Who Can Still Steer? State Capacity from Beijing to Buenos Aires
Emiliano López
Earlier discussions have pointed out that dependency constitutes a measurable structure. The Structural Dependency Index illustrates a country’s position within the global hierarchy. However, merely having a position does not equate to the ability to maneuver. The dependency tradition has consistently posed a crucial question that mainstream economics often overlooks: when a nation encounters the structural limitations of its situation, what actions can its state actually undertake to address them?
The previous note left a lever and a question. Tracing the lithium chain from the salar to the imported battery, it showed that closing a productive circuit at home requires a definite policy architecture, the one China assembled: capital controls, conditioned foreign investment, public ownership of the strategic links, directed credit toward the value-dense stages. That is the instrument. But an instrument is inert without a hand to wield it, and the note ended on the harder half of the problem: not what to build, but which state can build it. This note measures the hand.
The State Mediating Capacity index, one of the three tools in this framework, evaluates this aspect. It does not assess economic size or the quality of “institutions” according to the World Bank’s definition; rather, it focuses on something more precise: the institutional resources a state has amassed to facilitate its integration into the global economy. This index is derived from what the dependency tradition considers vital levers for any developmental initiative: the productive capacity that the state can mobilize (such as manufacturing value added and public investment), along with the autonomy it has preserved over essential sectors (including control of the capital account and a central bank that works in coordination with national policy). A high score does not imply that a country is wealthy or even that it is developing effectively; instead, it indicates that the state has established and maintained the capacity to act.
Viewing it this way, the global capacity map diverges significantly from the rankings typically presented. In fact, it resembles the opposite.
The capacity has shifted east
Currently, the states exhibiting the highest mediating capacity are not situated within the OECD core but in Asia (Figure 1). China leads the Global South with a score of approximately 0.43, alongside Vietnam and Bangladesh, while India, Thailand, and Malaysia closely follow. Economies from South Asia and Southeast Asia boast the highest regional averages in the entire panel. In contrast, Latin America ranks significantly lower, with the affluent core even further down: both the United States and Germany occupy the lowest positions in the figure.
Figure 1. State Mediating Capacity across selected economies, 2023.
Source: Dependency Lab database, based on Global South Insights.
This situation is not merely a result of measurement methods. It reflects the institutional remnants of a developmental history: economies that established a manufacturing base under state coordination, retained significant control over their capital accounts despite the orthodoxy’s push for liberalization, and aligned their monetary authorities with national objectives. However, this cluster is not homogeneous, and the distinctions within it are crucial. China possesses the most profound mediating capacity in the Global South, which it has directed toward achieving autonomy: a vast productive base, an unparalleled share of public investment, a managed capital account, and the lowest structural dependency in the entire South. Meanwhile, Thailand, Malaysia, and Vietnam exhibit comparable mediating capacity but have utilized it for a different purpose: solidifying their roles within export value chains rather than diminishing a dependency that remains high. Both countries exhibit real capacity; one fosters autonomy, while the others manage subordination more effectively. The index evaluates the means; the question of how those means are utilized is separate and fundamentally important. The lithium note gave this distinction a concrete name: Thailand and Malaysia are the assembly-node trap, states with the institutional means to capture the value-dense links of a chain that, for now, they let pass through them instead.
This leads us to the other side of the narrative. Capacity is not a permanent feature, nor is it a guarantee of prosperity.
South Korea, often viewed as the archetypal “graduate” from the periphery and an OECD member with globally leading firms, has experienced the most significant decline in capacity within the panel (Figure 2). At the beginning of the century, its mediating capacity was near the top of the entire sample, around 0.50, but it was halved over the subsequent decade and a half, dropping to approximately 0.28 by the mid-2010s. The specifics of this decline are notable: following the 1997 crisis and subsequent IMF program, Korea liberalized its capital account, granted its central bank the insulation prized by orthodoxy, and allowed the developmental-state apparatus that had previously regulated the chaebol to detach from national planning. While it retained firms like Samsung and Hyundai, it lost the state that had nurtured them. The United States scores in the low-to-mid 0.1s, with Germany even lower. The hegemonic core does not require active mediation; it defines the terms against which others must mediate, resulting in the atrophy of these instruments. Paradoxically, the lowest levels of mediating capacity globally are found in its wealthiest economies.
