The dominant literature on economic security, weaponisation and chokepoints, both in the public policy debate and in much of international relations research, remains trapped within a relatively shallow topography: ports, maritime straits, tariffs, sanctions lists, and, at most, dependence on first-tier (Tier 1) suppliers. This topography is misleading. It implicitly assumes that the vulnerability of a supply chain is proportional to its visibility, that what can be mapped is also what matters. The argument developed in this section reverses this assumption: the most decisive links in a production chain tend to be the least visible ones, located in Tier 3, Tier 4, or beyond, occupied by specialised or non-substitutable suppliers in the short or medium term, and opaque within supply chains. These nodes are referred to here as 'deep chokepoints'.
The thesis is that contemporary economic sovereignty is decided in the asymmetry of knowledge regarding the very architecture of production. Whoever is able to see, and, from that point onward, control the irreplaceable link in a supply chain acquires an instrument of coercion that is independent of the aggregate weight of their economy. However, this section introduces an additional hypothesis: that the opacity which makes these nodes vulnerable is, simultaneously, a strategic resource that can be deliberately produced and managed. To develop these propositions, this section articulates two bodies of literature that rarely engage with one another: the Anglo-American theory of structural power in networks, which explains why certain nodes matter more than others, and the French doctrine of economic intelligence, which describes how this knowledge is effectively produced, protected, and operationalised by state apparatuses.
Farrell and Newman (2019) begin with a simple observation regarding the topology of global economic networks: they grow asymmetrically, generating nodes with a degree of connectivity, in terms of network theory, of 'betweenness centrality' far greater than that of others. States exercising effective jurisdiction over these central nodes, and possessing domestic institutions capable of mobilising them, can convert this centrality into two types of coercive advantage. The first, the panopticon effect, enables the systematic surveillance of information and transaction flows passing through the node. The paradigmatic case is the United States' access to data from the Society for Worldwide Interbank Financial Telecommunication (SWIFT) financial messaging system after 2001, which made it possible to trace illicit financing networks. The second, the chokepoint effect, enables the denial of an adversary's access to the central node; SWIFT itself, in 2012, was mobilised by the United States and Europe to exclude Iranian banks from the international financial network, as part of the sanctions campaign against Iran's nuclear programme.
The theoretically relevant point is that this power depends exclusively on the structural position of the node within the network. It is this mechanism that this section seeks to shift from the domain of financial and informational networks (the original object of Farrell and Newman) to the domain of physical production. The question left unanswered in the authors' original formulation is: what replaces 'network centrality' when dealing with components, inputs, and engineering capabilities? The answer lies in a second body of theory, more recent and more formally oriented.
Clayton, Maggiori, and Schreger (2024), in their model of economic coercion and fragmentation, provide this bridge. The authors formally demonstrate that the coercive power associated with control over a sectoral input is non-linear: it grows disproportionately as a State approaches control over the total supply of that input, rather than gradually and proportionally to its market share. A supplier holding 60% of the market for a critical input possesses bargaining power that is more than proportionally greater than one holding 40%, while one approaching an effective monopoly, even over a technically trivial and low value-added input, may exercise coercive power disconnected from its weight in aggregate bilateral trade. It is this mechanism, referred to by part of the subsequent literature as 'granular power', that explains why control over apparently marginal links in the chain (a specific resin, an optical subcomponent, a metallic alloy produced by a single group of suppliers) can generate geopolitical leverage disproportionate to its nominal economic weight.
One empirical corollary advanced by the authors deserves attention: because coercive power only becomes extreme when approaching total control over an input, a substantial share of a country's economic security may be achieved through the diversification of sources, without the need to pursue decoupling (complete fragmentation) of the supply chain. The same authors further observe a structural asymmetry between great powers: the geoeconomic power of the United States rests primarily on its centrality in financial services, whereas Chinese power rests on its manufacturing centrality, a distinction that helps explain why the instruments of coercion mobilised by each power tend to target different sectors of the supply chain.
If the Anglo-American literature explains the structural logic of power in networks, it remains relatively silent regarding the process through which such knowledge is produced - that is, regarding the concrete practice of intelligence that precedes any act of coercion. It is here that the French school of guerre économique, associated above all with Christian Harbulot, offers an older and operationally more detailed genealogy.
Harbulot constructs his argument beginning in the mid-1990s, in a context in which the very idea of economic warfare was treated as taboo in the French public and academic debate, deemed incompatible with the liberal-cooperative vocabulary (Keohane and Nye) then dominant in discussions of globalisation. His first major study, prepared for the French Ministry of Research and consolidated in La machine de guerre économique: États-Unis, Japon, Europe (Harbulot, 1991), advances a thesis that today appears less eccentric than it did at the time: with the end of the Cold War, confrontation among great powers changed terrain, migrating from the military sphere to the economic-informational sphere. In later works, Manuel d'intelligence économique (Harbulot, 2012), L'art de la guerre économique (Harbulot, 2019), and La guerre économique au XXIe siècle (Harbulot, 2024), the author systematises an operational method based on four articulated functions: monitor (systematic monitoring of open and closed sources), analyse (the transformation of raw data into actionable intelligence), protect (the defence of one's own informational and technological assets), and influence (the capacity to shape the norms, narratives, and decisions of others). These four functions are dimensions that operate simultaneously and reinforce one another.
