What a technical argument among statisticians from 2003 to 2015 reveals about who gets to count, what gets capitalized, and why Decolonial Accounting is a method towards peace.
In Aesop’s fable of the Fox and the Stork, the fox invites the stork to dinner and serves soup in a flat shallow dish — a vessel perfectly suited to a fox’s tongue, while inconvenient for a stork’s long bill. The stork goes hungry. When the stork returns the invitation, the soup arrives in a tall narrow jar that only a beak can reach. The fox goes hungry in turn. Each host has set the terms of the meal, and those terms happen to suit the host.
There is a moment buried in the official record of the System of National Accounts revision process that has this quality. In February 2004, the Advisory Expert Group (AEG) convened to vote on whether warships, tanks, fighter aircraft, and missile systems should be reclassified as fixed assets in national accounts. The proposal passed 15 to 4, with 2 abstentions (Advisory Expert Group, “Decision of the AEG” 1). The United States Bureau of Economic Analysis not only supported the change — it had already implemented it unilaterally, years earlier. What the rest of the world was being asked to adopt, the US government had been quietly practicing since before the debate formally began. The dish had been set out; the question was only whether the other guests would find it nourishing.
The Technical Debate That Was Never Only Technical
The background is worth establishing precisely. The 1993 SNA treated most military hardware — tanks, missiles, warships, aircraft, guns. munitions, etc — as intermediate consumption by the government. This meant that when a government purchased an F-16, the full cost hit the books as spending, not capital formation. The rationale traced back to the post-World War II origins of national accounting: weapons were made to destroy, and the pioneer architects of the system — who had lived through the destruction — treated military operations as external to the economic process. As French national accountant André Vanoli wrote in his February 2004 letter to the ISWGNA, “military operations in war time and by extension military services in general were not considered to be an activity similar to economic activities… What happened internally in the sphere of military activities was judged external to the domain of economic activities” (Vanoli, “Military Expenditures” 1).
Imagine if you’re a shopkeeper on Main Street. When you buy paper bags and cleaning supplies to run the shop day to day, those are expenses — they get used up, they’re gone, they don’t show up on your balance sheet as something you necessarily owns. That’s intermediate consumption: spending that disappears into the cost of doing business. When you purchase a commercial oven that will bake goods for the next fifteen years, that’s a fixed asset — it goes on your balance sheet and depreciates it a little each year, and it counts as part of the shop’s productive wealth. Capital formation is simply the act of adding to that stock of productive assets, which makes your shop look wealthier, more creditworthy, and more economically substantial on paper. Now here’s where the military question bites: under the old SNA rules, when the government produces a warship or guns or bullets, they were treated like the paper bags — spent and gone, intermediate consumption, no lasting asset. Under the SNA 2008 rules, weapons become treated like the commercial oven — a fixed asset, depreciating over decades, adding to measured GDP and government net worth every year it sits in port. The material consequence is that countries with large militaries suddenly have bigger measured economies and stronger-looking balance sheets, not because anything new was produced for citizens, but because the accountants changed which shelf they put the warship on. For countries with small militaries that imports weapons on credit, the same reclassification makes the country look like it has more capital, but the debt to pay for that warship, and the biodiversity that the military and weapons system is degrading don’t even have a shelf at all.
The Canberra II Group, led by Brent Moulton of the U.S. Bureau of Economic Analysis, argued that this treatment was conceptually inconsistent and practically problematic. Their case had three main pillars. First, military weapons provide services continuously — including deterrence — and should therefore qualify as fixed assets under the same criteria applied to any other produced asset used repeatedly in production for more than one year. Second, the existing distinction between “destructive” and “dual-use” equipment was difficult to operationalize and created anomalies: an armored vehicle purchased by police was a fixed asset; an identical vehicle purchased by the army was intermediate consumption (Moulton, “Canberra II Group’s Recommendations” [revised 2004] 2). Third, the new International Public Sector Accounting Standards (IPSAS 17) already classified specialist military equipment as property, plant, and equipment, and harmonization between public financial accounts and national accounts was a stated priority.
At the April 2003 meeting in Voorburg and the October 2003 meeting in Paris, the Canberra II Group reached near-unanimous agreement on the reclassification. By February 2004, the AEG voted to accept the recommendation. By 2009, the UN Statistical Commission endorsed the SNA 2008, which incorporated military weapons systems into the asset boundary — formally. By 2014, most OECD countries had implemented the new standard (van de Ven, “New Standards for Compiling National Accounts” 2).
