What will the new approaches to macroeconomic statistics bring?

main conclusions

• In 2025, most countries will begin transitioning their statistical agencies to new standards for calculating macroeconomic indicators. The development of economic theory and economic relations requires periodic refinement of statistical approaches because statistics lag behind the reality they aim to measure. The previous System of National Accounts (SNA) was released in 2008. It took most countries four to seven years to implement its provisions in national methodology.

• This time, the most important changes to the SNA involve accounting for the depletion of natural resources and recording the production of data as investments in fixed capital. In addition, the approach to reflecting the activity of central banks will be changed, renewable natural resources will start to be interpreted as assets, and the method for assessing the production volumes of some non-market goods and services will also change.

• Although the new method for accounting natural resource depletion will not directly impact the size of gross domestic product (GDP), it will lead to net domestic product (NDP) falling by up to 20% of GDP for countries with high mining volumes. The logic of NDP is to reflect the full ‘price’ that an economy pays to maintain current production volumes. With the implementation of SNA 2025, NDP will become more informative. For Russia, the difference between GDP and NDP will increase by approximately 6 pps, to 18–21%.

• The assessment of the value of data ‘produced’ not for sale and the inclusion of this value in GDP is a necessary but complex novation, which will result in GDP being revised upward by a value of up to 3%, depending on country. Data collected, created and processed by companies or governments can work for the economy not only at the moment of creation, but also for a long time after that. This makes it similar to investments in fixed capital. The potential impact of the introduction of changes in the methodology on the modern Russian GDP is its revision by approximately 2% upward with the prospect of growing by multiple times over the horizon of a decade.

• The main practical conclusions for users of statistics are that (1) in the transition period, a lot more attention must be paid to cross-country comparisons, as countries implement new approaches at different speeds, (2) new opportunities for analyzing the digitalization and stability of economic growth will arise in macrostatistics.

FUTURE CHANGES TO THE SYSTEM OF NATIONAL ACCOUNTS

National wealth is the value of all assets owned by a country’s residents (minus liabilities).

In March 2025, the United Nations Statistical Commission approved the new version of the System of National Accounts¹. The previous version was adopted 17 years ago, in 2008. National statistics agencies use the system to develop their own methodologies for measuring core macroeconomic indicators: GDP, GNP, GRP, national wealth, investments in fixed capital, savings, etc. The development of economic theory and economic relations requires periodic refinement of statistical approaches because statistics lag behind the reality they aim to measure.

When implemented, SNA 2025 should improve the measurement of activities, the product of which is digital or based on data (the share of this is growing in the world), as well as enhance the analysis of the sustainability of production and economic growth rates.

¹ Key documents on developing and implementing the SNA: https://unstats.un.org/unsd/nationalaccount/default.asp.
Documents on developing new approaches as part of SNA 2025: https://unstats.un.org/unsd/nationalaccount/SNAUpdate/GuidanceNotes.asp.

A complete list of conceptual changes is given in Annex 4 of the System of National Accounts 2025. In parallel with the implementation of the new SNA, the implementation of a new version of the Balance of Payments Manual (BPM 7) will begin.

Here are the key conceptual changes.

1.  Depletion of natural resources will be perceived as a cost of production. This will not affect GDP, but will result in lower NDP. An assessment of how these changes impact macroeconomic indicators is given on pages 4–5.

2.  Digital data collected within a company or government agency will be considered a produced asset. Data production will be valued as an investment in fixed capital, which will increase GDP (see pages 6–7 for more details).

3.  The use of the method of assessing the output of industries by costs will become more universal. Capital costs will be included in the assessment of the cost of non-market production, as well as labor costs (currently only labor is taken into account). This, in particular, may increase the assessment of the added value created in the process of providing public services, and accordingly, GDP (by up to 3% depending on country).

4.  The activity of central banks will be considered entirely non-market. All regulation-related payments made by financial companies and banks will become transfers rather than their costs, which will increase the estimated value added produced by the sector and probably GDP (by up to 0.2%).

The use of the new system and the creation of new methodologies is voluntary, but the vast majority of countries are participating in this process. On average, countries needed four to seven years to implement the previous versions of the system (SNA 1947, SNA 1953, SNA 1968, SNA 1993, SNA 2008) after the publication of these documents. Given this experience and in accordance with the implementation strategy, most countries should switch to SNA 2025 in 2029–2030.

In Russia, the transition to SNA 2008 generally took place in 2014, but some recommendations are still being implemented.

The revision of GDP as a result of the implementation of the 2008 SNA in OECD countries reached 5 pps (on average it was closer to 3 pps)². ACRA estimates that implementation of SNA 2025 will have a similar effect.

Throughout the transition period, the already limited comparability of macroeconomic indicators across countries will become even worse. When making cross-country comparisons, data users will need to ensure, at a minimum, that the indicators for different countries are calculated using the same version of the SNA.