Figure 2. State Mediating Capacity over time, 1995–2023.
Source: Dependency Lab database, based on Global South Insights.
This is the point at which the framework’s key caveat must be acknowledged, as suggested by the China-versus-platforms contrast. Beyond the uses aimed at building autonomy and consolidating integration, there exists a third use: capacity captured by rent. Myanmar leads the rankings due to its robust capital controls and a significant state-directed economy; various oil-rich nations, including Angola and Algeria, score highly because oil rents are reinvested through state investment vehicles rather than through productive transformation. The index establishes the means; determining whose interests these means serve is a separate matter, and one of paramount significance. However, one cannot even address this question in a state that has relinquished its means.
Latin America has lost capacity, but not entirely
This brings us to the region where the decline has been most pronounced and to a conclusion that complicates the conventional narrative of Latin American decline.
Latin America registers the lowest regional mediating capacity in the panel, with an average score of 0.21, lower than any Asian region and below Sub-Saharan Africa. The neoliberal dismantling is evident in the data. Chile, often regarded as the continent’s model, scores only 0.13, having lost more than half of its mediating capacity since 2000; Peru stands at 0.12. These economies opened their capital accounts, privatized public investment functions, and granted their central banks the kind of independence that effectively means independence from national objectives. They generate surplus but lack the ability to steer it.
Nevertheless, the region is not uniform, and the exceptions carry substantial analytical weight. Argentina is the clearest case of a state that has lost capacity without losing all of it. Its mediating capacity was developed throughout the 2000s, peaking at around 0.34 in 2014, supported by capital controls, an active role in public investment, and a monetary authority aligned with developmental goals. The post-2015 shift violently eroded that capacity; it plummeted to barely 0.15 by 2018, marking the sharpest two-year decline recorded by any country in the panel, before partially recovering to around 0.25 today. This decline is genuine and should not be understated. However, 0.25 still places Argentina significantly above the Latin American average and approximately double the levels of Chile or Peru. The institutional memory of a developmentalist history, the capital-account tools, the public banks, and the practice of treating the exchange rate as a policy variable rather than a definitive judgment have been diminished but not obliterated.
The crucial comparison is not with the lowest scores but with the middle: Colombia and Mexico are just below Argentina, at 0.24, yet they lack the developmentalist legacy that informs Argentina’s score. Their capacity has been assembled in a more recent and narrow context, rather than stemming from the well-worn remnants of a national-developmental period. Although the levels are nearly identical, the origins, and thus the potential for development, differ significantly. Brazil exhibits a similar pattern from a higher baseline: its capacity has decreased from around 0.41 in 2000 to 0.30 today, diminished yet still among the higher levels in the region. Its trap, as the lithium note argued, is not the assembly node but the stagnant matrix: a relatively complete industrial structure that capacity is failing to re-dynamise rather than one it was never built to create. The mediating capacity is there, eroding; what it has not been turned to is arresting the slide.
This is a key point to retain. The countries of the Southern Cone did not arrive at their current state empty-handed. What they maintain is precisely an inheritance, the remnants of earlier periods when the state was established to mediate rather than merely to manage subordination. This inheritance is being depleted, and the trajectory is as significant as the level. However, a residue provides a foundation from which to build, and it is something that the Chiles and Perus of the region no longer possess.
Why this matters
The mainstream toolkit fails to recognize all of this, as it does not measure these factors. Central bank independence is viewed as an unequivocal positive in its indices; capital controls are seen as distortions; and a significant share of public investment is interpreted as inefficiency. The dependency framework interprets the same facts as the architecture of mediating capacity, producing a map crucial to understanding autonomy: not who is wealthy, but who retains the ability to act within the constraints imposed by their situation.
