Gillian Hart in Beijing: Negotiating capitalist models at the World Bank–China nexus

Recent formulations of state capitalism tend to present it as a distinct system, anchored in China, in opposition to a neoliberal model represented by the United States and its global armatures. As an alternative to this binarism, this paper argues for Gillian Hart's relational-comparative approach to the geographies of the new state capitalism. It outlines three of Hart's theoretical-methodological principles—multiple trajectories, conjunctural analysis, and articulation—and demonstrates how they can be used to analyze the interrelations between “statist” and “liberal” development trajectories, through an empirical account of conjunctural struggles in and between China and the World Bank during the Tiananmen Square crisis. It argues that Tiananmen was an inflection point in the relation between the development trajectories of the Chinese state and the World Bank, where conflicts over the continuity of economic reform were simultaneously struggles over the boundaries of the state. Examining these institutions conjuncturally shows that China's “state capitalism” is not an opposite of the Bank's liberal model, but has been, in part, produced through power-laden contests over the meaning and materiality of state and market in multiple arenas.


Introduction
The world, it seems, is once again being sorted into two economic models. Liberal capitalism, comprised of privately owned corporations, a regulatory state, and the rule of law, has stagnated since the 2008 Financial Crisis, while a new, muscular contender threatens its dominance. This model is characterized by gargantuan state-owned enterprises (SOEs), a bureaucratic developmental state, and an autocratic political class able to secure legitimacy through sustaining high rates of growth. The two are often imagined as opposites: the former model manages crisis indirectly by nudging markets through fiscal and monetary policy, while the latter-a new kind of "state capitalism"-does so through direct ownership and mobilization of the means of production. According to both economists and the business press, this grand rivalry threatens to encompass the whole globe, subordinating others within their gravitational pull (Kurlantzick, 2016;Milanovic, 2019). In these accounts, the former is exemplified by the United States and its global financial architecture, the latter by China and its internationalizing SOEs.
Such formulations often lapse into Eurocentrism and fail to recognize capitalism's uneven development (Alami and Dixon, 2020a). Dualistic models understate the mutually constitutive and dynamic character of Chinese and Washington-led forms of governance, freezing into independent ideal types what are two of many internally differentiated and co-evolving régimes. As one response to these "binary historical geographies," (Peck, 2021: 5) this paper argues for an alternative entry point for geographers interested in the relations between configurations of state and capital, inspired by Gillian Hart's postcolonial Marxist approach and her method of relational comparison. Built as a critique of comparative methods which frame cases as taxonomic units, relational comparison attends to the intertwined, grounded transformations of "connected yet distinctively different nodes in globally connected historical geographies," taking uneven development as a premise and outcome of analysis (Hart, 2018a: 373).
Reframing debates over competing capitalist models in a relational-comparative lens demands a series of moves away from what Hart (2018a: 376, following McMichael, 1990) calls analytical comparison-where discrete national economies are understood as instances of distinct capitalist varieties-toward a spatially sensitive conjunctural analysis of how development trajectories are forged in relation to one another. To this end, the paper outlines three maneuvers: first, a situation of cases in their intertwined historical-geographical trajectories; second, an analysis of political struggles in determinate conjunctures; third, an explanation of how these struggles rearticulate alliances, positions, and connections, and in turn rework places, identities, and trajectories. 1 It demonstrates this approach through an analysis of the inter-institutional negotiation of SOE reform strategies at the nexus of the World Bank and China's party state during the Tiananmen Square crisis. This empirical entry point is a strategic one. Both institutions are often approached as representatives of competing development models-the World Bank as the embodiment of the Washington Consensus and the Chinese state as the Beijing Consensus-and as the respective faces of liberal and state capitalism.
A Hartian analytic, however, finds these encompassing abstractions certainly insufficient and potentially misleading when applied to concrete "nodes." Tiananmen was a moment of the conjunctural remaking of the World Bank-China relation that belies accounts of China and the Washington Consensus as monolithic units, identifiable in terms of a perennial pro-statist/pro-market divide. It was, rather, a rupture that portended the transformation of China's political economy (Huang, 2008: 19-24;Hung, 2015: 61-69;So and Chu, 2016: 65-71), sparking governance conflicts in and beyond China. Within the Chinese Communist Party (CCP), the crisis aggravated factional struggles between conservatives and reformists over the consequences of marketization and center-local relations, which refracted through the World Bank as reanimated tension between its headquarters and the China Office over the viability of gradual reform. The outcome was a contested rearticulation of state and market policy rationalities within both the Chinese state and the World Bank that changed their subsequent trajectories and relations to each other, realigning them as accordant on the developmental role of markets but discordant on the institutional context of markets, hinging on the status of direct state ownership and control of productive capital. A Hartian approach not only provides a spatialized, dialectical account of the Chinese state and the World Bank (cf. Bottelier, 2018;Gewirtz, 2017;Weber, 2020), but suggests a methodological and theoretical turn for studies of the new state capitalism away from a liberal-statist dualism and toward granular, political, and processual accounts that capture "the dialectical and cumulative unfolding of different modalities of state intervention across space, scale and time, by tracing the different forms of interconnection between them" (Alami and Dixon, 2021: 4).
The next section details how recent discussions of state capitalism maintain the tendency to idealize both China and "the West" as stable, opposing models, ossifying the dynamism of their political economies, obfuscating their interdependencies, and exaggerating their differences. The third section argues that Gillian Hart's methodological strategies can bring into view how these two trajectories have been actively shaped through interconnected, grounded political struggles. The fourth section demonstrates these strategies by placing the historical geographies of the World Bank and the Chinese state in one frame, positioning Tiananmen as a conjuncture when their trajectories were contested, institutionalized, and consolidated in relation to each other. The conclusion makes the case for state capitalism studies to engage with the complexity, contradictions, and (inter)connections inside and between the asymmetrical social-institutional formations that are often taken for granted as "statist" and "neoliberal."