Guilhon and Moinet (2016), in their reference volume for French higher education, revisit the operational triad (inform oneself, protect oneself, influence) as the organising structure of the field, while changing Harbulot's focus (more centered on the State) toward a perspective that simultaneously encompasses the firm, the territory, and the State as legitimate scales of action. Fontanel and Fontanel (2013) add a dimension by showing that, in the French experience, economic intelligence has never been entirely public nor entirely private, but a hybrid arrangement in which the State provides the institutional framework while the primary collection of data is increasingly outsourced to private consultancies and due diligence firms. This public-private hybridisation is structurally identical to that identified, in another language, by Walton and Quinlan (2024): the authors observe that intelligence agencies have always treated supply chains as targets, but that the novelty of the present moment lies in the systematic mobilisation of private value chain management actors to produce, on behalf of Western governments, the mapping of their own structural vulnerabilities.
From the articulation of these two theoretical bodies, the structural and granular power of networks and the institutionalised practice of economic intelligence, emerges an analytical framework composed of three movements, each corresponding to a distinct stage in the conversion of structural opacity into an instrument of statecraft. To these three movements, a final reflective layer is added, shifting the conventional diagnosis.
The first movement consists of recognising and documenting that the opacity of the deeper layers of the supply chain derives from a structural property resulting from decades of cascading outsourcing, in which each contractual link is only required to know the immediately preceding one. The result is that the final purchasing firm frequently remains completely unaware of the identity of suppliers located in Tier 3 or Tier 4, the level at which the most specialised and least substitutable technical inputs are concentrated. This asymmetry, between what the firm knows about its own supply chain and what a State endowed with dedicated intelligence capabilities can discover about it, is the raw material upon which every subsequent coercive advantage is built.
The second movement converts this documented opacity into operational knowledge through two complementary types of operation. The technical decomposition of inputs consists of physically or technically disassembling a final product in order to identify its inputs and trace their actual origin, an exercise in reverse engineering that also relies on the analysis of patent filings as a public record of specific technical capabilities, as suggested by Guilhon and Moinet when they place intellectual property analysis at the core of the 'inform oneself' function. The mapping of control structures, in turn, consists of reconstructing the actual chain of ownership, financing, and influence over critical suppliers, going beyond the declared contractual chain in order to identify who effectively controls production and export decisions. This is the boundary between legitimate economic intelligence and industrial espionage, which becomes deliberately ambiguous, and it is also the domain in which, according to Walton and Quinlan (2024), Western States have historically invested less than their rivals.
The third movement is the effective conversion of informational advantage into bargaining power or coercion. Once the low-substitutability link has been identified, the available instruments range from export controls and licensing regimes to restrictions on access to financial or payment markets - replicating, at the scale of the specific input, the same logic as the chokepoint effect described by Farrell and Newman in the case of SWIFT. The specific contribution of Clayton, Maggiori, and Schreger (2024) at this point is to demonstrate that the intensity of the coercion obtained is proportional to the proximity of total control over the supply of the input, which explains why States with modest aggregate economic power may nevertheless possess disproportionate coercive capacity if they control a sufficiently concentrated link. The implication for States in the position of potential targets is relevant because vulnerability grows in a non-linear manner. The optimal response may instead be achieved through diversification directed toward the chokepoints identified in the first and second movements.
The three movements described above assume, either implicitly or explicitly, that opacity is a flaw to be corrected by intelligence, and that the role of the State is to illuminate what firms themselves ignore. This subsection introduces a hypothesis that complicates this diagnosis. The opacity of the lower strata of globalised production is, to a significant extent, deliberately produced and maintained by an ecosystem of intermediaries (traders, integrators, holdings, offshore jurisdictions) whose business model depends on non-transparency. There exists a private industry of opacity that extracts rents from the ignorance of others, and the purchasing firm pays, through risk premiums and intermediary margins, for the privilege of not knowing - and of being able to claim that it did not know.
Once this political economy of ignorance is recognised, the role of the State reveals itself to be more ambivalent than the intelligence literature on geoeconomics usually acknowledges. The same state apparatus that invests in intelligence to map its adversaries is, simultaneously, an active producer of opacity within its own strategic layers. Government agencies orchestrate, or tolerate, the creation of opaque corporate vehicles, front suppliers, and indecipherable ownership chains in order to shield sensitive sectors from foreign observation. Structural disinformation, the injection of noise into a competitor's tracking systems, is a first-order offensive tactic.
This observation makes it possible to expand the typology proposed by Farrell and Newman (2019). To the panopticon effect (to monitor) and the chokepoint effect (to deny), one may add a third, which we shall call the labyrinth effect (to confuse): the capacity of a State to generate artificial complexity within its own supply chains so that the adversary expends intelligence resources on dead ends while the true bottlenecks remain concealed.
Finally, there is an epistemological consequence of this inversion. If opacity can be produced as an asset, then the very concept of a 'chokepoint' comes to be understood as a construction: a supplier becomes a deep chokepoint when an actor possessing intelligence capabilities constitutes it as such, while simultaneously keeping other equally vulnerable nodes in the shadows.
The public-private partnership that Fontanel and Fontanel (2013) identify as central to economic intelligence acquires its full meaning: governing collective ignorance about supply chains is just as strategic as producing knowledge about them. The limits of national economic autonomy are ultimately defined by the depth with which a State is capable of coordinating its policy and decentralising to firms, the frontline of geoeconomics, the operationalisation of its objectives.
