took place and hence the resource rent generated during the accounting period as well as any other changes that may have occurred during 2020. The opening stock of 2020, which was estimated a year earlier (in November 2020), should not be revised in retrospect as a result of changes in physical estimates,31 as the intention of the monetary asset account is precisely to explain how the asset value changed over time. However, it is recommended to always use the most up-to-date national accounts data when calculating resource rent, as is further discussed in Section 6.3.
183. The treatment of taxes and subsidies related to production in the derivation of resource rent deserves consideration. We will first clarify terminology by placing the definition of taxes and subsidies by the SNA in a broader context. Then we will discuss specific taxes and subsidies. Finally, we will discuss a number of examples to provide further illustration as to how these should be treated.
3.4.1. SNA subsidies in broader perspective
184. There exists no international agreement regarding the definition of subsidies among international organisations:
The World Trade Organisation32 defines a subsidy as:
(a)(1) there is a financial contribution by a government ... where: (i) a government practice involves a direct transfer of funds (e.g. grants, loans, and equity infusion), potential direct transfers of funds or liabilities (e.g. loan guarantees); (ii) government revenue that is otherwise due is foregone or not collected (e.g. fiscal incentives such as tax credits); (iii) a government provides goods or services other than general infrastructure, or purchases goods (iv) a government makes payments to a funding mechanism, or entrusts or directs a private body to carry out one or more of the type of functions illustrated in (i) to (iii) above which would normally be vested in the government and the practice, in no real sense, differs from practices normally followed by governments;
or
(a)(2) there is any form of income or price support in the sense of Article XVI of GATT 1994;
and
(b) a benefit is thereby conferred.
The OECD in its policy work generally adheres to the same definition but calls this “support”; usage of the term subsidy may also vary across studies or domains.
The IMF Government Finance Statistics Manual (GFSM) adheres to the 2008 SNA definition of subsidies but also uses in its policy analysis33 a distinction between explicit and implicit subsidies, where the latter not only includes foregone tax revenue but also unpriced externalities.
The International Energy Agency (IEA)34 distinguishes energy subsidies and consumption subsidies where the latter are measured based on the “price-gap” approach i.e. as the difference between the price paid by consumers and the competitive market price.
185. The SNA definition of subsidies (sometimes called explicit subsidies) is much narrower than the definitions of subsidies/support employed above. It only covers subsidies related to production (which can be broken down into subsidies on products (D31) and other subsidies on production (D39)). The SNA distinguishes the following concepts:
Subsidies on products are subsidies payable per unit of a good or service. (§8.106)
Other subsidies on production consist of subsidies other than subsidies on products that resident enterprises may receive as a consequence of engaging in production (§8.112)
The SNA definition of taxes follows a similar approach, covering taxes related to production (which can be broken down into taxes on products (D21) and other taxes on production (D29)):
Taxes on products are taxes payable per unit of a good or service (§8.93).
Other taxes on production consist of taxes on production other than taxes on products that enterprises incur as a result of engaging in production. Such taxes do not include any taxes on the profits or other income received by the enterprise and are payable regardless of the profitability of the production (§8.102).
186. The SEEA CF (§4.138) uses the notion of “environmental subsidies and similar transfers” to include also social benefits to households (consisting of current transfers to households); investment grants (consisting of capital transfers to finance all or part of the acquisition of fixed assets); other current transfers (consisting of all current transfers other than current taxes on income, wealth, etc., social contributions and benefits, and social benefits in kind); other capital transfers (consisting of all capital transfers except capital taxes and investment grants (examples are legacies, large gifts and donations by households or enterprises intended to finance the purchase of fixed assets)).
187. To avoid confusion, in this guide we will use the term “subsidies” as defined in the 2025 SNA, and we will use “similar transfers” as understood in the SEEA, and “support” as a broad term for subsidies as defined by the WTO.
188. It should be also clarified that government revenue foregone or not collected is excluded in the 2025 SNA definition as these support measures do not consist of actual transactions, but often require the use of counterfactuals, and are therefore difficult to quantify (which is why these are sometimes called off-budget subsidies).
189. In addition, interest payments for loans are split in the national accounts into a payment for a service component (which is measured as financial service on loans and deposits35 recorded as intermediate consumption i.e. as a cost in the production account) and a residual called interest which is recorded as property income in the allocation of earned income account. Again, the national accounts record the income receivable, not whether the loan conditions are more advantageous than expected based on industry practice and therefore might be considered as a subsidy from a policy perspective. In any case, the non-service-fee interest component would not be considered in the derivation of resource rent using the top-down RVM method (see §128), as the GOS is obtained from the generation of income account (i.e. before allocation of income takes place).
190. In summary, we see that the SNA definition of subsidies is narrow in the sense that it is linked only to the production and generation of income, and subsidies can only be received directly by producers. Other support measures are recorded in the national accounts as long as they are transactions, such as current or capital transfers, but they do not enter the calculation of resource rent.
191. While arguments can be made for taking a wider view in considering support measures when estimating natural resource asset values, there are also two additional drawbacks to consider. First, the sectoral accounts where current or capital transfers are recorded do not have a breakdown by economic activity but by institutional sector, which would make it difficult to assign (some) of these support measures to economic activities to estimate natural resources asset values (and hence complicate the use of a top-down estimation method). Second, off-budget types of support require modelling and may have lower reliability.
192. Therefore, for the guidelines we will stick to the SNA (and SEEA CF) definitions and recommendations, recognising that there is a policy interest for standardised broader measures of support. It is suggested that these issues will be further assessed as part of the SEEA CF revision.
3.4.2. Specific taxes and subsidies
193. In the SNA, output is typically recorded at basic prices which are the prices that a producer actually receives. Likewise, intermediate consumption is valued at purchasers’ prices of inputs with the difference resulting in value added at basic prices. This leads to “a measure of gross value added that is particularly relevant for the producer.” (2025 SNA, §7.85)36 To get to GOS and GMI we would deduct remuneration of employees and net other taxes on production (D29-D39). Hence, the GOS/GMI measures what the producer retains as surplus or “profit” after all net taxes have been deducted. For the derivation of resource rent (Table 3-1) the 2025 SNA recommends correcting for specific taxes less subsidies, consistent with the SEEA CF. What is the rationale for doing so?
194. The SEEA CF reasons (United Nations et al., 2014, p. 152 (§5.119)[1]) that if taxes (both on products and other taxes on production) are considered specific i.e. applying only to this specific industry or products produced by the industry and not to a wider set of economic activities, they should be added back when deriving the resource rent. According to the SEEA CF (§5.119): “The deduction of specific subsidies from and the addition of specific taxes to the standard national accounts measures of gross operating surplus are such that the resulting measure of resource rent is neutral to these flows; that is to say, while these flows affect incomes of the extracting industries, they are effectively redistributions within the economy and should not influence the estimated return to the underlying environmental asset”. The situation is completely symmetrical in case of subsidies: we would deduct these subsidies (on products and other subsidies on production) when deriving the GOS/GMI for the derivation of resource rent.