Disabling binary geographies
In response to the contentious post-Fordism debates of the 1990s, Hart (1998) problematized two prominent claims about industrial restructuring: that global changes are being driven by restructuring in the core; and that specific sector evince the beginnings of an all-encompassing transition from mass production-based Fordism to flexible specialization-based post-Fordism. These problems, for Hart, were both conceptual and political. The first charge, against stylized "impact models," critiques the literature's sharp local-global dualism that presents globalization as an uncontestable force diffusing from the global North, only to be experienced by passive recipient communities elsewhere (Hart, 2002: 12-13;2004: 91). The second charge claims that "binary histories" universalize what is in fact a particular transformation, proving themselves incapable of recognizing the diversity of regional economic transformations and relational dynamics (Hart, 1998: 334;Sayer, 1989).
Hart's critique of dualistic and Eurocentric theory claims, made in light of South Africa's particular industrial trajectory, could well have been directed toward discussions of China's economy. In the same era, there was a widespread expectation that China would transition toward a liberal market economy, replete with privately owned firms and open financial markets. Greater market coordination would necessarily result in reduced transactions costs for private firms while increasing the costs for public or hybrid firms, and integrating with the world market would require harmonization with its property norms (Chai, 1998;Nee, 1989;Sachs and Woo, 1994). But sustained growth in foreign direct investment, the development of labor markets, and the emergence of private business have led neither to the disappearance of China's state sector nor to the liberalization of its financial markets, at least not in line with Western expectations (Huang, 2008;Lardy, 2014Lardy, , 2019. Since then, conceptions of the "China model," the "Beijing Consensus," and now "state capitalism" appear to supersede the twin fallacies of convergence and the binary history of central planning giving way to a market economy (Ferchen, 2013).
However, if it is now accepted that there is no singular capitalist trajectory, the problems of universalization and dualism have re-emerged in more recent macroeconomic imaginaries. Perhaps most noteworthy is Branko Milanovic's account of a tendential global polarity between "liberal meritocratic capitalism," denoted by private ownership of means of production and markets as the principal mode of coordination, and what he terms, after Weber, "political capitalism," combining private sector dynamism with a rational, bureaucratic party state (Milanovic, 2019: 91). While all countries are predicted to sway toward one of these models, Milanovic (2019: 10) claims they are represented most adequately "by the United States and China." Similar notions of a "state-led variety of capitalism" in comparative political economy theoretically position predominantly Asian late developers against a liberal-market economy model represented by Anglo-America (Musacchio et al., 2015;Walter and Zhang, 2012). These maneuvers entail a related tendency to select particular forms of intervention as evidence of "more state," while other forms are normalized or "hidden" (Alami and Dixon, 2020b;Peck, 2021). By reifying institutional forms that are geohistorically co-produced through both "state" and "market" processes, the binary history of central planning to market economy is reinscribed as a binary geography of state and liberal capitalisms.
Economic geographers have responded through the "variegated capitalism" framework, a radical institutionalist approach sensitive to inter-scalar dynamics and relational co-dependence. Built as a critique of a different binary model-that of the coordinated versus liberal market economy of the varieties of capitalism literature-these scholars have used the Chinese case to stress-test that approach's methodological internalism. Instead, they argue that explaining China's political economy requires attending to the path-dependent and multi-scalar institutional dynamics within and beyond China, constituted as it is by international trade, foreign direct investment, and localized experimentation (Lim, 2014;Zhang and Peck, 2016;see also McNally, 2019). A key injunction from this literature is the necessity for conjunctural approaches that assume no universal form of either the state or the market. Consequently, its problematization of institutional diversity and sociospatial complexity has helped to dislodge convergence narratives of post-socialist transition and subnational restructuring paths through cases in east and southeast Asia (Kenney-Lazar and Mark, 2021;Lim, 2019;Mulvad, 2015). A complementary "macro-constructivist" strand of geographical political economy employs Trotsky's notion of uneven and combined development to position China's state-led late industrialization in dialectical tension with the U.S.-led liberal international capitalist order, with attendant emphasis on geopolitical and inter-societal interaction (Peck, 2021: 12;see Dunford et al., 2021;Rolf 2021).
Yet, to provide an alternative to the competing models framing, it is not only important to reconstruct national models of "state capitalism" as relationally constituted and internally differentiated institutional hybrids of state and market logics; it is equally important to trace how apparently "statist" and "liberal" configurations are produced through their concrete interconnections. Following Hart (1998: 340), this means going beyond simply asking what the rules of relational structures are, the focus in addition is on how interaction takes place within and among social arenas, and on how definitions of social institutions and their boundaries are also the object of contestation.
The next section outlines the rudiments of this approach through three of Hart's distinct theoreticalmethodological moves, with particular focus on how they can enrich geographical analyses of state/ capital relations by attending to the constitution of institutional configurations and "state intervention" through power-laden practices in multiple interconnected arenas.

Relational comparison reiterated
Over the past three decades, Gillian Hart has developed a distinctive method as a rejoinder to the same types of dualistic and Eurocentric fallacies in political-economic geography that appear in extant studies of state and liberal capitalism (Ekers et al., 2020;Hart, 2018aHart, , 2018b. Relational comparison, informed by critical spatial theory, a Gramscian political sensibility, and agrarian political economy, situates cases as nodes within an interconnected historical geography, tracing how general processes are constituted by particular social struggles, practices, and relations in multiple arenas (Hart, 2002(Hart, , 2006(Hart, , 2018a. While any account of Hart's oeuvre is inevitably partial (see Ekers et al., 2020), three themes in her work help move beyond the "competing models" discourses that frame discussions of the new state capitalism by shifting the optic toward historically grounded practices and constitutive forms of power. First, by approaching empirical sites through their relations, it substitutes the search for models with a focus on interconnected trajectories of intracapitalist restructuring; second, by attending to power-laden material and meaningful practices in multiple arenas, it approaches restructuring and the state politically; third, by adopting an ontology of articulation, it centers relational connections between states and social forces, their conjunctural remaking, and their potential for transformation. Hart's (1998Hart's ( , 2002 conception of multiple trajectories draws from engagements with the work of Henri Lefebvre and Doreen Massey. It was originally posed as a challenge to orthodox narratives of industrial restructuring and, later, globalization that portrayed these processes as "impacting" places. Case studies in this vein problematically treated local processes as contingent outcomes of more general, global processes, thereby conflating the level of abstraction with geographical scale (Hart, 2018b). Instead, Hart argues that places-and "local" processes-must be considered constitutively global, produced through material and symbolic social practices, and historically situated in relation to processes unfolding elsewhere. From this perspective, formulations that approach "state capitalisms" as "the product of communist revolutions conducted in societies that were colonized or de facto colonized" (Milanovic, 2019: 67) or the result of "failed" post-socialist reforms (Musacchio et al., 2015: 115) are dubious, insofar as they tend to explain development on the basis of internal evolution, obscuring their imbrication in extra-local processes. The strategy is not to conceptualize variants of capitalism that unfold through singular logics, but to analyze how particular forces come together to remake development trajectories in particular places, with global reverberations. Doing so redirects empirical attention to path-shaping conjunctures (Hart, 2020).