What was excluded from the revision shelf was the one thing that could help to balance out military systems: enviromental accounts. The 1993 revision had already been the moment when ecological accounting came closest to the core. Joint UNEP and World Bank workshops through the late 1980s had built enough momentum that environmental and natural resource accounting was an explicit item in the SNA revision process. The resulting 1993 SEEA — the *Integrated Environmental and Economic Accounting* — acknowledged directly that its own architects had tried and been turned back: “considering the current state of knowledge on environmental accounting and the divergent views on a number of conceptual and practical issues, it has not been possible to reach an international consensus at this time for a fundamental change in the SNA” (United Nations Statistics Division, *Integrated Environmental and Economic Accounting* iii). Ecological accounting was assigned to satellite status — a complement to, not a component of, the central framework. The 1993 SNA itself repeated this architecture, devoting a separate section to environmental satellite accounts while keeping the production boundary intact (United Nations, *System of National Accounts 1993*, chap. XXI). The stork’s jar had been acknowledged, noted, studied, and set aside. By the time the SNA 2008 revision cycle opened a decade later — the same cycle that capitalized tanks and missile systems — the household accounts and wellbeing indicators that had also been raised as serious proposals for the core framework were again routed to supplementary tables and further research agendas. Depletion of natural assets was recorded in the “other changes in volume” account, with no effect on GDP; depreciation of a warship now was. The dish had not changed. The guests who could only use a long-billed jar were told, once more, that the table had been set as fairly as the current state of knowledge allowed.
The Dissenters Identified the Problem Correctly
What makes this archive instructive is not the outcome but the dissent, which was sharper and more structurally aware than the revision literature typically acknowledges.
Vanoli’s intervention was the most systematic. He argued that weapons are not economic assets — they are political assets. Their function, properly understood, is destruction in wartime, and the pre-1993 tradition recognized this by treating military outlays as current expenditure at the moment they arose. The deterrence framing deployed by the Canberra II Group, he argued, obscures rather than clarifies: deterrence is an outcome, not a service, and the analogy with insurance is logically flawed because “defence expenditures in the prevention approach are incurred in order to avoid the risk itself, not to cover the possible losses if prevention fails” (Vanoli, “Military Expenditures” 3). When war actually breaks out, he noted with precision, prevention fails., and that will undermine peacetime accounting models by collapsing them into contradiction. A tank destroyed in combat, under the proposed treatment, becomes an “other change in volume” — an externality — rather than the integral use of the asset for its designed purpose. This, Vanoli concluded, was “a kind of conjuring trick in order not to bear all the consequences of the proposed treatment” (Vanoli, “Military Expenditures” 4).
The Federal Statistical Office of Germany raised a different but equally significant objection: the new treatment would substantially affect both GDP and government net lending/borrowing, and in the context of the European Union’s excessive deficit procedures, this was politically unacceptable as a technical accounting revision (Federal Statistical Office of Germany, “Updating of the SNA-93” 1). Germany’s objection was institutionally motivated, pointing to an unfair national accounting balance sheet that favored high-defense-spending economies.
Fadhil Mahdi of ESCWA, representing developing country perspectives in the expert comment record, was direct: “this proposal did not meet unanimous acceptance from AEG members… military expenditure does not generate tangible production. In addition, these expenditures increase drastically the value of capital formation in developing countries. While the value added generated by the total investment becomes very small especially in the non-producing countries importing the weapons” (Mahdi, in “Expert Comments for Issue: Military Expenditures” 1). This is the structural asymmetry in plain language. A weapons-importing country gets a GDP uplift and a capital formation entry for hardware it did not produce, financed by debt, with no domestic value-added chain behind it.
The Banco de Portugal, in its country comment, accepted the reclassification while noting a different problem: the borderline between fixed assets and inventories was conceptually unstable. Bombs and missiles stored for deterrence, Portugal’s central bank observed, are themselves deterrents — which means “the argument for including military weapons systems in gross fixed capital formation can also be used in relation to expendable durable goods” (Banco de Portugal, “Military Expenditures Comment” 2). This is not a minor technical quibble. It exposes the arbitrary line the Canberra II Group drew between capital and inventory, a line that was ultimately more politically expedient than logically necessary.