Comparability is limited, including due to the different rates and areas of data revision. For further details, see ACRA’s research A significant positive revision of industry data is the norm for Russia and many other countries from October 20, 2020.

² See the paper The impact of implementing the 2008 SNA on GDP, Sim Benson (UNSD) at the Regional Workshop on National Accounts and the development of Economic Statistics Infrastructure within the SDGs Framework.

NEW ACCOUNTING FOR NATURAL RESOURCE DEPLETION

What has changed? Depletion of non-renewable resources (such as oil, gas, coal) will now be reflected as production costs³ in the same way as depreciation of fixed capital. Depletion of renewable resources (for example, forests) will be recorded in a similar way if their use exceeds the rate of reproduction. Consequently, when calculating NDP, the ‘cost’ of using resources will need to additionally be subtracted from GDP.

SNA 2008: NDP = GDP — depreciation of fixed capital.
SNA 2025: NDP = GDP — depreciation of fixed capital — depletion of natural resources.

³ In the 2008 SNA, resource depletion is treated as ‘other changes in the volume of assets’, which implicitly reflects the essence of what is happening.

What is the logic behind this? The change described above is designed to adjust the fact that in the SNA, the extraction of non-renewable resources increased GDP and other indicators, but did not clearly enough record the associated losses of national wealth. For example, the confident growth of Mongolia’s economy in 2011–2018 at an average rate of 6.5% annually was largely due to booming copper and coal exports, mainly to China. The economy received revenues at the expense of growing depletion of natural resources, but revenues were not sufficiently saved or invested. Most of the SNA indicators show that growth was impressive, but they mask the fact that overall national wealth declined, by 3–12% over 2011–2018 according to different definitions. There are many historical examples of such periods of resource-based economic growth that did not lead to long-term growth of national wealth.

Figure 1. What could be the reduction in net domestic product look like when calculated in accordance with the new approaches?*

* The graph shows average annual depletion of natural resources for 2012–2021 for 42 countries; estimates for all countries are provided in Appendix 1.
** Average indicators are given for groups of countries: arithmetic mean/weighted average by size of economy.
Sources: World Bank, ACRA

According to the original concept, NDP should show how much of the product produced in the economy can be ‘eaten up’ without losing future production capabilities, but also without increasing them. This is why, when calculating NDP, the depreciation of fixed capital that has occurred is subtracted from GDP, since in order to prevent losses of production capacity, part of the gross product must go towards restoring the means of production (machines, tools, etc.). At the same time, in reality, even if this depreciation is offset, national wealth and future opportunities may decline because in the process of production in the economy there was not only wear and tear of production assets, but also depletion of some natural assets. In order to avoid a decline in national wealth, income from the extraction and sale of natural assets must be transformed by the economy into other assets, either financial or physical. The new formula for NDP captures precisely this need.

In the new SNA, the use of natural capital is viewed on par with the use of physical capital and therefore the essence of NDP is changing. Now it shows how much more of the gross product could be ‘eaten up’ without losing production capacity and resource wealth. This is a step toward more transparent measurement of national welfare, brining macrostatistics closer to sustainable development metrics⁴. NDP may become a new basis for formulating budget rules for natural resource extracting countries as a kind of analogue of ‘non-oil GDP’.

⁴ Some studies consider the possibility of including other types of capital, such as human or environmental capital, in macroeconomic statistics. Experimental measures take into account the fact that education expenditures is an investment that increases human capital, national wealth, and productive capacity, while, for example, particulate matter emissions cause a decrease in natural wealth along with the depletion of natural resources. See, for example, the World Bank’s paper The Changing Wealth of Nations 2024: Revisiting the Measurement of Comprehensive Wealth.

Consequences for statistical indicators

1. National GDP will not change, but the new methodology will affect the structure of the production and distribution accounts of the SNA.

2. NDP will decline in most countries to the extent that a country’s production involves mining and deforestation (Fig. 1), and the difference between NDP and GDP will increase. For Russia, in 2021 NDP would have amounted to 79% of GDP as per SNA 2025, while under SNA 2008 it amounted to 88% of GDP (Fig. 2). Natural resource depletion in Russia averaged 6% of GDP annually from 2012 to 2021.

3. The information content of NDP and its growth will increase. Currently, the growth rates of NDP in most cases are very similar to the growth rates of GDP, but this will change. The difference in growth rates will carry information about what part of the GDP growth is ‘paid for’ by losses of natural resources.