However, the index measures a precondition, not an outcome, and this distinction forms the crux of the political argument. Consider the two movements traced by the framework. The first progresses from subordinate periphery toward the reconstruction of institutional capacity: rebuilding public control over finance, industrial policy capacity, fiscal sovereignty, and regulation of capital flows, effectively reconstructing precisely what the SMC measures. The second movement leads from there toward non-hegemonic autonomy: utilizing that accumulated capacity to subordinate external integration to a domestic developmental logic, which is what Amin referred to by delinking. The State Mediating Capacity serves as the tool for analyzing the first movement. It indicates who has assembled the means to initiate action.
Figure 3. The structural-dependency / mediating-capacity space, with the four configurations of the framework’s typology.
Source: Dependency Lab database, based on Global South Insights.
What it cannot predict is whether those means will be employed for the second movement or squandered. Here, the case of China, explored in detail elsewhere in these notes, is decisive: not as a model for imitation but as evidence that the second movement possesses a mechanism rather than an enigma. China’s structural dependency declined across all six dimensions simultaneously because the state maintained authority over the nodes where surplus is captured in the accumulation circuit: state-owned finance capital during the investment phase, and control over land, finance, utilities, and strategic commodities during the production phase. High mediating capacity aimed at autonomy manifests in a distinct way: a capital account managed through an unprecedented opening of trade, alongside the institutional ability to reproduce, on demand, the technology that the imperial core treats as a chokepoint. Integration deepened; sovereignty was preserved. This same capacity, however, in a state dominated by a comprador bourgeoisie reliant on a dependent relationship, effectively manages subordination, all while portraying it as modernization. The class configuration wielding these means differs significantly, even though the means themselves are identical. That illustrates why capacity is a necessary condition for autonomy but never sufficient.
This framework is also pertinent to Argentina. In the structural landscape, Argentina ranks high, with a Structural Dependency Index hovering around 0.37, placing it within the persistent dependency band, not far from Chile and Peru. Yet, the key differentiator lies not in its structural position but rather in its mediating capacity: standing at 0.25, whereas those of its neighbors have dwindled to half that level. Argentina has not reached the second movement nor has it moved beyond the periphery. Instead, it retains the means to pursue the first movement: public banks, capital-account instruments, and a developmentalist reflex, which, though eroded, have not been extinguished. The future of whether this inheritance deteriorates into the Chilean condition or is reconstructed into something akin to the Chinese model remains unwritten in the index. It is fundamentally a matter of which class forces dominate the state. The index cannot answer it. It can only tell you that the means are still there to be fought over.
More than a developmental state
It is easy to interpret all this as a revival of the developmental state—the bureaucratic structure praised by the revisionist literature on the East Asian miracle (Amsden, Wade, Chang) for its role in disciplining private capital and guiding late industrialization. The State Mediating Capacity index could thus serve as a measure of how much “developmental state” a nation possesses. However, this interpretation is not incorrect, but rather inadequate, and this inadequacy constitutes the crux of the argument.
The developmental state, in its traditional form, imposed discipline on capital to facilitate accumulation. It selected winners, conditioned credit, and protected nascent industries, all aimed at accelerating a capitalist catch-up whose logic, valuing capital, it never questioned and ultimately supported. Consequently, Korea’s apparatus was dismantled once it had fulfilled its role: after the chaebol achieved global competitiveness, the developmental state had completed its purpose and was reintegrated into the open, financialized order it had initially sought to challenge. A capacity designed to serve accumulation is relinquished when that accumulation no longer requires it.
This also encapsulates why the wealthy core scores lowest on the index, and it is essential to articulate this clearly. The point is not merely that the hegemonic centers have no dependency to mediate; it is that in the centers the state’s capacity is itself subordinated to the dynamic of capital. The apparatus has not diminished due to lack of use; rather, it has been reconfigured to facilitate accumulation instead of directing it: a central bank detached from national goals, an open capital account, public investment redefined as a subsidy for private valorization. Such a state can be incredibly capable, as evidenced by the United States, yet still exhibit low mediating capacity, since its abilities are aligned with capital’s self-propulsion instead of countering it. The low scores of the core do not indicate weakness but rather a specific subordination, with the state functioning as an instrument of accumulation in nearly its purest form. The developmental state temporarily subordinated itself to accumulation and was dismantled once the task was complete; in contrast, the core does so structurally and permanently. Both represent variations of the same subordination, and the framework’s proposition is its reversal.