When attending to these conjunctures, Hart focuses on their determination by "relations of force at various levels" (Hart, 2018a: 388), which is centrally figured by her Gramscian conception of the social terrain. 2 A key notion inherited from Gramsci is a conception of politics operating through hegemony, a process through which particular groups struggle for social power through mobilizing ideas and language, reconstituting identities, and instituting practices that attempt to neutralize antagonism to their rule on the basis of consent and coercion, but ultimately only displace rather than eliminate contradiction (see also Thomas, 2009: 160-7). For Hart, as for Gramsci, the state figures not as a "unitary, coherent and homogeneous" thing, but as a complex ensemble of institutionalized relations and a key site of hegemonic struggle and contradiction in class-divided social formations (Gramsci, 1971: 342;Hart, 2002: 26). Forms of "intervention," should, then, be understood in relation to struggles for hegemony, operating through and beyond the state.
Central to Hart's Gramscian approach is the notion of articulation, which inflects both her conception of space-time and of power and subjectivity. Drawing on Stuart Hall's ( [1980) reinterpretation of the Althusserian concept, Hart uses articulation to name two relations. First, it refers to the unstable connections between relations of production and complexly structured social practices, interests, and beliefs that hegemony implies. For Hart, articulation is the way in which "material and meaningful relations cohere"-or give expression to one another-in particular historical-geographical conjunctures (see Ekers et al., 2020Ekers et al., : 1581Hart, 2002Hart, : 26-33, 2007. Since their constitutive articulations are non-necessary, hegemonies are by definition contradictory formations which must be constantly defended, remade, and reorganized. Working on the terrain of hegemony, then, requires a non-reductionist approach to political-economic analysis, instead of demanding "attention to how material economic practices, power relations, and the production of meaning and difference constantly play upon one another" (Hart, 2002: 297).
A final theoretical point-which aligns with the second notion of articulation-is that Hart does not presume determination of particular cases through an "a priori general or universal process," (Hart, 2018a: 381) but instead understands them as complex unities-in-diversity, articulated from the interconnections between spatially and historically specific developments unfolding in various sites. 3 Indeed, this directly follows from Hart's rejection of general abstractions as sufficient explanatory mechanisms for concrete developments. Articulation as a methodological injunction approaches general processes as the provisional, unstable, and contestable outcome of grounded social relations and struggles (see also Werner, 2019). By extension, it cautions against conceiving of divergent development trajectories as the result of a singular endogenous dynamic or external force as much as it disqualifies binary conceptions of state versus market. Instead, approaching state intervention and restructuring as a process of articulation reframes it as an uneven process of recombining state and market logics in novel ways, forged through active processes of struggle, acquiescence, negotiation, and contestation in multiple connected arenas.
Relational comparison follows from these theoretical prescripts. By maintaining an extraverted sense of place with attention to the potentially transformative play of social forces within and between them, it centers on how specific trajectories are shaped by interconnected hegemonic struggles. It is admittedly less useful for delineating the internal structures of different economic systems or for producing typologies of state intervention, though these are vital for understanding state capitalism as an institutionalized social form (Nölke et al., 2019;Petry et al., 2021). For this reason, this approach is best understood as one possible methodological alternative for querying the production of different capitalist configurations. But because it does not assume a priori abstractions like "state" or "liberal" capitalism, relational comparison demands different entry points. The next section illuminates these principles through an analysis of the World Bank-China nexus centered on the Tiananmen Square crisis, but before doing so, explicates the rationale for this particular entry point.

China/World Bank refractions
If Hart's methods undermine dualistic and teleological conceptions of global processes by tracing the concrete, relational practices through which development trajectories are remade, there are many possible entry points for problematizing the binary geographies of state and liberal capitalism. The one taken in this paper is the World Bank-China nexus at an inflection point in their relation, the Tiananmen Square crisis. Indeed, the liberal/statist binaries imposed on the economies of China and the United States often extend to their development armatures, with the "Beijing vs. Washington Consensus" indexing this inclination (Ferchen, 2013;Huang, 2010). Studies comparing their development financing models have mostly avoided this rigid schema, yet it persists in discussions of their "dual transformations," whereby China embraces market-based instruments of the "Western model" while "Western institutions" like the Bank have turned toward institutionbuilding and stakeholder engagement strategies more characteristic of Chinese development finance practices (Chin and Gallagher, 2019;Groce and Köstem, 2021). While this paper does not compare financing models, a relational-comparative lens subverts the ideal-typical framing of "Western neoliberalism" nor "China's state capitalism" by positioning these as unstable articulations within a complexly structured world rather than as discrete, static entities. 4 Concretely, the paper traces how China's trajectory of economic reform and the World Bank's trajectory of neoliberal development were both transformed in relation to one another during the Tiananmen Square crisis, using recently-released (January 2020) country strategy program reports and correspondence from the Bank's digital archive. 5 It does so through three cuts, each employing a distinct Hartian move to illuminate how China's "gradualist" reform strategy emerged, consolidated, and came into conflict with the World Bank in this conjuncture. The first cut, using the method of multiple trajectories, briefly historicizes the World Bank-China nexus in relation to the shifting contours of Chinese socialism and Bretton Woods liberalism during the 1970s and 1980s, showing how their uneven development during this era facilitated new circuits of finance and expertise between the reformist bloc of the Chinese state and the China Office of the World Bank. The second cut, through a conjunctural analysis of the Tiananmen Square crisis, analyzes how intra-party struggles over economic reform reverberated through the World Bank as a struggle between the headquarters and the China Office over the viability of gradual SOE reform, with the boundaries and definitions of the state and the market becoming objects of contention within each institution and between them (see Meulbroek, 2022). The third cut shows how, with the unstable resolution of these struggles, the World Bank and the Chinese state were rearticulated-they now found themselves in a new relation with each other, expressing different logics of market and the state.