Banco de Portugal’s criticism is similar to the fox drawing a line down the middle of the table and calling one side provisions and the other side supplies — insisting the distinction is technical and principled, when its actual function is to determine who eats and who doesn’t. A ballistic missile held in a silo for twenty years as a strategic deterrent meets every criterion the Canberra II Group established for a fixed asset: it is used continuously in the production of defense services, a measurement of national security, it provides ongoing deterrence, and it has a plannable service life. A bomb held in a warehouse for five years waiting to be expended in training meets most of the same criteria. The Canberra II Group resolved this by arguing that deterrence counts, while expendability doesn’t. There is no logic that values lives or environments, it was simply a warmongering beancounter’s convention aligning with the largest defense-spending economies, managing their accounting.
What the GDP Numbers Actually Show
From a macroeconomic perspective, the OECD’s 2015 Statistics Brief shows how the 2008 changes to the SNA altered GDP across the advanced economies, but it does not account for the ongoing downturn or systemic debt that followed the 2008 financial crisis. Across OECD countries, the capitalization of military weapons systems under SNA 2008 began in 2013 and raised GDP levels by an average of 0.3 percentage points (van de Ven, “New Standards” 9). The largest impacts were in Greece (0.6 percentage points) and the United States (0.5 to 0.6 percentage points), the latter noted with the caveat that “military expenditures had already been capitalised in NIPAs” before the international standard caught up (van de Ven, “New Standards,” Table 1, 11). In other words, the US did not change its practice to match the international standard — the international standard changed to match US practice (National Income and Product Accounts is the system that the U.S. Bureau of Economic Analysis uses, and often—not always— influences the SNA). The stork had already eaten. The question put to the multilateral table was only whether everyone else would accept the jar as the correct vessel.
It should be no surprise, but what the accounting of military systems reveals is that there is an historical pattern of how international statistical standards get made: leading economies, particularly those with large state apparatuses and well-resourced statistical offices, shape norms that smaller, less-resourced, or structurally different economies then absorb as technical requirements. The SNA military asset revision is one instance of a recurring dynamic in which accounting standards, presented as methodological improvements, carry embedded assumptions about what constitutes productive activity, what deserves to be counted as wealth, and whose national balance sheet should look stronger.
The worldwide military expenditure data in the archive gives this additional weight. In 2004, total world military spending was estimated at approximately $1.1 trillion, with the United States alone accounting for $623 billion by FY2008 — more than all other countries combined (World Wide Military Expenditures). The countries at the bottom of that list — Fiji at $36 million, Papua New Guinea at $17 million, East Timor at $4.4 million — are either weapons importers or near-demilitarized Pacific and small island states. Under the SNA 2008 framework, a small island economy that finances weapons imports now records those purchases as capital formation, improving its measured savings and capital stock even as the underlying productive capacity remains unchanged and the debt service accumulates.
This matters directly for Pacific contexts. When Fiji, Samoa, or Tonga record imported military hardware as fixed assets, GDP goes up marginally, measured net saving improves, and government balance sheets appear stronger. What does not appear is the opportunity cost: the healthcare infrastructure not built, the watershed management not funded, the fisheries data system not capitalized, environmental data left on the side of a road. The SNA 2008 expansion of the asset boundary for military systems occurred while ecological assets — forests, fisheries, freshwater systems — remained outside the production boundary entirely, consigned to satellite accounting or are simply unmeasured, waiting for new international standards to absorb that data into their own bureaucratic systems.
The Intemerate Accounting Critique
From the perspective of Intemerate Accounting, the SNA military asset debate illustrates precisely the problem that the Intemerate framework is designed to address. The SNA’s production boundary is does not represent the production strengths of a united nations — it reflects a historically specific set of priorities about what kinds of activity counts as economically productive, what kinds of capital deserve to be depreciated and therefore counted as contributing to output, and what kinds of damage remain invisible in the core accounts.
The military asset reclassification expanded the asset boundary in the direction of state-organized violence while the natural capital boundary remained contested, fragmented, and largely satellite-bound. The UN Statistical Commission endorsed weapon systems as fixed assets in 2009, justifying it against a fictional asset value of national security; it has not endorsed an equivalent mandatory treatment for environmental security: forest depletion, fisheries drawdown, or soil carbon loss. The fox served the meal, and the dish was flat. This asymmetry is neither oversight nor accidental, it was built as a system to reflect which constituencies have the institutional weight to shape international statistical standards, and which do not.
Vanoli was right that military equipment and weapons are political assets, not economic ones — but he stopped short of the deeper conclusion. The same critique applies to the entire production boundary. GDP is a political construction, not a neutral measure of economic reality. Its expansion to include weapons capital while continuing to exclude ecological reproduction costs is a statement about whose claims on the future are recognized as productive and whose are not.
