How power is measured.
The Structural Power Project operationalizes Susan Strange's four structures of power — security, production, finance, and knowledge — across twelve nations. It scores the capacity to set the terms others must operate within, not mere possession.
The four structures
Strange held that structural power — the power to shape the frameworks within which states, firms, and people must act — runs through four channels. Each is a "side of the diamond"; together they decide who sets the terms. The Project scores each structure equally. (The fuller account is in the Report; this is the working definition.)
Who provides protection to whom, and on what terms. The structural question is not who holds the largest arsenal but who writes the order — who sets the rules of who may hold weapons, guarantees others' security, and commands the routes and norms everyone else must accept. The score is rule-authorship and protection-provision.
Who decides what is produced, by whom, and on what terms. Power here lies in controlling the indispensable input or chokepoint, governing the value chain, writing the trade rules, and capturing the surplus — not in raw output volume. The lever is the ability to grant or deny others access to the means of production.
Who controls the creation of credit and the structure of money. Strange's purest structure: the power to allow or deny access to capital, to denominate trade and debt, to rescue (or refuse to rescue) others' banks, and to exclude through the payment rails. The dollar's reach and lender-of-last-resort capacity sit here.
Who controls what is known, believed, and how it travels. Power flows to whoever authors the standards and platforms others must conform to, originates the frontier technology, can deny access to it, and shapes the beliefs and channels of communication. The score is authorship, denial, and the conferral of authority.
Scoring
Structural power is rule-setting — the power to shape the frameworks others operate within (Strange, 1988) — so components are scored absolute (0–95) by default: a nation's level on a lever, independent of the others. The two genuinely zero-sum levers (a formal IMF/World-Bank vote share, command of a single maritime chokepoint) are the only ones scored positionally as a share-of-universe. A metric is the mean of its scored components; a structure is the equal-weight mean of its metrics; the headline is the absolute level — the mean of a nation's four structure scores, on the same 0–95 scale. A secondary share-among-the-twelve figure is kept for bloc framing. The universe is disclosed: these twelve nations, not the world.
Metrics by structure
Each metric links to its definition, its grounding in Strange's text, and its component sources.
The dataset
The full machine-readable dataset — every nation, metric, and component, with scores and per-cell reasoning — is published as a single file. A reader who would weight the structures differently, or recompute the headline on a different basis, can do so directly from it.
Download data.json · the wire format consumed by this site
Known limitations
This index is circulated for scholarly feedback. Three limitations are disclosed openly rather than hidden.
⚑ The proxy–possession problem
Strange warned—of the knowledge structure, in words that apply index-wide—that such power is "unquantifiable" and that "the indicators that can be found are only the roughest of guides." The deeper difficulty: the moment a metric reaches for a proxy, that proxy carries some possession basis — because the data the world publishes is possession data (stock, share, volume), while the structural fact (who shapes whom) is rarely measured directly. Our metrics are framed structurally — every component asks who authors the rule or wields the denial — but at the point of measurement, leakage enters. A component-level audit grades 36 of 60 components fully structural, 19 mixed, and 5 possession-dominated; roughly 40% carry some leakage, concentrated in finance and production (reserve-currency share, outward-FDI stock, asset ownership). Some of this is fixable by rescoring behaviour rather than stock (transnational-firm-power, technological-primacy); some is irreducible — the share of trade forced into a currency simply is the denomination fact, and a chokepoint cannot be held without the resource. We put this to scholars as an open question: how should a structural-power index handle levers whose only available proxy is a possession measure?
⚑ Bottom-tier score inflation
Provision components are normalised against the leader (raw ÷ max × 95), which compresses the lower band: a nation with little structural role on a lever can still surface in the 15–25 range rather than near the floor it deserves. An audit of the four lowest-ranked nations found a recurring pattern of ungrounded "participation" scores — cells where the rationale prose never even names the country. A structure-by-structure correction pass has been applied: fifty such cells were re-grounded downward against each component's own evidence, while genuinely held levers (Russia's energy and OPEC+ weight, India's GVC and IETF role, South Africa's platinum chokepoint) were kept. The deeper proxy-possession leakage above remains; bottom-tier ordering should still be read as indicative, not precise.
⚑ EU attribution
Euro authorship levers (trade-rule authorship) are split among Germany/France/Italy by EU Council population weight; euro exposure levers (reserves, payment rails, the benchmark curve) remain attributed full-bloc, since a counterparty faces the euro as one indivisible unit. The mix is principled but not the only defensible rule, and one consequence — post-Brexit Britain reading below France and Germany in finance — is disclosed as a known artifact of the full-bloc-exposure choice.