Multiple trajectories: The encounters of late-Maoist China and McNamara's World Bank
The Republic of China was a founding member of the World Bank, but the latter's enrollment in anti-communist liberal internationalism meant that, upon the establishment of the People's Republic in 1949, representation in the Bretton Woods institutions moved with the Nationalist government in Taipei. During the Cold War, mainland China began a period of relative autarky. A full overview of China's political economy under Mao is not possible here, but suffice to say, but a few points are contextually important for its subsequent relation to the World Bank. First, China was never governed under a uniform system of "planned economy" but instead most SOEs were under the control of local governments (Lim, 2017: 70). While responsible for producing the bulk of the nation's output under prices formally set by the State Pricing Commission, SOEs were under informal pressure to overproduce beyond the production targets set by national-level plans, using market mechanisms like piece wages to achieve them, resulting in uneven regional development between the industrial northeast and the Western provinces (Li, 2008;Lim, 2019). Second, China's industrialization drives were made possible by imports of plant and equipment from the Soviet Union in the 1950s, funded through large procurements of grain from peasants toward urban consumption and export (Ash, 2006). This strategy managed to achieve gains that had not been possible under informal colonialism, with industry as a share of GDP growing from 20.9% to 47.9% and industrial productivity growing 236% from 1953 to 1978. However, agriculture, functioning as the accumulation fund for urban industrialization, only witnessed 25.5% productivity growth over the same period (Hung, 2015: 45).
China's state-socialist model faced significant challenges all along, especially from unstable geopolitical relations with the Soviet Union beginning in the late 1950s. But by the late 1970s, after the displacement of labor power and fixed capital toward small-scale production during the Cultural Revolution, lower-than-projected petroleum discoveries, and growing trade deficits from equipment imports after the resumption of trade with the West, China's political economy was in a state of stagnation. Deng Xiaoping and the pro-reformist faction outmaneuvered the Gang of Four in succeeding Mao in the late 1970s, and, to avert food riots, upscaled finance minister Zhao Ziyang's "Sichuan model"-granting rural enterprises the ability to sell excess agricultural production at their own prices-to national policy under the "dual-track system". The party then set its sights on reviving productivity in the urban industrial sector as a hedge against social discontent and incipient Soviet aggression on the continent (Rolf, 2021: 96-98;Vogel, 2011: 454). The pro-reformist faction sought expertise and capital through the World Bank to assist in a program of generalized industrial and macroeconomic restructuring. 6 The Chinese state (re-)encountered the World Bank at the culmination of a period of rapid organizational change at the Bank during the 1970s. During Robert McNamara's tenure as President (1968President ( -1981, it had transitioned from a conservative, productivist, project-based lending bureaucracy to a large quantitative economic research center and development agency focused on devising national policy strategies. The geopolitical-economic threat of decolonization, economic nationalism, and mounting inequality in the developing world led the former Secretary of Defense to expand the Bank's agricultural and social infrastructure lending, with new programs focused on agricultural credit provision, public health, family planning, sanitation, housing provision, and education aimed at the political neutralization of the peasantry (Pereira, 2020;Sharma, 2017). After 1973, the availability of cheap credit through recycled petrodollars reduced the Bank's stature as lender, prompting McNamara's efforts to promote its role as a provider of technical assistance and research. It had also been extending its reach from the "Third" to the "Second World" that decade, advising the socialist Hungarian and Vietnamese governments, which, in turn, persuaded China to consult the Bank to assist with the sequence and execution of economic policy reform at the national scale (Gewirtz, 2017). Yet China was unique from other socialist and non-aligned polities seeking assistance from the Bank. As a result of Mao's strategy of "self-reliance," China's stock of debt was less than three percent of gross national income by 1981, while governments that had borrowed recycled petrodollars from Banks in the global North accumulated massive amounts of foreign debt during the 1970s, with Poland's and Argentina's figures hovering at around 45% (Hung, 2015: 43-51). Facing a punishing repayment schedule after the Volcker interest rate shock in 1979, these governments found themselves forced to take emergency conditional loans from the World Bank and the International Monetary Fund (IMF). China, however, stood apart as a highly solvent and lucrative client for the Bank in a period of financial uncertainty. In early 1980, McNamara took personal charge to arrange and expedite China's membership while negotiating Taiwan's exit (Bottelier, 2018: 74;Husain, 1994: 28;Sharma, 2017: 158).
The twin crises of China's post-Cultural Revolution stagnation and that of the late-Keynesian international political economy, then, conditioned a new relation between the World Bank and the Chinese state. In May 1980, the World Bank approved relocating representation to Beijing and commenced a lending and technical assistance program with China's economic policymakers. But this was not to be a hubristic World Bank clashing with a rigid planning bureaucracy. Though some of China's Marxian economists were indeed skeptical of "western" economic knowledge, there was actually a high degree of affinity between party reformists, economists, and the World Bank (Bottelier, 2018: 79-81;Gewirtz, 2017: 134-135). The Bank's economists overseeing the China project were not market liberals, but structuralists like Hollis Chenery (who developed the Bank's budgeting calculus linking lending to increases in production; see Allan, 2019: 193), Shahid Husain (who oversaw the Bank's funding of ISI in Brazil; Bottelier, 2018: 73-74), Parvez Hasan (Pakistan's chief economist during the "development decade" of the 1960s), Edwin Lim (who worked on post-reunification Vietnam; Gewirtz, 2017: 111), and later, Shahid Javed Burki (a proponent of the adoption of McNamara's "basic needs" paradigm; Burki, 2002: 6-7). For these economists, fidelity to market principles was secondary to enhancing productivity within each national context, and it was vital, as then-Bank economist Adrian Wood (2019: 3-4) put it, "to guard against losing the strengths of the existing system" (see Gewirtz, 2017: 139-146). The target of productivity gains through both planning and market mechanisms articulated with the cautious approach taken by the party, which eventually adopted the Bank's prescriptions for competitive bidding on projects and joint ventures with foreign firms (Bottelier, 2018: 76;Husain, 1994: 29). After the well-received first report, entitled China: Socialist Economic Development, was published in 1983, the World Bank produced a report on China annually, and partnered through financing and assisting investments in agricultural modernization and rural credit extension, transport and energy infrastructure, public services, education, and sectoral reform. 7 But the accord between the Bank and the Chinese state would come under pressure from forces within both of them later in the decade.