4. Net national income and net savings will decline.

Figure 2. Approximate calculation of Russia’s NDP for 2021 (before and after changing approaches)*

* The calculation is based not on the latest available GDP estimate for 2021, but on the estimate from the collection Indicators of National Accounts of Russia in 2016–2023 dated August 30, 2024, since this is the most current estimate, which has a comparable official estimate of NDP (shown in the second line of the graph). The calculations are illustrative in nature and are intended to show the approximate ratio of NDP calculated using different methodologies.
Sources: Rosstat, World Bank, ACRA

Difficulties of implementing and using the new approaches

1. The valuation of extracted resources can vary within rather wide limits depending on the assumptions of the calculation, since in the 2025 SNA it is based on calculating the difference between future discounted rental incomes at the beginning and end of the period, and not on the current market price of the resource. A relatively complex approach to estimating resource depletion will reduce its volatility over time, but this is achieved at the cost of accepting assumptions about unobservable parameters.

2. The interpretation of physical growth of NDP as a sustainable component of GDP growth is tempting and partly justified, but its new definition does not reflect the fact that the depletion of natural resources in extractive countries indirectly reduces future production capabilities not only in these countries but also in countries that mainly import these resources. The analysis of the sustainability of growth should not be reduced to the analysis of NDP.

DATA PRODUCTION AS AN INVESTMENT

Essence of the changes. Digital data collected and processed for internal use by companies and the government (for example, a database of clients or taxpayers) will now be considered produced assets, similar to software. This means that the costs of producing and processing such data will be included in the gross fixed capital formation.

Since the cost of the data collected and processed for internal use is unknown (there is no market where it could be determined), it will be calculated based on the corresponding labor costs. To determine the added value created by this labor, costs will be multiplied by the typical ratio of output in industries where the data is a final and marketable product to the wage fund in these industries (in fact, it is multiplication by the estimated labor productivity in a different industry)⁵.

This approach to estimating the output and added value (cost-based) applies to most types of activities resulting in non-market products, for example, public services.

⁵ For more information about the approaches the statistics agencies may apply to estimate the value of produced data, see Handbook on measuring data in the System of National Accounts.

Logic of the changes. The purpose is to reflect the growing value of data outside the digital economy. In many sectors, even traditional ones, including agriculture and industry, data about a product, technology, customers, competitors, or suppliers is becoming an important production factor and a competitive advantage. Previously, the cost of creating data for internal use was accounted for as an ongoing expense (and such data itself was not considered an asset), which resulted in underestimation of the amount of investment in intellectual capital. In reality, the data collected, created and processed by companies or the government can work for the economy at the time of creation, as well as for a long time afterwards. This is what they have in common with investments in fixed capital. Changing the approaches provides a wider picture of the scale of digitalization in the economy and, in particular, the share of digital assets.

For more information about the growing importance of the digital economy and IT sector, see ACRA’s research Structural changes in the Russian economy in 2022–2024 from March 14, 2025.

Impact on the statistical indicators

1. Gross fixed capital formation will increase and, as a result, GDP will grow. The first estimates of the value of annually produced data in countries where statistical agencies have made experimental calculations are in the range of 0.5–3.0% of GDP (Table 1). ACRA does not exclude the possibility of this figure increasing several times over time for certain countries after the full-fledged introduction of the methods. Similar calculations conducted by the Agency with a large number of expert assumptions show that the range of estimates in Russia is 1.0–3.0% of GDP in 2016–2021 (Appendix 2).

2. The estimated national wealth will increase because digital data will be considered assets. The first estimates of their gross value in different countries amounted to up to 13% of GDP.

3. The share of industries that use their own data more actively (compared to averages) to provide services or produce goods will grow in the total value added and GDP. Presumably, this will primarily affect the financial sector, public administration, and retail trade.

Table 1. Experimental estimated value of data produced for internal use

Difficulties of implementation and application of the new approaches

1. Rather complex studies of the very essence of labor, rather than its formal characteristics, will be required to evaluate the labor costs associated with data creation. Currently, very few of these labor market indicators are collected. Statistical agencies will not only have to assess which professions in each industry involve data collection and processing, but also, what is much more difficult — to determine how much of the working time in each profession is spent on data-related tasks, that is, to examine the structure of individual worker’s working hours.

2. In economic terms, fixed capital should only include data that can be used for more than one year. However, given the technical complexity of dividing data by usage life, SNA 2025 recommends that all data should be treated equally. This may lead to an overestimation of investments in fixed assets and, as a result, of the estimated value of accumulated data.

3. The valuation of accumulated data, which will become a part of the valuation of national wealth, contains another component that is difficult to measure — the estimated rate of depreciation. Relevant practice, including accounting, has been built in relation to physical objects, however, for intangible assets of a non-market nature, it will be necessary to accept parameters such as the service life of data, which will be very difficult to justify in practice, although they will greatly affect the result, changing the asset value multi-fold.

* Depletion of resources means the annual average decline in the value of resources due to their extraction. The value is estimated as a discounted sum of future rental incomes.
Sources: World Bank, IMF, ACRA