195. The rationale for doing so is that the resource rent should capture the economic value that is generated by using the natural resource in production without redistributions by government. As an extreme example, suppose the extraction of a natural resource generates a lot of output; in case all of this is taxed, we would have a zero GOS, and hence, if we would not adjust the GOS by adding these specific taxes, also a zero-asset value, which would be counterintuitive. In case certain economic activities are not (yet) profitable, society may decide to subsidise them (e.g. renewable energy production). As a result, it will be profitable for companies to engage in them, which will be reflected in the national accounts, as these specific subsidies will be included in the GOS. However, the resource rent that is generated (obtained through deducting these specific subsidies from GOS) may well be low (or even negative), and as a result the economic value of the natural resource will be zero (see Box 3-2; country example from the Netherlands).
196. Specific taxes related to production used in the derivation of resource rent are a subset of taxes on production and imports (D21+D29) in the national accounts, and Specific subsidies used in the derivation of resource rent are a subset of subsidies (D31+D39). We would expect the numbers for specific taxes and subsidies to be lower than the numbers for taxes on production and subsidies on production respectively.
197. The notion of specific taxes and subsidies related to production should be operationalised as follows. First, check if it is clearly a specific tax/subsidy using the following criteria: is it a tax/subsidy related to exploitation of a specific, named natural resource? Examples are: pollution taxes imposed by government on coal mining companies and for subsidies, amounts payable by government to producers of solar panels for each panel sold or to compensate producers (this may include households in their capacity as producers of renewable energy) for losses due to price caps.
198. If it is still unclear whether it is a specific tax/subsidy related to production (examples = carbon tax, mining, North Sea activities), then use the following approach: evaluate its scope at ISIC Section level and apply a threshold of 90% of the total value of the tax/subsidy. For instance, a tax levied only within ISIC Section B (say on Division 05 - Mining of coal and lignite and Division 06 - Extraction of crude petroleum and natural gas) but not outside ISIC Section B would be considered a specific tax for both of these Divisions.
199. As well as specific standalone taxes/subsidies on natural resources production, we may also have variants of more generic taxes levied on natural resource producers at different rates than non-natural resource producers (called sur-taxes). In these instances, the amount of the specific tax can be calculated as the difference between the value of the tax at the specific rate less the value at the generic rate. It is possible that the specific rate could be lower than the generic rate and hence form an implicit subsidy.
Introduction
Renewable energy resources are becoming more and are important for the Netherlands. Renewable energy production has increased from 35 PJ in 2000 to 308 PJ in 2023. Fossil energy extraction (natural gas, oil) on the other hand has sharply decreased over the last year as for most reserves the extraction is no longer socioeconomically feasible. Statistics Netherlands has started to investigate how to monitor and value renewable energy resources. Here the results of a pilot project are presented for the valuation of wind resources on the Dutch part of the North Sea (Statistics Netherlands, 2024[9]).
Methodology
The method applied to value wind resources on the North Sea is based on the net present value of future resource rents. In this example we focus on the derivation of resource rent. A top-down approach is not possible, as no comprehensive data is available for all wind projects at sea. Therefore, a bottom-up approach was applied, which uses multiple data sources including data from separate wind projects to approximate national accounts concepts as best as possible.
The first step is to determine the monetary value of annual electricity production from wind energy resources. This involves multiplying the physical production in kWh from the energy statistics with the basic prices for electricity plus related subsidies on products. In the Netherlands the major subsidy to stimulate renewable energy production is the so called SDE++ subsidy. The subsidy that is provided is calculated as the difference between the cost of the technology that reduces CO2 (the “base rate”) and the market value of the product that is generated by the technology (the “corrective amount”). The base rate is fixed for the entire subsidy period, but the corrective amount is set annually. The unprofitable component decreases when the market value rises, as does the amount of the subsidy you receive. This subsidy for renewable energy is classified in the SNA as a subsidy on products and thus should be included in output in basic prices.
We also need to make a correction for the profile and imbalance costs of wind energy, which refer to the difference between the average electricity price received by the wind producer and the average electricity price on the wholesale market. Intermediate costs include operation and management costs (O&M), such as fixed costs (e.g. insurances) and variable costs (e.g. warranty and maintenance contracts for turbines). These costs thus also include compensation of employees. After subtracting these costs, we arrive at the gross operating surplus.
We consider the SDE++ subsidy as specific subsidy, as this scheme is dedicated to renewable energy generation. In addition, there are certain investment grants that are received by the producers of wind energy, which we also consider as specific subsidies. These subsidies should be subtracted to calculate the resource rent (see also discussion below).
To calculate the capital services (or user costs of fixed assets), we have compiled the capital stock by analysing annual investments in newly installed wind turbine capacity (in MW) and multiplying it by the investment costs per MW of that particular year. These costs include expenditures on foundations, turbines, electric infrastructure, mains connection, construction interest, land acquisition costs and civil engineering. Since investments take place some years before the completion, the investments are spread over multiple years. We then construct the fixed capital stock using the perpetual inventory method and assume a linear depreciation of 20 years. This fixed capital stock is necessary to calculate the return to capital and the depreciation, which together form the capital services rendered by an asset. The return to capital can be measured using the weighted average cost of capital (WACC), which the PBL Netherlands Environmental Assessment Agency publishes for wind energy in general (PBL, 2022). Subtracting from the output (including subsidies on products), intermediate costs, specific subsidies and capital services, results in the resource rent estimates.
Results
Renewable energy production from wind resource has increased significantly over the past years (Figure 3-1). The construction and operationalisation of new wind parks resulted in upward shifts of the annual production.
The results for the resource rent calculation for wind energy on the Dutch part of the North Sea are shown in Table 3‑5. After subtracting for the specific subsidies (SDE++ and investment grants) the resource rent is negative for all the years. This is not an unexpected result as the main purpose of these subsidies is to subsidise the unprofitable component of different renewable energy technologies.
We have also calculated what is sometimes called a “social resource rent” (Section 3.4.2) namely the resource rent that would be obtained by not correcting for specific taxes/subsidies. Here, the idea is that the distribution of income function of the government indicates that society as a whole has a certain preference for in this case renewable energy production. The reduction in negative externalities inherent in the use of wind energy will be expressed in the valuation and this valuation will be closer to the social preferences regarding energy production. Not correcting for the specific subsidies results in a positive resource rent for wind energy at sea, except for the last year. This is to be expected as wind turbine operators need to be able to generate profits to stay in business.
Another observation is that the calculated resource rents are very volatile. Between 2019 and 2021 the production (in KWh) of offshore wind energy increased sharply, but this also leads to an increase in intermediate consumption, which is based on the physical production and capacity as described in the methodology. However, as a result of higher electricity prices less subsidies have been provided, which has caused the social resource rent to decline and become negative in 2021.
200. The 2008 SNA already included depletion of natural resources (as an economic disappearance of non-produced assets, K21) in the other changes in volume of assets account. The 2025 SNA includes depletion as a cost of production in the current accounts, thereby impacting net aggregates such as net domestic product, net national income and net saving. This is an important change as the 2025 SNA will place greater focus on net measures. Countries are encouraged to show depletion of natural resources in addition to depreciation, so that users can understand its impact on net measures. It is therefore important to obtain accurate measures of depletion.