Social unrest had been fomenting in cities since the Chinese state initiated an expansion of price and organizational reforms to the urban industrial sector in 1984. These were, indeed, counter-movements to processes of marketization but should be understood in relation to the particular conditions of production and social reproduction and their transformations in the early reform era (Hart, 2002: 199). In part an effort to build a reform-invested counterweight to the conservative old guard, the pro-reform central state devolved investment and fiscal authority to local governments, which responded through undertaking new accumulation strategies, like repurposing commune and brigade industries as profit-oriented township and village enterprises (TVEs) or transforming local state-owned urban enterprises into capitalist firms. With unencumbered access to socialist-era public investments and cheap labor from the decollectivizing agricultural sector, these enterprises earned super profits that resulted in an investment boom (Hart, 2002: 206-231;Hung, 2015: 54-61;Xu, 2018), causing "state" employment as a share of the labor force to increase dramatically through the early 1980s (So and Chu, 2016: 60-61). Simultaneously, however, the introduction of the contract labor system, management reforms, and the hardening of SOE's budget constraints resulted in the dismantling of Mao-era worker entitlements, including the guarantee of lifelong employment security and protection from management repression. Ambitious rounds of price liberalization stoked heavy inflation, peaking at over 28% in April 1989 (Weber, 2021: 253). The pressures of a declining social wage, the growth of a cadre-capitalist class, and perceived party indifference to drastically changing conditions of livelihood began to spark ideologically diverse student and worker protests in urban areas, reaching crisis proportions in the late 1980s (Wang, 2004). In response to these pressures, party conservatives such as Chen Yun, who had previously aligned with Deng and Zhao after Mao's death, perceived local overinvestment and inflation as evidence of waning "national strength" and loss of central control (Fewsmith, 2008: 39;Zhao, 1993). Even pro-reform party factions were divided on whether inflation should be solved by stabilizing aggregate demand (austerity) or by increasing aggregate supply (boosting production; Weber, 2021: 239-243). In response to the emerging crisis, Chinese and World Bank officials worked collaboratively on diagnosing the source of inflation, but reached contending conclusions.
While the Bank had advocated the dual-track model and enhanced worker protection during the first round of reforms, it changed course remarkably by mid-decade. McNamara's retirement portended a regime change within the Bank, with his successor, Tom Clausen, appointing Anne Krueger to the position of chief economist. 8 As former Bank economist and China program specialist Shahid Husain recollected, Krueger, who had spent her career theorizing state intervention as a form of rent-seeking, was "very hand-in-glove with the U.S. Treasury," introducing strong positions against import substitution and state-led industrialization efforts while getting "rid of anybody who had been associated with Hollis Chenery" (Husain, 1994: 31-32). For her part, Krueger directly scrutinized 1985's Long-Term Development Issues and Options report-which took a patient approach to enterprise privatization-but it was the more general problematization of price distortion that marked this period, in part through the influence of Eastern European economists working with the Bank (Lim, 1993: 16;Weber, 2020Weber, : 4-5, 2021. For the Bank, China's unrest was rooted in the CCP's "partial strategies" of price reform: rationalization of commodity prices prompted surges of local government investment in enterprises, but because SOEs were not incentivized to compete through improving productivity, deregulation without privatization resulted in inflation (World Bank, 1989: 27-29, 127). The State Council, on the other hand, viewed inflation as the effect of overheating TVEs and the export-oriented economy, and called for austerity measures (Hung, 2021(Hung, : 1008Weber, 2021: 240). The Bank, in contrast to the Chinese state, understood the crisis in terms of the incompleteness rather than the quality of urban market reforms.
Approaching China and the World Bank as connected yet distinct nodes in globally interconnected historical geographies brings into view how their alliance was forged in the global context of normalized state-led capitalist modernization facing exhaustion. The forces of intra-party pragmatic reformism and the Bank's structuralism formed the basis of a reform program under a shared vision of industrial restructuring, selective marketization, and institutional renovation. But as these reforms began to threaten social stability, conservatives in the Chinese state and ascendant neoliberals in the World Bank drew on distinct discursive resources-Marxist-Leninist appeals to plan-rationalism, and neo-classical reasoning corroborated by ongoing reforms in Eastern Europe, respectively-to articulate visions of readjustment that now conflicted with the hegemonic liberalreformist alliance. Evincing an emergent contradiction between the two bureaucracies, the tension would come to head after the Tiananmen Square protests (and the collapse of the Soviet Union) catalyzed a crisis in the Chinese state in 1989. This global conjunctural moment sparked an intraparty backlash against the pro-reform faction's ambitious liberalization measures, preceding China's turn toward more muscular forms of statism in the 1990s, but also reverberated in the World Bank. These struggles would transform both trajectories and redefine the Bank/China relationship.

