The Gross Domestic Product (GDP) is one of the most widely used and analyzed macroeconomic statistics. Understanding some of its less-intuitive measurement rules can be quite elucidating as to what GDP actually measures.
In this article, I reference widely used national accounting standards and guides such as the System of Accounts framework (SNA) (jointly published by the UN, IMF, World Bank, OECD, European Commission) [SNA08, SNA25], the U.S. BEA's NIPA handbook [BEA], and Understanding National Accounts, 2nd edition (authored by Lequiller & Blades and published by the OECD) [L&B], among other sources. A reference section is included at the end of this article.
A refresher
Recall one common definition of the GDP, the "expenditure approach to GDP":
where
C is domestic household consumption,
I is private domestic investment1,
G is government consumption and investment,
X is exports, and M is imports.
GDP measures the value of production occurring in a window of time (e.g., within a year).
GDP measures the production of goods and services; transactions in financial assets are ignored. The act of buying and selling financial assets is viewed merely as an exchange of financial claims/liabilities rather than an act that produces something [BEA]. Thus, the private investment component I excludes the buying and selling of equity and bonds, and is instead concerned with non-financial capital assets such as machinery, and with inventory accumulation.
The adjective "gross" in GDP refers to not deducting depreciation ("gross" as in "gross of depreciation").
The adjective "domestic" in GDP refers to production attributed to residents of a country. Such production may physically occur abroad. Conversely, production occurring inside a country need not necessarily be attributed to its residents.
As the term "resident" appears frequently in this writing, it might be helpful to recall how it is defined:
A resident of a country is a household, private firm, bank, government unit, or other institutional unit whose centre of predominant economic interest lies within that country's economic territory [SNA08, BPM].
Cooking your own dinner does not count towards GDP but paying someone to cook for you does
Household services such as cooking, cleaning, child care and gardening contribute to GDP when someone is paid to provide them. They are, however, excluded from GDP when households produce these services for themselves [L&B, SNA08].
Reasoning: Household production of services for own use is generally kept outside the production boundary, largely because valuing such services is a tricky matter [L&B, SNA08].
Hence the saying, "marrying your housekeeper may decrease GDP".
Freelance work contributing to GDP can cease to contribute to GDP when performed as an employee
A freelancer working from home for a firm that's a resident of a foreign country increases his own country's GDP by increasing exports X (not very surprising).
However, if he had instead performed the same work as an employee of this foreign firm, his work would no longer be treated as domestic production, and thus would not affect GDP [BPM]. (His employee remuneration instead contributes to GNI.)
Ordinary missiles are inventory, but scarier ones that provide deterrence are fixed capital
Most single-use ammunition, missiles, rockets and bombs are treated as inventory. But certain highly destructive ballistic missiles can instead be treated as fixed capital if they provide an ongoing deterrence service [SNA08].
Goods becoming more valuable while in storage usually does not increase GDP, but maturing wine can
Suppose a retailer buys an item for $100 and keeps it in inventory. If its market price later rises to $150 simply because prices have risen, the additional $50 is a holding gain, not production, and therefore does not increase GDP [SNA08].
Some goods are treated differently when storage itself is part of the production process. Wine that becomes more valuable as it matures is a classic example: part of the increase in value during maturation can be treated as output produced while the wine is being stored [SNA08].
A foreign student can live in a country for years and still have their groceries count as exports
An international student's spending on tuition, accommodation, restaurant meals, groceries, and other living expenses in the host country is recorded as exports2 rather than domestic Household Consumption C.
For example, when a foreign student who has lived in Australia for three years goes grocery shopping, the groceries are treated as an Australian export.
International students are generally treated as residents of their home country even after having studied abroad in a host country for several years [BPM].
Medical patients receiving treatment abroad are treated similarly [BPM].
A homeowner is treated as paying rent when staying in their own house
For the purposes of GDP calculation, a house produces a stream of housing services that its occupants consume.
If rent is not actually paid, an imputed rental value is still added to Household Consumption C. This imputed value is estimated by reasoning about rents for comparable dwellings [BEA, L&B]. This treatment also applies to homeowners occupying their own houses.
Fun fact: This is the largest source of imputation in US GDP [BEA].
Reasoning: This makes GDP invariant to whether a dwelling is rented or owner-occupied.
Additional note3.
Goods can cross an economic territory without being counted in imports or exports
Suppose a semiconductor company resident in Korea sends $1M worth of wafers it owns to a company resident in Singapore for processing. The Singapore company performs $100,000 worth of processing and sends the wafers back to Korea.
If ownership of the wafers never passes to the Singapore company, Singapore is not treated as importing $1M of wafers and later exporting them again. Instead, Singapore records only the $100,000 processing fee as a service export [BPM].
Thus, goods can physically cross a country's border without the goods themselves ever being counted in imports M or exports X.
Reason: International trade in goods in the national accounts generally requires a change in economic ownership, rather than the mere physical movement of goods across borders [BPM].
Goods can be treated as a country's exports without ever entering the country
Suppose a merchant resident in Singapore buys $1M worth of jewelry from a company resident in Japan, then sells it to a customer in the US for $1.2M. The jewelry is shipped directly from Japan to the US and never enters Singapore.