3.5.1. Defining depletion
201. The 2025 SNA defines depletion as follows:
Depletion .., in physical terms, represents the decrease in the quantity of the stock of a non-produced natural resource over an accounting period that is due to the extraction of the natural resource by economic units occurring at a level greater than that that of its growth; in monetary terms, it corresponds with the decline in future economic benefits, due to extraction in excess of its growth, that can be earned from a resource, the value of which is based on the physical flows of depletion using the price of the natural resource in situ. (2025 SNA, §7.286)
This definition covers both non-renewable resources such as minerals and oil as well as biological resources that are able to reproduce and grow such as fish resources. When it comes to produced biological resources, depletion should be understood as a reduction in the capacity of the non-produced underlying asset to generate future economic benefits (e.g. timber) due to overextraction of the produced overlying asset. The overextraction is always assessed in quantitative terms. Although the 2025 SNA supports the view that agricultural land can be subject to depletion (see §11.180, §11.197 and §35.62), this guide does not recommend measuring depletion of agricultural due to practical measurement challenges.37
202. As shown in Table 2-1 it should be noted that renewable energy resources, radio spectra and permits to use natural resources are not subject to depletion. The depletion of water resources is not addressed in the 2025 SNA and therefore its valuation is not recommended, and any value changes are to be recorded as other changes in volume. Depletion also does not apply to produced biological resources yielding repeat products, which are subject to depreciation, and bare land.
203. It is important to distinguish between depletion and degradation.38 Depletion should be understood in quantitative terms - the using up of assets in production39 - while degradation captures the qualitative dimension. The 2025 SNA notes (§13.32) “All degradation of land, water resources and other natural assets due to less predictable erosion and other damage to land from forest fires or other natural events should also be considered as a quality change, and thus recorded in the other changes in the volume of assets and liabilities account. However, changes in the value of, for example, forest land, resulting from a change in the regenerative potential due to its use in production is to be recorded as (negative) depletion.” For example, agriculture or forestry practices may consist in use of fertilisers or other chemicals to boost yield which may lead to (predictable) erosion and hence depletion. At the same time, this may also cause a reduction in quality of the land (e.g. due to chemical pollution) or surrounding natural resources such as water (e.g. eutrophication), which may be picked up in associated market prices. This reduction in value would not be considered as a cost of depletion but as degradation to be recorded as other changes in volume.
204. In case of non-renewable (mineral and energy) resources, the 2025 SNA (§7.287) is clear that while discoveries may permit the ongoing extraction of the resources, these increases in volume are not considered regeneration and cannot be used to offset depletion.
205. In the case of forest land used for timber production or non-cultivated fish stocks that are not sustainably managed, depletion is assessed by comparing removals (for timber) or catch (for fish) with the natural growth of the resource. SNA 2025 (§7.288) states: “Only the amount of extraction that is above the level of growth is recorded as depletion; in the case the amount of extraction is below the level of growth (for example, to allow the resource to regenerate and thus allow higher future extraction), it is recorded as negative depletion.” As further detailed in Section 5.2, when it comes to timber resources the 2025 SNA distinguishes between work-in-progress (the standing timber) and the underlying asset (the forest land), which captures the capacity of the land to continue providing timber into the future. Overharvesting may have two effects: a net withdrawal from work-in-progress and a reduction in the value of the non-produced forest land in case future growth of timber is reduced, which is recorded as a cost of depletion. Depletion can therefore only arise as a result of overharvesting, which is assessed by comparing the removals (or catch in case of fish resources) with the growth of the resource. Likewise, regeneration only applies in case of underharvesting, ensuring a symmetric treatment.
Source: SEEA CF Figure 5.2.
206. The 2025 SNA provides further details on how to measure the growth of biological resources directly drawing on the SEEA CF:
While the rates of extraction can be observed directly, measurement of the rates of growth can be complex and usually requires consideration of biological models. These models will usually account for both the structure and the size of biological resource populations; and exhibited by their general form, when the stock or population of the specific type of resource is small, the rate of growth will be small but, as the population increases, the rate of growth will also increase. Eventually, as the population within a given area reaches the carrying capacity of the area, i.e., as the density reaches a maximum, the rate of growth in the population will slow substantially. (2025 SNA, §7.289)
The 2025 SNA continues to explain that depletion (and regeneration) can be assessed by comparing the harvest with the sustainable yield instead of the mere growth, as illustrated in Figure 3-2:
Based on this general model, for any given population, it is possible to calculate the number of animals or volume of plants by age or size class that may be removed from the population without affecting the capacity of the population to reproduce itself (i.e., opening stock equals closing stock). In effect, there is a “surplus” or excess that can be harvested from the existing stock. In biological models, this surplus is known as the sustainable yield. The level of the sustainable yield rises and falls in line with the overall size and structure of the population. For example, in populations where the growth rates are low, the sustainable yields are also low. It is noted that the same level of extraction will have a different relationship to the sustainable yield depending on the population size. (2025 SNA, §7.290)
In these guidelines, as further discussed in Chapter 5, we treat the use of population models for deriving depletion (and regeneration) as advanced methods.
3.5.2. Deriving depletion
207. A key innovation introduced by the SEEA CF and followed in the 2025 SNA is to anchor the cost of depletion in the changes in physical units: when there is no physical extraction there can be no depletion.
208. The SEEA CF method for estimating depletion consists of the following steps (based on Annex A5.1). First it is important to see that the resource rent can also be written as the physical extraction multiplied with the unit resource rent:
Equation 3.3
with the physical extraction (in tonnes or cubic meters) and the unit resource rent.
209. Next, we estimate the so-called price of the resource in situ by expressing the asset value as follows:
Equation 3.4
with the price of the resource in situ, which is obtained by estimating the asset value by the total physical stock of the resource at that point in time.
210. This allows us to decompose the change in asset values into two components:
1. A change in the price of the resource in situ (multiplied with the average of the opening and closing stocks), and
2. Changes in physical stocks (due to extraction, discoveries etc.) multiplied with the average in situ resource price.
To see this, we can write:
Equation 3.5
However, we could have also decomposed this (see orange instead of blue) as
Equation 3.6
As both decompositions are equally valid, following SEEA CF we average them whereby we obtain:
Equation 3.7
where denotes the average of the price in situ during the accounting period, and the average physical stock during the accounting period.
It is important to see that this is an exact decomposition, the opening stock value plus the changes in stocks equal the closing stock value. Finally, to obtain the cost of depletion, we can further disaggregate the first element as:
Equation 3.8
with = growth (in case of biological resources able to reproduce) at time t, = discoveries at time t; = reappraisals at time t, = extraction / harvest at time t.
The cost of depletion then becomes:
Equation 3.9
In case we have a situation of regeneration or negative depletion. In case of non-renewables equation 3.9 simplifies to the physical extraction during the accounting period, multiplied with the average price of the resource in situ:
Equation 3.10
211. It is important to realise (as discussed in Section 3.5) that the point of departure for estimating depletion are the two NPV estimates that are made at different points in time with different information including different assumptions about the future flow of resource rents.
212. It can be readily seen that due to discounting, in principle: .40 This supports the interpretation that the resource rent consists of an income (net return to natural asset) and a depletion element as shown in Table 3-1. By subtracting the cost of depletion, only the income element of the resource rent is included in net measures.
213. However, due to smoothing and averaging (between beginning and end of accounting period), in some cases leading to a situation in which depletion is higher than resource rent implying a negative net return to the natural asset. If this arises, no adjustment is suggested, as a negative net operating surplus (after deducting depreciation – and in this case also depletion) can occasionally happen. For example, farmers may have negative value added in some years, one does not introduce adjustments for such instances either.