Conjunctural struggles: The Tiananmen crisis and its reverberations in the World Bank
The immediate response of the World Bank to the Tiananmen Square demonstrations was to suspend $780 million in lending and pull 18 staff members from its offices in China, at the behest of the parliaments of its major G-7 donors (Figure 1). The Bank, however, pushed to resume lending as soon as possible, and President Barber Conable quickly subverted American Congress to approve a mission to China, resuming normal lending operations only weeks after the incident. 9 But Tiananmen catalyzed intra-party struggle between the reformist bloc-an alliance of provincial leaders, local government corporatists, and central revisionists who had held the upper hand within the party since Deng's ascension-and conservatives, who understood Deng's project as inciting a "peaceful evolution to capitalism" and betraying the revolution (Fewsmith, 2008: 39). Vindicated by the crisis and fears of repeating the pitfalls of the Soviet Union's perestroika, conservatives maneuvered for the recentralization of state authority over production and distribution. While neither faction resisted project-specific lending from global multilaterals, they mobilized criticism against incipient "bourgeois liberalization" and attacks against China's planning infrastructure. Mediating this intra-party struggle was ideological discord over the relations between modes of coordination (markets and planning) and modes of production (capitalism and socialism), with party conservatives equating the former and the latter, while reformists advanced a position that saw them in no necessary relation (Misra, 1998;Zhao, 1993). While the State Council initiated an austerity program to control inflation, it postponed further structural reforms to industry and finance. The conjunctural consequences of this interregnum would reshape China's trajectory, and its relation with the World Bank and the liberal order.
Hegemonic conflicts between reformists and conservatives were not contained within the party. They quickly reverberated through the World Bank, creating uncertainty over whether continued market reform or restoration of central planning would emerge as the party's post-crisis economic strategy. China's finance minister Wang Bingqian-Zhao's replacement-assured the China Office that reform would continue, but "with different emphasis on the market and the plan at different times," noting that state property rights had been "inappropriately reduced" through experimental ownership reforms. 10 In a memo to Headquarters, China Office Director Shahid Javed Burki interpreted this as evidence that a "new generation" of pro-reform party members were struggling with "the old guard," the outcomes of which were not yet certain. 11 He worried that line struggles and short-term austerity measures might delay necessary reforms to SOEs and the banking system. The first post-Tiananmen report, titled China: Between Plan and Market, conceded that reform needed to be kept "within the bounds of political tolerance," but it was clear they would not be solved until China recognized the "restraints provided by effective markets, hard budget constraints, and independent financial institutions" (World Bank, 1990: 5, 10). After China's Eighth Five Year Plan was published in March 1991, it affirmed support for state enterprises through measures like the Investment Orientation Tax (aimed at directing enterprises toward fixed investments), guaranteed credit provision, and exemptions from federal taxes. For the China Office, this confirmed the "renewed influence of China's elderly and more conservative leaders," casting doubts on China's ability to confront the "deep structural problem" of SOE overinvestment. 12 The neoliberal reform paradigm gaining adherents at the time dictated that the entrenched legacy of "central planning" called for "shock therapy." Buttressed by Washington's foreign policy, this approach was gaining traction as the most theoretically and financially appealing means of postsocialist reform, advocated for the macroeconomically unstable Soviet bloc by economists on the payrolls of the Bank and the IMF (Lipton and Sachs, 1990;Weber, 2021). For them, the low profitability of SOEs was the expected result of inefficient state industries insulated from price signals. "Stabilization" required realigning property with incentives through a rapid program of price and interest rate liberalization, with ensuing deficits financed by the sale of state assets, leaving no room for the state as a producer or distributor of goods and services. This approach was summarized in the 1991 World Development Report, which argued that privatizing SOEs was a "necessary and highly desirable" part of reform, and urged developing economies not to use SOEs "to pursue distributional goals at the expense of efficiency" (World Bank, 1991a: 144, 143). The Report made clear that if "trenchant political opposition" was not displaced, the second-best solution would be to prolong the timeline of reform: countries like China would transition to a market economy within 10 years (World Bank, 1991a: 145-147; see Figure 2).
Rather than understanding the transition problem in terms of reform or restoration, the China Office fashioned a position in close dialogue with the perceived direction of the party, recognizing that the urban industrial reforms and price liberalization measures of the 1980s had gone too far, too fast. The initiating memo for the 1991 country strategy paper (CSP) stated that the problem was not whether a given reform was pro-planning or pro-market, but whether it achieved "appropriate balances between central agencies, provincial and local governments, and economic units-related to this-between administrative measures and market in a very large and complex economy in transition". 13 The final draft put it in deliberately contrarian terms: China is not trying to create a western, market economy. It is not trying to eradicate the vestiges of the socialist system. Instead, it is trying to introduce reforms and new institutional mechanisms to modernize and increase the efficiency of the socialist system, which includes a very significant role for markets in resource allocation and for the private (or non-State) sector… talk of "shock therapy" for China makes little sense, for there is no clear idea of what the new institutions would be like, nor is it apparent what disease such therapy would be designed to cure (World Bank, 1991b: 7).
The CSP urged the World Bank headquarters to continue to lend at the levels agreed upon before Tiananmen, and support projects intended to expand China's industrial capacity, whether the industries were owned by the state or not. This endorsement of what Weber (2021: 146) calls China's "experimental gradualist" approach followed the China Office's decade-long engagement with the party. China's horizontal relation with the Bank meant that the China Office had been socialized into a pragmatic relation with Chinese policymakers, who tended to reject policies that rested on criteria of convergence with an idealized market equilibrium. Edwin Lim (1993: 14), who had authored the first two reports later reflected: "what [China] had to do, you could do under the socialist framework… there was no need to privatize the entire industrial sector. We didn't see that they had the conditions to do it." For parts of the Bank closer to (U.S.) power, however, this approach was unacceptable. The rationale was that, in a context where state ownership of industry still existed, industrial lending would prop up uncompetitive firms: shock therapy was designed to cure inefficiencies in "the market." Gerardo Sicat, an economist at the East Asia regional headquarters in Manila, wrote a review of the CSP arguing that China was the conservative laggard of the liberalizing world, meriting the withholding of Bank capital. 14 The CSP's reception was most acerbic at the Bank's research headquarters in Washington. There, Frank Lysy, an advisor at the Development Policy Group (DPG), cautioned chief economist Lawrence Summers on the China office: Burki is developing and pushing the position that, in contrast to the more common Bank view that comprehensive and relatively rapid policy reform is needed in centrally planned economies, China is correct in pursuing an approach of gradualist change. 15 Drawing on lessons from the Bank's ventures in Yugoslavia in the 1970s, Lysy asserted that lending should neither directly nor indirectly support SOEs since they were precisely the target of reform. Subsidizing them through loans to the energy sector or infrastructure in a marketizing environment would amount to an unmitigated "disaster," inevitably draining resources and increasing national debt. 16 Lysy projected that "it may take decades for problems of a similar nature to appear in China," but when they did, it would tie up over 10% of the Bank's lending portfolio and cost International Development Assistance (IDA, the Bank's concessional arm) billions in wasted funds, leaving the Bank unable "to support through increased lending a new Government which wished to follow a fundamental change in policies in a market-oriented direction." 17 The lines of struggle were thus drawn around different articulations of the materiality of lending and the meaning of reform. For the China Office, China's gradualist approach signaled that reforms may be reversed or upscaled in line with party command, and the country was therefore at little risk of default. Yet for the headquarters, any reforms that did not privatize SOEs indicated a lack of substantive reform, leading to a debt crisis and possibly political revolution.