In Singapore's national accounts, the merchant's $1M purchase is recorded as a negative export and the $1.2M sale as a positive export, resulting in a net effect of increasing Singapore's exports by $200,000 [BPM].
Reason: International trade in goods in the national accounts generally follows changes in economic ownership, rather than the mere physical movement of goods across borders [BPM].
Illegal economic activity can contribute to GDP
The illegality of an economic activity is not sufficient reason to exclude it from GDP.
Under the SNA/ESA framework, consensual market transactions involving illegal goods and services can still count as production.
For example, EU countries include estimates of certain illegal activities such as illegal drug sales, prostitution and smuggling in GDP [SNA08, ESA10].
Reasoning: GDP measures economic production, legal or otherwise. Excluding illegal market activity complicates comparisons when an activity is legal in one country but illegal in another.
It should be noted that certain countries like the US does not attempt to comprehensively measure illegal economic activity in official GDP numbers.
An insurance policy's contribution to the GDP isn't simply its charged premium
Suppose you pay an insurer a premium of $1,000 for a policy. If the insurer expects, on average, to pay about $700 of that amount back to policyholders as claims, as well as earn $50 of investment income on reserves held on behalf of policyholders, then the measured insurance output would be:
$1,000 - $700 + $50 = $350.
Thus, it is $350, not the full $1,000 premium, that is treated as the value of the insurance service produced.
Actual claims are not directly used to subtract from premiums. If actual claims were deducted directly, a year with an unusually large natural disaster could collapse measured insurance output or even make it negative, even though insurers may actually have provided more insurance services than usual. National accountants therefore use an adjusted measure of claims (represented by the $700 in this example) that smooths exceptional losses when estimating insurance output [BEA, SNA08].
Scientific research contributes to GDP, unlike many other business expenses
Companies often record R&D as an expense. In national accounting, however, (qualifying) R&D is treated as capital formation rather than intermediate consumption and therefore contributes to GDP [SNA08, BEA].
For a private business, R&D increases Domestic Investment I. For the government, R&D instead increases government expenditure G.
The scientific discovery of novel knowledge is thus treated similarly to the construction of productive machinery.
Conclusion
Whether an act of production contributes to GDP, and to which country's GDP or which component (C, I, G, X, M), depends on concepts such as the production boundary, residence, economic ownership, imputation and capital formation.
References
[BEA] U.S. Bureau of Economic Analysis; 2024; NIPA Handbook: Concepts and Methods of the U.S. National Income and Product Accounts; (URL not provided because the US BEA does not version-pin the NIPA Handbook)
[L&B] Lequiller, and Blades; 2014; Understanding National Accounts, 2nd ed.; OECD Publishing, Paris; https://www.oecd.org/content/dam/oecd/en/publications/reports/2014/10/understanding-national-accounts_g1g43f55/9789264214637-en.pdf
[SNA08] European Commission, International Monetary Fund, Organisation for Economic Co-operation and Development, United Nations, and World Bank; 2008; System of National Accounts 2008; https://unstats.un.org/unsd/nationalaccount/docs/SNA2008.pdf
[SNA25] European Commission, International Monetary Fund, Organisation for Economic Co-operation and Development, United Nations, and World Bank; 2025; System of National Accounts 2025; https://unstats.un.org/unsd/nationalaccount/sna2025.asp
[ESA10] Eurostat; 2013; European System of Accounts — ESA 2010; https://ec.europa.eu/eurostat/documents/3859598/5925693/KS-02-13-269-EN.PDF.pdf/44cd9d01-bc64-40e5-bd40-d17df0c69334
[BPM] International Monetary Fund; 2014; Balance of Payments and International Investment Position Compilation Guide; https://www.imf.org/external/pubs/ft/bop/2014/pdf/GuideFinal.pdf
[IMF-QNA] International Monetary Fund; 2017; Quarterly National Accounts Manual: 2017 Edition; https://www.imf.org/external/pubs/ft/qna/pdf/2017/QNAManual2017text.pdf
[UN-SUT] United Nations Statistics Division; 2018; Handbook on Supply, Use and Input-Output Tables with Extensions and Applications; https://unstats.un.org/unsd/nationalaccount/docs/SUT_IOT_HB_Final_Cover.pdf
[OECD-CAP] OECD; 2009; Measuring Capital — OECD Manual 2009, 2nd ed.; https://www.oecd.org/en/publications/measuring-capital-oecd-manual-2009_9789264068476-en.html
[ONS] Office for National Statistics; 2025; UK National Accounts, The Blue Book: 2025; https://www.ons.gov.uk/economy/grossdomesticproductgdp/compendium/unitedkingdomnationalaccountsthebluebook/2025
[DESTATIS] Federal Statistical Office of Germany (Destatis); 2025; ESA 2010 Methods and Sources for the German GNI (Gross National Income) and Its Components — Edition 2025; https://www.destatis.de/EN/Themes/Economy/National-Accounts-Domestic-Product/Publications/Downloads-National-Accounts-Domestic-Product/esa-2010-methods.pdf
[SINGSTAT] Singapore Department of Statistics; 2019; Benchmarking of Singapore's National Accounts to Reference Year 2015; https://www.singstat.gov.sg/-/media/files/publications/economy/ip-e46.pdf