3.5.3. Recording of depletion in the sequence of economic accounts
214. The 2025 SNA recommends (§1.26) recording depletion of natural resources as a cost of production in the production and income accounts. The attribution of the costs to the legal owner and the extractor is recorded in line with the split asset approach (2025 SNA, 4A.34), and illustrated in 2025 SNA Table 27.1, and consistent with the reasoning of Chapter 5 of the SEEA CF, which says:
Record the total cost of depletion in the production and generation of income accounts of the extractor as deductions from value added and operating surplus. This ensures that the analysis of extractive activity and economy-wide aggregates of operating surplus and value added fully account for the cost of depletion. Further, since the government has no operating surplus in regard to the extraction activity, not recording depletion in the production account of the government ensures that estimates of government output (which are calculated based on input costs) are not increased owing to depletion;41
Record the payment of rent from the extractor to the government in the allocation of primary income account. This entry is the standard national accounts entry;
Record an entry, entitled “Depletion borne by government”, in the allocation of primary income account to reflect (i) that the rent earned by the government includes the government’s share of total depletion which must be deducted to measure the depletion-adjusted saving of government; and (ii) that the depletion-adjusted saving of the extractor would be understated if the total amount of depletion were deducted in the extractor’s accounts. Another way of viewing this entry is to consider that the rent earned by government must be recorded net of depletion (i.e., depletion-adjusted rent is derived) in the derivation of depletion adjusted saving for government. (SEEA CF, §5.220)
215. Regeneration is recorded in full in the production and generation of earned income account of the extractor (as negative depletion), with part of the negative costs allocated to the legal owner in the allocation of earned income account (in line with their respective shares of the asset value).
216. The recording of depletion in the sequence of economic accounts will be illustrated in Chapter 6.
3.5.4. Depletion and tiers
217. In the absence of physical asset accounts (Table 3‑2) to estimate depletion in physical and then monetary terms (as in the standard Tier 2 approach described above), the recommended Tier 1 basic method consists in measuring depletion based on the asset value divided by the asset life (for opening and closing values). The asset value would need to be obtained, in the absence of suitable market prices, by applying the NPV of future resource rents, without use of physical data. This implies that the resource rent is not split into a unit resource rent and extraction (Equation 3.3) but that projecting of future resource rents needs to be undertaken directly in monetary terms.
218. This should provide in general a good approximation of the standard method. To see this, if we are assuming a constant level of extraction (or removal or catch), the asset life consists of the total physical stock divided by the extraction level. Therefore, we can write:
Equation 3.11
219. This Tier 1 approach is relatively easy to apply, it requires in addition to the asset value a (default) asset life. For instance, if we have a resource which generates a resource rent of 100, a discount rate of 2%, with an expected asset life of ten years, using Equation 3.1 we can estimate the asset value to be 916. Due to the assumed asset life of ten years, we find that the cost of depletion is 91.9 (in the first year).
220. An example is shown in Chapter 5 for fish resources. However, when there are significant changes during the accounting period (e.g. discoveries, reclassifications, price changes) that cause the extraction level to change, the depletion estimate derived in this manner will deviate from the standard (Tier 2) approach, which is preferred.
221. An advanced method (Tier 3) for estimating depletion for renewable resources such as forest land underlying timber production and non-cultivated fish resources may consist in applying a biophysical model. Such a biophysical model would describe the population in terms of different age brackets in order to estimate the sustainable yield and would allow for making dynamic projections of stocks. The assessment of depletion would be based on a comparison of harvest with the sustainable yield. This will be further discussed in Chapter 5.
222. 2025 SNA Chapter 27 section C (§27.15-27.19) specifies that there are three basic types of treatments when it comes to rights to use a natural resource:
1. The legal owner may permit the resource to be used to extinction. In this case, the economic ownership is transferred which is recorded as the sale of the resource.
2. The legal owner can extend or withhold permission for the use of the resource from one year to the next. In this case, the legal owner retains economic ownership which is recorded as a resource lease (with rent payment).
3. The legal owner may allow the resource to be used for an extended period of time in such a way that in effect the user controls the use of the resource during this time with little if any intervention from the legal owner. In this situation the economic ownership of the natural resource is shared, leading to the so-called split asset approach.
223. What is important is that “under all three options the total value of the natural resource should be recorded against the relevant natural resource class, such as land or mineral and energy resources. The exception to this treatment concerns radio spectra.” (2025 SNA §27.20) It should be noted that in 2008 SNA rights (permits, licences or quotas) to use the natural resources could qualify as assets in their own right (under certain circumstances, for instance when transferable and valid for more than one year), which lead to a separate registration under AN21 contracts, leases and licenses (permits to use natural resources AN222).42 This treatment does not apply in 2025 SNA (with the exception of radio spectra where the permit could be recorded under certain circumstances as a separate asset from the value of the radio spectra itself and recorded under AN392).43 The main reasons behind this change are as follows. First of all, to ensure comparability of balance sheet data across countries. If country A has instituted a quota regime for its fish stocks, and country B has not, it would be odd to see country’s A’s balance sheet consisting of permits to use natural resources, and country B’s consisting of fish resources. Second, due to the physical understanding of depletion, it is hard to conceive of permits as subject to depletion, while natural resources clearly are.
224. In case of the third situation described above, the 2025 SNA recommends (§4A.34) the split-asset approach i.e. to apportion the value of the natural asset to the legal owner and the extractor (producer) of a natural resource based on their shares of the resource rent (see example in Box 3-3). The difference between the resource rent and relevant payments received by the legal owner (to be recorded as rent D45, see below) provides the foundation of the split-asset approach. The asset is split based on how much of the resource rent is captured by the legal owner (usually government but may also be households for instance in case of indigenous peoples that receive rents (royalties) from extractors). As discussed in Section 3.4. it is recommended to calculate and compare the resource rent with rent payments and apply a threshold: if rent payments are on average more than 90% of the resource rent, it may be assumed that economic ownership remains with the legal owner and there is no need to split the asset between legal owner and extractor.
225. If rights to access the natural resource are tradeable, one could expect the value of the rights to be equal to the (net present value of) the difference between resource rent and rent payments, as this would be the surplus value that the right provides to the extractor.
226. The split asset approach generally applies to all types of non-produced natural resources (as long as conditions of the third situation are met), although as explained in 2025 SNA Chapter 27 occurrence may differ between resources. Agricultural land is frequently subject to a resource lease, although splitting of assets in case of forest land is possible (2025 SNA §27.27). In case of mineral and energy resources “the legal owner ... often does not appropriate the full resource rent” (2025 SNA §27.31) which would lead to splitting assets. Also in case of renewable energy resources (2025 SNA §27.32) and in case of fish stocks (2025 SNA §27.40) splitting of assets may apply. In case of water resources, the SNA deems splitting of assets less likely, and this guide therefore does not recommend splitting.44 Splitting of assets does not apply to work-in-progress.
Suppose the resource rent is calculated as 100 and that the part appropriated by the legal owner, in this case the government, is 40 (equals rent, D45). Assume further an asset life of 100 years and discount rate of 5%. Therefore, the total asset value = 2 000.