Disputes over the role boundaries of the state and market peaked in the drafting of the 1992 CSP, which was to outline the China program for the rest of the decade. The China Office's first draft proposed two lending scenarios, a high option ($2.7B in commitments) and a base option ($1.9B in commitments), assessed based on China's "improvements" in building a general framework for industrial reform. While the draft allowed competition and corporate governance as possible criteria for "improvement," the DPG retorted that "the proposed lending program…would bring us to the ceiling on possible exposure" since ownership reform was not tied with lending. 18 Summers, despite sharing Lysy's skepticism over whether "evidence that such a framework, short of private ownership, can be identified much less implemented," negotiated a middle ground. 19 Recognizing the strategic role of China's portfolio, he requested the China Office to "specify precisely what the Bank should expect in order to… proceed with lending in the enterprise and finance sectors." 20 Following the CSP's final review meeting, a compromise between the Office and the DPG was reached: the Bank would lend at the base option, but this would be contingent on fostering "an environment in which the non-state sector enterprises would grow," and tying lending with ensuring SOEs "function according to the dictates of the market." 21 The Tiananmen conjuncture did not simply create a divide between a "statist" China and a "liberal" World Bank, but multiple struggles within the Chinese state that unevenly "refracted" (Hart, 2002: 300) through the Bank itself, bringing existing political and intellectual divisions between the headquarters and the China Office to the surface. In the party, conflicts between conservatives and reformers over inflation, social stability, and the meaning of socialism catalyzed struggles between the pro-shock therapy Washington headquarters and the pro-gradualist China Office. These material conflicts over lending and development policy were simultaneously conflicts over the boundaries and domain of the state within post-socialist societies. This conjuncture also led to the rearticulation of the World Bank/China relation, inflecting both of their subsequent trajectories. It is to this remaking that the paper now turns.

Rearticulating the Bank-China relation: Gradualism versus the Washington Consensus
If Tiananmen put the China Office's gradualist-sympathetic position into question, the return of the pro-reform faction to party leadership in early 1992 appeared to validate the Bank headquarters' pro-market predilections. With the symbolic affirmation of liberalization at the 14th Party Congress and Deng's "Southern Tour", apprehension over the continuity of marketization seemed to vanquish, and its partnership with the World Bank resumed: total commitments to China at the end of the 1992 fiscal year totaled $2.5 billion, mirroring the 1992 CSP's "high-case" scenario; and in 1993, the World Bank and Chinese policymakers convened at the Dalian Conference, the outcome of which was a sixteen-point program of far-reaching reform for the national economy, including the privatization of smaller enterprises, a recentralization of fiscal policy and introduction of local bond markets, and the creation of new policy banks to complement the commercialization of state-owned banks (Gewirtz, 2017: 254-258;So and Chu, 2016: 72). A high-profile publication commemorating the Bank's 50th anniversary admitted that the organization had "learned that where circumstances permit, there are merits in a gradual approach to economic change" (Thomas and Stephens, 1994: 18).
By 1994, however, it became clear to the China Office that the balance of political forces within the party was not favorable toward the rapid, full adoption of the Dalian program, but rather, would continue gradual SOE reform while concentrating central state capital in strategic firms and the four large state-owned banks. With this constraint, the Office shifted emphasis from direct industrial lending toward the "creation of an enabling environment for the reform and restructuring of state enterprises," in hopes that the tide would eventually turn toward privatization. 22 Yet the outcome was not what was anticipated. "Grasping the large and letting go of the small," China's strategy of privatizing smaller SOEs while maintaining ownership in large ones, aimed to transform state enterprises like China Mobile and SINOPEC into "national champions" modeled on Western corporations, shifting the policy environment in what has been termed a "neo-statist" direction during the mid-1990s (Huang, 2008;Hung, 2015: 61-69;McNally, 2019). While a round of rotations moved the Bank's long-time China staff members-with Lim moving to West Africa and Burki moving to Latin America (Burki, 2002: 22)-their replacements appeared to reflect the Headquarters' neoliberal strategy in East Asia under the "Washington Consensus," in alignment with the U.S. finance capital and the Treasury (Berger and Beeson, 1998).
After Tiananmen, then, the Bank/China relation was rearticulated from policy-based lending coupled with mutually informing exchanges of economic expertise and a shared vision of directional marketization, toward one where lending was marked by frustrated attempts at policy intervention at the macro-level. But it would not be accurate to claim that this rearticulation was simply an antagonism between market fundamentalism and Chinese statism (cf. Huang, 2010). While the Bank would continue to court China as a dutiful borrower (and, indeed, the largest for much of the 1990s to today) and accepted many of its microeconomic principles-the development of labor markets, the booming private sector, and welcoming of foreign industries-it became increasingly critical of the links between China's state property and its macroeconomic policies: its capital markets and balance of trade. At the forefront of subsequent evaluations of China's reform was a sense that "incomplete" or "partial" reforms of SOEs and the banking system left the Chinese (and international) economy vulnerable to collapse, buttressed by a historical imagination that understood these institutions as "the remnants of the planning system" (World Bank, 1992: 3, 52;1993: xii). For example, attempts by the World Bank to develop the China Investment Bank into a financial intermediary for industrial investments were frustrated due to their continued allocation through the planning system at below-market interest rates (World Bank, 1997). While China specialists continued to acknowledge the country's efforts in SOE reform through corporatization policies, those at headquarters-affiliated departments worried that the Bank was complacent with the consolidation of large SOEs, especially in the industrial Northeast, and their support through the statist financial sector. Complaints of "official foot-dragging" on SOE reform, with no concern for "the continuing weak profitability of the sector as a whole" 23 permeated the Bank's guidance for China during this era-especially in anticipation of its entry into the WTO-despite declining outlays through subsidized credit during the 1980s and 1990s (Jefferson and Rawski, 1994: 62;cf. Naughton, 2007).