The Intemerate Accounting framework addresses this through its multi-layer compliance chain and its insistence that ecological, social, and community reproduction costs be measured, attributed, and accounted for at the local level, with data sovereignty vested in the communities whose resources are being drawn upon. Where the SNA 2008 asks “what criteria must an asset meet to enter the production boundary?” and answers with technical criteria about repeated use and service lives, Intemerate Accounting asks the prior question: who decides what counts as an asset, and in whose interest is that boundary drawn?
The practical implication is, for example, a community in Fiji whose mataqali land sits adjacent to military infrastructure — or whose fisheries are affected by naval activity — is not a participant in the SNA’s asset accounting. The land exists; the fisheries exist. But under SNA 2008, what is recognized as capital is the military hardware, not necessarily the reef system it may degrade, unless of course the reefs have enforceable ownership and generate economic benefits. The Intemerate framework proposes to invert this: ecological and community assets are primary; military expenditures are a liability that should be disclosed, not a productive input to be capitalized.
And just to be clear, when we talk about environmental or ecological assets, that ownership applies to the physical boundaries as much as the ownership and management of its data.
The Decolonial Stakes of Statistical Standards
The 2003-2015 SNA military assets debate is an accounting dispute about GDP measurement, and it is also something else. It is a record of how dominant states use technical multilateral institutions to normalize the priorities of their own political economies as universal methodological standards.
In summary, the technical structure of hegemony remains: the Advisory Expert Group voted; the Canberra II Group recommended; the UN Statistical Commission endorsed; but the United States had already implemented it. In this scenario, the fox had already eaten, and the vote was about whether to call the dish a universal standard.
The question that Intemerate Accounting raises — and that the brand new World Data Organization, inaugurated in Beijing on March 30, 2026, now creates space to ask in a different institutional context — is whether alternative measurement frameworks can achieve the same kind of institutional recognition, not through unilateral implementation by powerful states, but through multilateral deliberation that includes the communities and South-South economies whose ecologies and social reproduction costs are currently invisible in the core accounts.
Works Cited
Advisory Expert Group on National Accounts. “Decision of the AEG on the Classification of Military Weapon Systems as Fixed Assets.” United Nations Statistics Division, February 2004.
Advisory Expert Group on National Accounts. “AEG Meeting February 2004: Recommendation of Consultation.” United Nations Statistics Division, 2004.
Banco de Portugal. “Comment on Military Expenditures.” Country comment submitted to ISWGNA/AEG, 2004.
Federal Statistical Office of Germany. “Updating of the SNA-93, Issue 19: Military Expenditures.” Wiesbaden, 16 July 2004.
Mahdi, Fadhil. Expert comment on military expenditures issue. In Expert Comments for Issue: Military Expenditures. United Nations Statistics Division, 2004.
Moulton, Brent R. “Canberra II Group’s Recommendations to Treat Military Weapon Systems as Fixed Assets.” SNA/M1.04/09, United Nations Statistics Division, 17 December 2003 (revised 10 March 2004).
Moulton, Brent R. “Canberra II Group’s Recommendations to Treat Military Weapon Systems as Fixed Assets.” Canberra II Group on Non-Financial Assets, 17 December 2003.
Paparella, Christof, and Viet Vu. “Military Weapon Systems as Fixed Assets.” Workshop on 1993 System of National Accounts Update, Bangkok, 19-22 April 2005. United Nations Statistics Division.
United States Bureau of Economic Analysis [Landefeld, J. Steven]. Letter to Dr. Paul Cheung, Director, Statistics Division, Department of Economic and Social Affairs, United Nations. 6 October 2006.
van de Ven, Peter. “New Standards for Compiling National Accounts: What’s the Impact on GDP and Other Macro-Economic Indicators?” OECD Statistics Brief 20 (February 2015).
Vanoli, André. Letter to members of the ISWGNA on military expenditures. N° 01-2004-AV, 4 February 2004.
United Nations Statistics Division. *Integrated Environmental and Economic Accounting: Interim Version*. Studies in Methods, Series F, No. 61. New York: United Nations, 1993.
United Nations. *System of National Accounts 1993*. Brussels/Luxembourg, New York, Paris, Washington D.C.: Commission of the European Communities — Eurostat, IMF, OECD, United Nations, World Bank, 1993.
“World Wide Military Expenditures.” Data compilation from GlobalSecurity.org/CIA World Factbook, c. 2007-2008.