The respective shares of resource rent are therefore: 60 for the extractor (producer) and 40 for the government (i.e. part appropriated by government = 40%). The total asset value will therefore be split based on these respective shares of resource rent (60 vs. 40) as follows:
Producer: 1 200, Government: 800.
The cost of depletion will be attributed accordingly. Assume depletion cost of 80 (NP1depl). First, the full 80 will be recorded in the production and generation of income account of the producer. Second, 32 (D45depl (-)) i.e. with a negative sign will be allocated to the government in the allocation of earned income account as depletion borne by the legal owner (i.e. reducing the rent received by government (40) by the amount of depletion (32) assigned to government), while 32 (D45depl (-)) (again with a negative sign) will be shown with the producer (i.e., reducing the rent paid (40) by the amount of depletion assigned to government (32)). Third, the capital account then shows depletion of 48 for the producer and 32 for the government (with negative signs, on expenditure sides of the account).
Thus, in 2025 SNA there is an alignment (consistency) between depletion and economic ownership of a natural asset (split in the same proportion as depletion) in the balance sheets between producer and legal owner (government), which was not the case in 2008 SNA.
It is important to realise that where the legal owner is the government, total government revenue from exploitation of the natural resource can be larger than the resource rent, for instance due to the application of general taxes. Here we are interested in estimating only the part of the resource rent appropriated by government, based on rent received (D45), while other revenues are not to be considered.
227. How should the share of the resource rent to the legal owner be measured? The EGNC considered initially the following 5 components which were mentioned in earlier Eurostat guidance (Eurostat 2003):
Specific taxes on products
Specific other taxes on production
Rent (royalties)
Specific taxes on income
Dividend payments
228. As regards the definition of rent (D45), a wide range of different types of transactions described as royalties or natural resource related taxes may exist in countries, oftentimes not explicitly referred to as rent in source data. When compiling national accounts, it is therefore recommended to apply wide search criteria.45 The 2025 SNA stipulates that “any payments made by the user/extractor of a non-produced natural resource to the owner of the natural resource, which are linked to the use/extraction of that resource, in particular to the quantity and/or value of that resource, should be recorded as rent. These would include, for example, royalties, sur-taxes, and permits” (§8.173). This implies that such payments that may be classified as taxes (product, production, or income) in other frameworks should be recorded as rent (D45) in national accounts, as long as government is the legal owner of the natural resource.46 In addition, although the 2025 SNA classification of distributive transactions does not require further breakdowns of rent by a natural resource, for the compilation of a split asset approach, it is necessary to have such breakdowns, as also recommended in Guidance note WS.12 Environmental classifications (SNA Update, 2023[10]).
229. The inclusion of dividend payments (D4121) was considered as in many countries’ government has a stake in companies managing or investing in extraction activities. However, this could cause complications as mentioned in Liu (2023[4]): “For instance, if part of dividends from the oil and gas companies are also sent to their foreign shareholders as resources in the rest of the world account of primary incomes and current transfers, it leads to a creation of foreign ownership of Norwegian oil and gas resources in proportion to their respective share in the total resource rent, if the split-asset approach is strictly respected.” The inclusion of dividend payments in the estimate of government’s share of resource rent is therefore not recommended.47
230. Therefore, in principle, in case government is the legal owner of the resource, it should be sufficient to use rent payments (D45) to determine the government’s share.
231. In case government is not the legal owner of a natural resource, but households or the private sector, specific taxes may occur that would be recorded under D21 or D29 in the national accounts. This could lead to a situation where the value of the resource to the legal owner and extractor would not align with the asset value obtained through application of the residual value method. Therefore it is recommended when government is not the legal owner, not to correct for specific taxes less subsidies when deriving the resource rent (see Table 3-1). Likewise, in case of specific subsidies, these should not be deducted when deriving resource rent, as long as government is not the legal owner.
232. The main rationale for this recommendation is that such a valuation approach better measures the exchange value of the asset to the owner. For instance, in case an extractor (or investor) foresees having to pay specific taxes, he would subtract the specific taxes as costs when calculating his profit i.e. not treat specific taxes different from other production taxes. As a result, in concept, the value of the resource would be equal to the value measured by a tradable license or permit to extract.48
233. Resource rents and the part of resource rents captured by the legal owner will change from year to year, and this would make the resulting split asset values volatile. It has already been recommended to smooth the unit resource rent used for projecting the resource rent flows required to calculate asset values by the NPV method (see Section 3.3.2); and here it is recommended to use average resource rent shares to split asset values as well. For the averaging, it is recommended to be consistent with the averaging (smoothing) that is used in estimating the asset value itself (e.g. between three and ten years).49 This implies that the average share will be updated on an annual basis, but due to the smoothing is expected to change very little. Any resulting changes in split asset values due to changes in the share are to be recorded as other changes in volume.
234. The same average share is to be used to split the cost of depletion, revaluation and other changes in volume of assets and liabilities elements.
235. The 2025 SNA recommends recording ownership changes of parts of the resources other changes in volume of assets and liabilities, and not as capital transfers. Likewise, changes in rents (royalty) regimes should be recorded as other changes in volume. Application of the split-asset approach will be further illustrated in Chapter 6 for different natural resources.
236. Existing compilation experience by countries has shown that negative resource rents often occur. They can arise due to a number of reasons, but it is always important to further investigate, as, in theory, persistent negative resource rents imply that there is no economic rationale for exploiting a specific resource for society. It is recommended to ask the following questions to ensure that they are “genuine” negatives:
Negative resource rents may be due to high values for user costs of capital, which are deducted from GOS and GMI to arrive at resource rent. Standard assumptions of the PIM such as average service lives should be checked. The service life being used can be compared with the range and recommended average service life (as stated in (UNECE, 2024[11])) and possibly adjusted to a longer service life thereby reducing user costs.
What is the ratio between depreciation (CFC) costs (coming from a PIM model) and depreciation costs as reported in business accounts (e.g. company reports)? It is important to realise that depreciation costs are not actually incurred, they are estimated. In case depreciation costs from national accounts are significantly above depreciation costs from business accounts (e.g. by more than 50%), it may be necessary to reassess the assumptions that are being used in the PIM model which will lead to changes in the resource rent.
Do stranded assets exist? Depreciation may also be overestimated due to the existence of stranded assets. Equipment (e.g. unused oil rigs) linked to stranded assets should be identified and be removed from the PIM calculations when estimating the user costs of produced assets.
Is the rate of return used for fixed capital realistic? The rate applied (based on the “Everything but” approach, see Section 3.3.1), could be compared with the rate of return in ISIC Section C Manufacturing. If the difference is more than 2 percentage points, the rate should be reassessed.
Would the resource rent still be negative if (net) specific subsidies were not deducted? In some situations, negative rents can exist if an economic activity is (heavily) subsidised by government for other socio-economic or environmental reasons, such as maintaining employment or due to positive impacts on related economic activities (see Section 3.4.2). An example could be small scale (e.g. artisanal) fishing, where government may decide to subsidise the activity because of its positive impacts. If the resource rent would still be negative if (net) specific subsidies were not deducted, there may be a problem with the data or estimation process.