As the Washington Consensus ran its course in the 1990s, it attempted to balance China's success (on its own terms) as a "gradualist" reformer as it continued to propagate the virtues of rapid reform and privatization elsewhere. Mirroring debates over the East Asian Miracle, 24 the 1999 World Development Report argued that …certain policies that helped Japan develop in the 1950s and 1960s, generated growth in East Asia in the 1970s and 1980s, and sparked China's economic boom in the 1980s and 1990s were specific to time and place. They may not have worked well in other countries, nor are they likely to be appropriate in the opening decades of the 21 st century (World Bank, 1999: 2).
These competing meanings of reform were undergirded by material divergence between the Bank and China during this era. In the 1990s, a bloc led by the United States pressured the World Bank to cease providing grants to China through IDA, in an attempt to push the country into deeper reliance on international capital markets, which, in the Bank's own words, "placed strains" on its relationship with China (World Bank, 2005: 10;Xu, 2017). Meanwhile, regular lending was constrained by rules limiting the share of its portfolio dedicated to a single country, leading to a $2.3 billion decrease from 1993 to 2001. In this setting, the Bank's calls for macroeconomic reform increasingly fell short of the Chinese state's expectations, offering little by way of precision or implementation strategies. 25 The view of the Washington Consensus-era China Office toward the country's direction can best be summarized by representative Pieter Bottelier (1993Bottelier ( -1997, potential conflict [between state and market] has already become a reality, which explains the seemingly schizophrenic nature of China's current half-reformed economic system. The state and the market cannot both have a leading role in the economy at the same time (Bottelier, 2018: 87).
Yet, Beijing's relation with the Bank is not that of a new Cold War, even if the Bank faces continuous pressure from the United States to cease lending to China (Chin and Gallagher, 2019). The country continues to borrow and donate to the World Bank Group: from 2013 to 2019, the Bank committed, on average, $1.78 billion per year to the country, while China itself was the sixth largest donor to IDA in 2019. 26 Moreover, as the Chinese state has developed its own aid capabilities, the World Bank has, in turn, strategically readjusted by accepting new roles for direct state ownership and management in areas like infrastructure and finance, even as it remains wary of their effects on competition and efficiency . While current competitive conditions in development finance and foreign investment may, then, warrant distinctions between the Bank's and China's approaches, a Hartian frame suggests that the market-state continuum is an inappropriate metric for characterizing them. It further suggests that, as hegemonic social forces, they are articulated unities: not given once and for all, but are "contradictory, and must be continually renovated, renewed, and re-enacted" (Hart, 2002: 28).

Conclusion
This paper has made the case for a different theoretical and methodological approach for geographical studies of the "new state capitalism," an emerging problematic that seeks to understand "the aggregate expansion of the state's role as promoter, supervisor, and owner of capital across the spaces of the world capitalist economy" (Alami and Dixon, 2021: 7). It argued that Gillian Hart's work provides a set of orientations, tools, and concepts for geographers interrogating the basis for this aggregate expansion through tracing interconnected trajectories, political struggles, and hegemonic projects within and across state boundaries. To illustrate this, the paper outlined three theoretical and methodological moves-multiple trajectories, conjunctural analysis, and articulation-and used them to illuminate the politics of economic development strategy at the World Bank-China nexus, focusing on the remaking of their trajectories and relations through the Tiananmen Square crisis. While the trajectories of late-Maoist China and the late-Keynesian World Bank set the terrain for exchanges of finance and expertise between the two organizations, their initial shared vision-cum-program of liberalization faltered as inflation and social crises unfolded in China. In the wake of Tiananmen, conjunctural struggles within and between the World Bank and the Chinese state remade their trajectories and relations. While the reformistconservative compact ended in a strategy of redoubled state ownership of leading SOEs and banks with partial privatization, the neoliberalizing Bank was unsatisfied with anything less than full privatization. The result was a new articulation of state/market norms within both organizations and new relations between them, with China's "gradualist model" posed against the China Office's alignment with the mainstream Washington Consensus approach. By putting China's gradualist and the World Bank's neoliberal consolidation in the 1989 conjuncture into the same analytical frame, the paper has offered one window into how these projects-what are often termed "state" and "liberal" capitalist models-have been produced with each other.
In doing so, the paper provided less of a statement on state and liberal capitalism as discrete economic systems than it did an analysis of how forms of intervention, institutional structures, and development strategies are actively produced through practices that combine state and market norms, rationales, and logics across multiple, asymmetric nodes in the global political economy. It suggests that geographers can intervene by analyzing how state-capitalist trajectories are (re) made-indeed, with far-reaching, global consequences-through interconnected arenas of everyday life, with the state and the market as qualitative, relational phenomena to be assessed even in the most "liberal" and "statist" contexts. This means there is a variety of entry points through which the "new state capitalism" can be problematized, not in the least through continued attention to the interconnected struggles and negotiations of development projects, trade agreements, intellectual property governance, and international regulatory standards, and how state maneuvers are related to contestation from below. Attending to the struggles that constitute these "varieties" might, then, help to denaturalize the Hobson's choice between two capitalist models differentiated on the basis of a greater role for party-state dictat versus that of a self-devouring financial sector. If the institutionalized relations which these models name have been actively forged, it might be within the scope of political action to rearticulate alliances, projects, and strategies and transform them.