What the framework seeks is something fundamentally different and more challenging: not the state that disciplines capital for accumulation, but rather the state that subordinates the accumulation dynamic itself to a long-term plan. This distinction is political rather than bureaucratic, best understood through Jessop’s strategic-relational interpretation of the state (Jessop; and behind him, Poulantzas). According to this perspective, the state is a social relation rather than a mere entity or subject; its institutional form is never neutral but inherently carries strategic selectivity, a built-in bias that favors certain strategies and class forces over others. The actors operating within this framework engage in strategic calculations shaped by the structural dynamics established through past struggles.
Three implications arise regarding how the index should be interpreted. First, mediating capacity is not a neutral toolkit but a selectively biased landscape: the same instruments—capital controls, public banks, directed credit—do not perform identically everywhere, as the social bloc concentrated within the state skews them toward one project or another. In the hands of a comprador bourgeoisie tied to a dependent relationship, they enforce subordination and label it modernization; in the hands of a bloc capable of subordinating accumulation to planning, they foster autonomy. Second, the legacy that Argentina and Brazil possess (Figure 4) reflects a structurally inscribed selectivity: the capital-account reflexes, public banks, and the developmentalist tendency to treat the exchange rate as a policy variable remain as a foundation that keeps a developmental strategy conceivable, even when no bloc is actively pursuing it. This is precisely why a residue constitutes something to build upon, while an exhausted apparatus does not. Third, this underscores why China’s experience is significant rather than exemplary: its trajectory is distinguished not by a more competent bureaucracy—both Korea and China had formidable systems—but by the establishment of a social bloc, distinct from prevailing Western and comprador configurations, where long-term planning is paramount and accumulation serves that purpose, accompanied by a strategic selectivity designed to maintain that relationship.
Figure 4. The composition of mediating capacity for four economies, latest available year, across the six components of the index.
Source: Dependency Lab database, based on Global South Insights.
The lithium note that we presented previously identified the policy architecture; however, the critical point is that the architecture remains inert without the bloc, and the bloc operates within a terrain selectively shaped by it. This aspect, rather than the size of the bureaucracy, is what the index fails to capture and what the next note must address.
The threshold
This delineates the threshold that the next note will traverse. The ability to mediate dependency has shifted to Asia, declined in the West, and persists in Latin America primarily as a legacy, with Argentina and Brazil drawing down from a still-high baseline, while others have completely depleted it.
The longer trajectory is unmistakable, and this arc will be the subject of the next note. The systematic decline of state capacity throughout Latin America is the cumulative result of half a century of neoliberal restructuring and right-wing governance, commencing in the 1970s and never fully reversed. The regional average plummeted from approximately 0.28 in the mid-1990s to barely 0.20 by the early 2000s, and throughout the subsequent progressive cycle, it never rebounded: at its peak, it reached 0.22 before resuming its decline. Individual recoveries were genuine, particularly Argentina’s, but they were localized and reversible, absorbed into a regional aggregate that remained stagnant. Progressive and popular governments mitigated the erosion in certain areas; however, they did not alter the overarching trend. This trend is driven by external factors: an imperialist offensive that, coinciding with the hegemonic project’s decline, deepens economic dependency and constricts the space available for any peripheral state to cultivate sovereignty and autonomy within the global accumulation circuit. The erosion of domestic capacity and the tightening of dependency from abroad represent two facets of the same process, which will be the focus of the next notes.
Capacity, in any case, serves as a means, and means do not dictate ends. The actual outcomes of a development path in individuals’ lives, whether mediating capacity translates into autonomy or merely results in a more comfortable form of subordination, is the essence of what the framework’s third tool, the Social Reproduction Outcomes, is designed to assess. Having the means constitutes a precondition; effectively utilizing them entails political action. The ultimate test is whether they benefit the people for whom they are ostensibly held.
Emiliano López is a researcher at CONICET–Universidad Nacional de La Plata and Chief Economist at Tricontinental: Institute for Social Research.
The State Mediating Capacity index is one of three tools within the Dependency Lab, alongside the Structural Dependency Index and the Social Reproduction Outcomes. All figures can be reproduced on our platform using Global South Insights.