Would the resource rent still be negative when using a price at the upper range of the commodity prices observed in the past ten years? Even after smoothing (say of three to ten years), negative resource rents can occur due to a prolonged slump in commodity prices, but if resource rents are always negative there may be an error somewhere in the compilation process.
In addition to applying the checklist, compilers are encouraged to conduct sensitivity analyses to assess how small fluctuations in key variables (e.g. prices, extraction volumes) could affect the rent estimates. This will help determine if the negative rents are persistent or a result of temporary anomalies. If after all these checks, negative resource rents still persist, it is recommended for the purpose of projecting future resource rents to set the unit resource rent to zero. This means that the asset value becomes zero, i.e. does not have (at the moment) an economic value from a national accounts point of view, although it may be bringing many other (non-economic) benefits, especially in case of biological resources (e.g. certain ecosystem services such as air filtration, water regulation and purification, flood mitigation, soil erosion control, noise attenuation and global climate regulation that contribute to a range of benefits including health).
237. A zero-asset value provides a signal to policy makers that the current way the resource in question is managed is suboptimal. Which policies to implement to change the situation would be a question for modelling, for scientists and experts to advise on, and eventually for policy makers to decide. For instance, Grimsrud et al. (2015[12]) calculate a time series of resource rent generated in Norway by fishing which was negative for the period 1984-2014 apart from 2010-2011. In the paper they model what the resource rent could have been in a counterfactual scenario of optimal management (using an optimisation model) and find that resource rent would be much higher (and positive).
238. In case the asset value is set to zero, this implies that there will also be no cost of depletion (but there can still be depletion in physical units). In fact, allowing for negative asset values could lead to a situation where the physical using up of the asset would lead to negative depletion (hence regeneration) which is clearly counterintuitive. Therefore, it is recommended not to record negative asset values but to compile physical SEEA based asset accounts to complement monetary national accounts.
Aggregating negative and positive resource rents
239. When following the bottom-up approach, it is possible that for some individual resource (e.g. solar in case of renewable energy) a negative resource rent is derived, while others would generate a positive resource rent. This raises an issue of aggregation when projecting future resource rents (for the resource as a whole): should the negative resource rents be set to zero before aggregating, or should we aggregate the sum total of negative and positive rents?
240. As national accounts compilation should be scale invariant (that is, we should obtain the same result independent from the level of disaggregation), and we would obtain a different result when putting negative resource rents to zero before aggregating, it is recommended to aggregate positive and negative rents. If the resulting rent is negative, it is recommended to investigate based on the checklist mentioned above.
Valuation methods (Section 3.2)
The preferred valuation method for natural resources is using market prices or market-equivalent prices. When these are not available, it is recommended to apply the Net Present Value (NPV) method of future resource rents.
Resource rent measures the surplus value accruing to the extractor (producer) of a natural resource and is equal to the SNA concept of capital services. Resource rent and rent (D45) are two different concepts.
The recommended method for estimating resource rent is the residual value method (RVM). The RVM measures resource rent by deducting from output all costs related to the extraction of the resource. These costs include the user costs of capital (excluding the natural resource one is trying to value) used in production. A distinction can be made between a top-down method and a bottom-up method for applying the RVM.
When rights to exploit a natural resource have been allocated by an open competition process or when rights are transferable, the value paid by the extractor may provide a proper valuation of the resource (i.e. of the part covered by the right). It is however recommended to calculate and compare the resource rent with rent payments (when the contract is annual) and apply a threshold: if rent payments are on average more than 90% of the resource rent, it may be assumed that economic ownership remains with the legal owner and there is no need to split the asset between legal owner and extractor.
User costs of capital (Section 3.3)
User costs of capital consist of two elements: depreciation and a net return to capital used in production.
Capital used in production may include both produced capital and non-produced capital such as purchased goodwill and marketing assets, although the latter are often immaterial and/or difficult to measure. The scope of produced assets is restricted to fixed assets on practical grounds, unless the use of non-produced assets is expected to be material.
It is important to derive high-quality estimates for the fixed capital used in production, as input for deriving the costs of capital, by using the Perpetual Inventory Method – see (OECD, 2009[3]).
For estimating the rate of return to capital, it is recommended to apply the “Everything but” approach, which derives the rate of return by dividing the net operating surplus of all economic activities except the natural resource extraction and/or harvesting industries and minus the net operating surplus of non-market production by the value of net fixed assets of the same set of activities and sectors. Use of an activity specific return is considered an advanced method.
Projecting future resource rent (Section 3.3)
Projections of future resource rent are calculated as the average unit resource rent (a price) multiplied by the physical extraction or (net) growth (in physical units) expected in each future year (over the expected life of a natural resource).
The average unit resource rent should be based on the calculated unit resource rents for a number of years (called “smoothing”). The exact number of years will depend on the type of resource, but typically would range from 3-10 years.
It is recommended to assume that unit resource rents remain constant in the future unless specific policies have been implemented which would allow to estimate a specific path of future unit resource rents.
Furthermore, it is recommended to assume a constant level of extraction or (net) growth equal to the last period’s, unless a specific extraction path is available.
Discounting (Section 3.3)
Market-related risk is assumed to be captured by the rate of return to fixed capital, rather than by the discount rate used for valuing natural resource assets.
For the common stable real discount rate for calculating natural resource assets and depletion, it is proposed to use 2%.
In case countries prefer to choose their own rates, the following applies:
If a government-prescribed discount rate for valuing natural resources exists, apply this rate.
Otherwise, the rate used should be a stable rate.
The preferred option is an average yield of government debt securities (real), representing a risk-free rate; in light of the long asset lives of natural resources, the securities should have a maturity of at least ten years; the average should be taken over the last ten years (minimum).
Another alternative would be to apply a social rate of time preference; here it is recommended to apply the so-called Ramsey formula as expressed in OECD (2009[3]).
Asset valuation (Section 3.3)
It is encouraged to compile SEEA monetary asset accounts (as an intermediate step) for individual natural resources, which depict opening and closing stocks and a full reconciliation of how asset values change during the accounting period as a result of any discoveries, depletion or regeneration of the asset, catastrophic losses, reclassifications and reappraisals. Information from the monetary assets accounts is then used to compile the integrated framework of national accounts.
Specific taxes and subsidies (Section 3.4)
When the government is the legal owner of the natural resource, specific taxes less subsidies (both on products and production) should be included in the calculation of resource rent. The main criteria to identify them is whether it is clearly a tax/subsidy related to exploitation of a specific, named natural resource.
Depletion of non-produced natural resources recorded as cost of production (Section 3.5)
Depletion in physical terms is the result of the extraction/harvest of a natural resource by economic units occurring at a level greater than that of regeneration.
Physical asset accounts according to the SEEA encompass the required physical information for national accounts compilation. Due to various synergies, this guide recommends compilation of SEEA accounts for relevant natural resources in case these do not yet exist.
In the standard approach, depletion in monetary terms (i.e. cost of depletion) is derived by multiplying the physical depletion with the average price of the asset in situ (the price of the asset before extraction, as opposed to the price received for the sale of the resource after extraction).
Split asset approach (Section 3.6)
In case of shared economic ownership between the legal owner (usually government) and the extractor, the value of natural resources should be split in line with the appropriation of resource rent.
In case the government is the legal owner of the resource, it should be sufficient to use rent payments (D45) to estimate the government’s share. In case the government is not the legal owner, no correction should be made for specific taxes less subsidies when deriving the resource rent.
In order to split the asset values, compilers should use an average of the calculated shares by applying the same period as the one used for smoothing unit resource rents.
Negative resource rents (Section 3.7)
When applying the RVM method, negative resource rents may occur. A check list is provided to assess whether negative resources rents are “genuine” negatives.
If after all these checks, negative resource rents persist, it is recommended – for the purpose of projecting future resource rents – to set the unit resource rent to zero. This means that the asset value becomes zero, and that there will be no cost of depletion.
References
[8] Drupp, M. et al. (2015), Discounting disentangled.
[18] Eurostat (2022), Delineation between resource taxes and rent, Eurostat.
[2] Eurostat and OECD (2015), Eurostat-OECD compilation guide on land estimation, https://ec.europa.eu/eurostat/documents/3859598/6893405/KS-GQ-14-012-EN-N.pdf.
[12] Grimsrud, K., L. Lindholt and M. Greaker (2015), Resource Rent in Norwegian Fisheries. Trends and policies, Statistics Norway, Research Department (Papers No. 827).
[6] HM Treasury (2023), The Green Book (Annex A6. Discounting), https://www.gov.uk/government/publications/the-green-book-appraisal-and-evaluation-in-central-government/the-green-book-2020#a6-discounting.
[14] ISWGNA (2025), Guidelines for estimating the net return to capital for non-market production, https://unstats.un.org/UNSDWebsite/statcom/session_56/documents/BG-3a-ISWGNA_BG_Net_return-E.pdf.
[15] ISWGNA (2023), Background document to the report of the Intersecretariat Working Group on National Accounts on the recommendations for the update of the 2008 SNA, https://unstats.un.org/UNSDWebsite/statcom/session_55/documents/BG-3f-NationalAccounts-E.pdf.
[4] Liu, G. (2023), Testing the split of economic ownership for petroleum resources in Norway, Documents 2023/24, Statistics Norway, https://www.ssb.no/en/nasjonalregnskap-og-konjunkturer/nasjonalregnskap/artikler/testing-the-split-of-economic-ownership-for-petroleum-resources-in-norway.
[5] Liu, G. (2016), The wealth of Norwegian raw oil and natural gas: 1970-2015. Statistics Norway, https://www.ssb.no/en/energi-og-industri/artikler-og-publikasjoner/_attachment/286645?_ts=158cf3323a8.
[7] OECD (2018), Cost-Benefit Analysis and the Environment: Further Developments and Policy Use, OECD Publishing, Paris, https://doi.org/10.1787/9789264085169-en.
[3] OECD (2009), Measuring Capital - OECD Manual 2009: Second edition, OECD Publishing, Paris, https://doi.org/10.1787/9789264068476-en.
[13] SNA Update (2023), WS.11 Guidance note on the treatment of renewable energy resources as assets, https://unstats.un.org/unsd/nationalaccount/SNAUpdate/GuidanceNotes.asp.
[10] SNA Update (2023), WS.12 Environmental Classifications, https://unstats.un.org/unsd/nationalaccount/SNAUpdate/GuidanceNotes.asp.
[9] Statistics Netherlands (2024), SEEA Ecosystem account Dutch North Sea, 2023, https://www.cbs.nl/en-gb/custom/2024/12/seea-ecosystem-account-dutch-north-sea-2023.
[11] UNECE (2024), “Final report, Item 3 of the provisional agenda Improvement of measures of consumption of fixed capital ECE/CES/GE.20/2024/6”, in Directors of Macroeconomic Statistics Task Force on fixed assets and estimation of consumption of fixed capital under European System of Accounts 2010, presented at the Conference of European Statisticians / Group of Experts on Statisticians / Group of Experts on National Accounts, Twenty-third session Geneva, 23-25 April 2024.
[16] United Nations (2019), System of Environmental-Economic Accounting — Accounting for Energy, https://seea.un.org/sites/seea.un.org/files/documents/seea-energy_final_web.pdf.
[17] United Nations et al. (2024), System of Environmental-Economic Accounting—Ecosystem Accounting (SEEA EA), https://seea.un.org/ecosystem-accounting.
[1] United Nations et al. (2014), System of Environmental-Economic Accounting 2012 — Central Framework, United Nations, https://seea.un.org/sites/seea.un.org/files/seea_cf_final_en.pdf.
← 1. It should be noted that at the time of finalising the guide, some of the codes associated with the classification hierarchies of the 2025 SNA were still under review and may undergo change. Users are advised to consult the final published version of the SNA when it becomes available.
← 2. In fact, one of the changes in the 2025 SNA is to include also rent and a net return to capital used in production as an additional cost element in the sum of costs method (§4A.6).
← 3. Nominal value and indirect valuation only apply to financial assets.
← 4. The SEEA CF (§5.157) uses in situ to describe the price of the asset before extraction, as opposed to the price received for the sale of the resource after extraction.
← 5. The SNA Update WS.11 Guidance note on the treatment of renewable energy resources as assets (SNA Update, 2023[13]) surveyed the literature but concluded that the evidence was mixed: “Overall, the farmland and residential property price studies we reviewed support the SEEA’s contention that renewable energy production can positively influence land values where the benefits of that production accrue to the owners of the land. However, they also support our contention that even on private land, the size (and even direction) of this change is unpredictable today because markets are not yet in equilibrium. These studies point to the importance of real-world buying and selling decisions, which are influenced by both policy contexts and market participants’ knowledge, both of which continue to evolve within and among countries.”
← 6. In situ resources primarily consist of AN32 mineral and energy resources, and AN3322 non-cultivated biological resources yielding once-only products
← 7. The 2025 SNA follows the terminology of the SEEA CF. Economic rent (as defined in the SEEA CF, §5.113-114) measures “the surplus value accruing to the extractor or user of an environmental asset (e.g. natural resource) calculated after all costs and normal returns have been taken into account”. If we are dealing with environmental assets, the SEEA CF uses the terminology “resource rent”.
← 8. Use of rent payments as proxy for resource rent to value the natural resource is referred to as an appropriation method in the SEEA CF.
← 9. In the split asset approach the appropriation method is de facto used in order to partition the value of an asset, as well as the costs of depletion etc. between legal owner and the extractor (producer) of the resource.
← 10. However, in the split asset approach the appropriation method is de facto used in order to partition the value of an asset, depletion etc. between legal owner and user (or extractor) of the resource.
← 11. The 2025 SNA uses remuneration of employees instead of compensation of employees in the 2008 SNA.
← 12. In the national accounts, taxes on production and imports (D2) = taxes on products (D21) + other taxes on production (D29), while subsidies (D3) = subsidies on products (D31) + other subsidies on production (D39). For details, see 2025 SNA Chapter 8.
← 13. GOS does not include income taxes, property income (like e.g. investment income including interest or similar returns), or rents. Entrepreneurial income (B4g) as defined in the SNA provides a better alignment with the concept of profits in business accounting.
← 14. When the production units of households are not legal entities (and cannot be treated as such), they are described as unincorporated enterprises. They remain part of the same institutional unit as the household to which they belong (2025 SNA, §5.24).
← 15. If the remuneration of employees component in GMI is not negligeable, one may try to use techniques described in 2025 SNA Chapter 17 Capital Services to disaggregate GMI and attribute part as return to labour (remuneration of employees).
← 16. 2025 SNA (§4A.33) emphasises the importance of a direct link: “In the case of non-financial assets, using the method of the present value of future economic benefits can only be used if there is a direct link between the future economic benefits and the asset in question, in the sense that one can assume that there are no other assets which may have generated the residual income.”
← 17. In some cases, one type of natural resource may be used in the exploitation of another. If so, also a return needs to be applied to the supporting natural resource.
← 18. Inventories of products that increase in value may in some cases be recorded as production such as aging cheese and wine (2025 SNA §A7.8). These situations however can be neglected for the purpose of estimating the rate of return.
← 19. Includes depreciation of decommissioning costs, which are in SNA capitalised (see section 4.3.3).
← 20. Guidance and a supporting tool for countries is being developed by the IMF using the simplified PIM set out in Annex C of Measuring Capital (OECD, 2009[3]).
← 21. The situation for non-market producers is different, see Guidelines for estimating the net return to capital for non-market production (ISWGNA, 2025[14]).
← 22. It is also not clear how one could isolate the return to the natural resources from the overall return to the activity.
← 23. If we assume that the return falls in the middle of the accounting period (reflecting the average price level of the period), it is recommended to calculate the return by taking the average of net opening and net closing stocks (both in current prices), with the result that this represents a real rate of return. For example, suppose we have an asset value of 100 at the start of the accounting period and 120 at the end of the period due to net investments of 10 and price increases of 10, we want to use the average asset value (105 when expressed in previous year’s prices) in the price level of the middle of the accounting period (hence adding 5 of inflation) i.e. 110.
← 24. Gross catch is the total live weight of fish caught (gross removal less pre-catch losses) – see Section 5.3.2.
← 25. A disadvantage of smoothing is that a change in market conditions (e.g. prices shift higher or lower for the long term) is picked up with a delay. However, it is difficult to tell whether a change in price is temporary or structural, and in general the advantages of smoothing (less volatility, better predictor of future prices) are considered to outweigh the drawbacks. ISWGNA (2023[15]) mentioned “the importance of using long-term averages of resource rents (applying the residual value method), to avoid volatility in the value of mineral and energy resources as a result of short-run price fluctuations of commodity prices.”
← 26. The period of three to ten years can be seen to coincide with the business cycle. Under certain circumstances there may be good reasons not to smooth, for instance when futures markets provide a different signal compared to the long-term price trend or due to expected changes in the regulatory regime, but this should be the exception not the rule.
← 27. In periods of high inflation, it is especially important to choose a proper price index to express the unit resource rent in constant prices, for example an industry specific deflator.
← 28. This recommendation is consistent with existing guidelines such as provided by the SEEA Energy (United Nations, 2019[16]) (§6.41) which notes “If extraction profiles are available from experts, energy agencies, geologic institutes, etc., those profiles should be used.”
← 29. To avoid confusion, it is essential for any discussion of this topic that all references to discount rates (proposed or in use) should make clear whether they are expressed in real or nominal terms.
← 30. The monetary asset account for timber resources: work-in-progress is slightly different, as it shows growth, removals and natural losses rather than regeneration/depletion. This is discussed in Section 5.2.
← 31. As explained in Section 6.3 the monetary value may be updated following regular revision process of national accounts data that are used for estimating resource rents for instance during benchmark revisions.
← 32. Agreement on Subsidies and Countervailing Measures, Article 1: Definition of a Subsidy. Available at: https://www.wto.org/english/docs_e/legal_e/24-scm_01_e.htm.
← 35. Which was known as FISIM (financial intermediation services indirectly measured) in the 2008 SNA.
← 36. NB: The purchasers’ price excludes any VAT (value added tax) deductible by the purchaser.
← 37. In concept, agricultural land can be subject to depletion for instance when harvesting practices (linked to provisioning services) lead to soil erosion or nutrient depletion (and thereby reduce future yields). However, measurement is more difficult compared to forest land due to the short rotation cycles of agricultural crops and would require information about elements such as the condition of the soil, which would bring us close to taking an ecosystem accounting approach (e.g. measurement of ecosystem condition).
← 38. Although arguably the SNA, SEEA CF and SEEA EA may not be entirely consistent at this point, and the distinction between depletion and degradation will be further discussed during the SEEA CF update process.
← 39. The SEEA EA (United Nations et al., 2024[17]) (§10.29) explains that the definition of depletion “can be seen as sitting within the definition of ecosystem degradation to the extent that the quantity of a stock of a natural resource is considered part of the structure and composition of an ecosystem asset. The term depletion is retained to refer solely to the cost of using up natural resources. This measure will be narrower in scope than ecosystem degradation since it will only relate to the loss of future provisioning services. However, an economy-wide measure of depletion will be broader in scope to the extent that it includes declines in the net present value of the stock of non-renewable resources due to extraction, in particular mineral and energy resources, since these fall outside of the scope of ecosystem assets. In the framing of the SEEA EA, depletion therefore relates to provisioning services, that are the contributions ecosystems provide to benefits that are extracted or harvested from ecosystems.”
← 40. To see this . If we assume for ease of exposition that extraction and resource rent is constant, we obtain which shows that unless r (the discount rate) is 0.
← 41. It should be noted as mentioned before (see footnote 10) that that the treatment of depletion as a cost of production may also have an effect on GDP.
← 42. Thus 2008 SNA asset permits to use natural resources AN222 does not feature anymore in the non-financial non-produced assets (AN2) classification in 2025 SNA. Under 2025 SNA permits to use natural resources AN392 are solely dedicated to permission to use radio spectra.
← 43. As this appears to be inconsistent with the general approach followed in the 2025 SNA, the treatment of radio spectra has been added to the post 2025 SNA/BPM research agenda.
← 44. The treatment of radio spectra differs from other natural resources in 2025 SNA and is the same as in 2008 SNA, therefore no splitting applies.
← 45. Please note that the recording of specific taxes as D45 instead of D29 does not change the resource rent. In case (conform 2008 SNA) they were treated as D29 they would not be part of the GOS but are added back in for deriving resource rent. In case of treatment as D45 they would be included in GOS (and also feature in resource rent). So the GOS would be different in both situations, but the same resource rent would be obtained.
← 46. This is also the recommendation contained in Eurostat (2022) on the delineation between resource taxes and rent.
← 47. Eurostat (2022[18]) clarifies that under certain circumstances dividend payments may be recorded as rent: “When a lease of natural resources extended to a 100% owned public corporation gives rise to large dividend payments compared to the royalties or other surtaxes and seem disproportionate compared to the equity invested, a reclassification of the non-financial flow under rent should be envisaged.” (Eurostat, 2022[18])
← 48. It should be noted here that a drawback of this recommendation is that we may overstate or understate total wealth. For instance, in case of specific net subsidies, this would then raise the value of the natural resource to the extractor and private owner and overstate total national wealth unless these (expected) subsidies are somehow counter booked as a (contingent) liability for the government. The treatment of (contingent) liabilities is an item on the post 2025 SNA/BPM research agenda.
← 49. As noted in Section 3.3 on Projections of resource rent, actual flows of resource rents (for past reference periods) are not to be changed.