Lost and Found: Bourgeois Dependency Theory and the Forgotten Roots of Neodevelopmentalism

Neodevelopmentalism emerged in Brazil and Argentina in the aftermath of the demoralization of the Washington Consensus. Although its intellectual proponents place it within the long tradition of Latin American developmentalism, an important theoretical origin of neodevelopmentalism—dependency theory—has so far been ignored. The term appeared for the first time in 1978 as an expletive in the heated controversy between Ruy Mauro Marini and Fernando Henrique Cardoso and José Serra in the Revista Mexicana de Sociología. Breaking with the supposition that underdevelopment could be overcome only through social revolution, Cardoso and Serra embraced a perspective of long-term social transformation based on class alliances with fractions of the national bourgeoisie and international capital. This perspective was gradually weakened and finally abandoned in favor of full-fledged neoliberalism when Cardoso became president of Brazil in 1994, only to be resuscitated by so-called pink-tide administrations after 2002. O neodesenvolvimentismo surgiu no Brasil e na Argentina após a desmoralização do Consenso de Washington. Embora seus proponentes intelectuais o coloquem dentro da longa tradição do desenvolvimentismo latino-americano, uma importante origem teórica do neodesenvolvimentismo - a teoria da dependência - até agora foi ignorado. O termo apareceu pela primeira vez em 1978 como um palavrão na polêmica acalorada entre Ruy Mauro Marini e Fernando Henrique Cardoso e José Serra na Revista Mexicana de Sociología. Rompendo com a suposição de que o subdesenvolvimento só poderia ser superado por meio da revolução social, Cardoso e Serra abraçaram uma perspectiva de transformação social de longo prazo baseada em alianças de classe com frações da burguesia nacional e do capital internacional. Essa perspectiva foi gradualmente enfraquecida e finalmente abandonada em favor do neoliberalismo completo quando Cardoso se tornou presidente do Brasil em 1994, apenas para ser ressuscitada por administrações da chamada maré rosa após 2002.

The recent rise of the extreme right in Latin America is prompting some writers to cast a benign eye on the neodevelopmentalist experiments undergone in Brazil and Argentina between 2003 and 2015. By siding with Lula in Brazil and Kirchner in Argentina, part of the left has tacitly endorsed the opposition between neoliberalism and neodevelopmentalism, falling into a discursive trap. Indeed, at least since the demise of the revolutionary-Marxist stream of dependency theory in the late 1970s, various versions of liberalism and developmentalism have shared a duopoly of legitimate development discourses in Brazil and Argentina. Based on a reified distinction between states and markets and on unfulfilled promises of catch-up capitalist development, the sharp division between (neo)liberalism and (neo)developmentalism leaves no room for oppressed social groups to assume protagonism in defining their own development priorities. If this protagonism is to be recovered, it is time for Latin America's radical tradition of anticapitalist development thought to be reclaimed.
Neodevelopmentalism is not simply the most recent reincarnation of the Latin American classical developmentalism whose intellectual foundations lie in the heterodox structuralism of Raúl Prebisch and Celso Furtado. 1 One of the most important roots of neodevelopmentalism has so far been ignored: Fernando Henrique Cardoso's reformist dependency theory. Indeed, the term "neodevelopmentalism" appeared for the first time in 1978 as an expletive in the heated controversy between Ruy Mauro Marini (1978) and Cardoso and Serra (1978) in the pages of the Revista Mexicana de Sociologia (for a review see Hadler, 2013). This controversy was the culmination of a long debate about the capitalist development possibilities of Latin America and marked the final split of dependency theory into diametrically opposed revolutionary-Marxist and bourgeois-Weberian schools.
As presciently noted by Marini, the reformist stream of dependency theory represented by Cardoso and Serra came to the rescue of the discredited developmentalist thesis at a critical moment. The exhaustion of the state-led phase of import substitution in the 1960s 2 and the subsequent rise of murderous capitalist dictatorships backed by the United States (Gill, 2004) seemed to dispel any illusions about the potential of progressive, anti-imperialist alliances between the working class and the national bourgeoisies. In the 1970s, the sociological backbone of developmentalism was broken. When even modest social mobility seemed to threaten the privileges of national bourgeoisies, they quickly stepped back from progressive development perspectives and revealed their deeper, historical connection to both international capital and the traditional oligarchies (Bambirra, 1978: 104). The crisis of classical developmentalism confirmed Frank's (1969) disjunctive "underdevelopment or revolution." Acutely spelling out the only real options for Latin American peripheral socie ties, Frank and the first generation of Marxist dependency theories emptied the core promise of developmentalism, the perspective of capitalist catch-up (see Antunes de Oliveira, 2017;Bambirra, 1974;1978;Dos Santos, 1970;1977;1978).
In a nutshell, Latin American ruling classes proved incapable of severing their umbilical link to archaic structures of political domination based on a subordinated integration into the international market. Therefore, they could not lead broad social alliances with organized labor aiming at an autonomous process of capitalist catch-up development, as expected by the developmentalist school. This situation of bourgeois insufficiency, previously identified by Marx in relation to other internationally subordinated ruling classes and acutely captured in Trotsky's theory of permanent revolution (Löwy, 2010), returned historical agency to the working class of peripheral countries, the key social actor with real interest in breaking these relations of subordination. The culmination of the Marxist stream of dependency theory was this clear realization, captured with precision in Marini's "Dialéctica de la dependencia" (2009 [1973]).
At that crucial juncture, there was consensus on the left that the only hope for Latin America to break the chains of capitalist underdevelopment was profound social and economic change led by the historically subordinated classes. Nevertheless, a split in the dependency literature allowed for a renewed faith in what Benjamin Selwyn (2016) has recently labeled elite-led capitalist development. Faletto's Dependencia y desarrollo en América Latina (1979 [1967]) departed from the radical conclusions of other dependentistas, proposing the concept of "Dialéctica de la dependencia," which reconciled the possibility of capitalist development with the condition of dependency. This argument would be further elaborated by Cardoso in a series of texts published in the 1970s. Conflating economic growth with development and surreptitiously adhering to a stageist view of development that assumed the United States and Europe as models to be followed by the rest of the world, Cardoso took aim at the Marxist stream of dependency theory. Implying that it precluded any form of accumulation in Latin America, Cardoso built a straw man of Frank's and particularly Marini's ideas. The final stroke came in 1978 with a virulent critique of Marini's "Dialéctica de la dependencia" promptly answered by Marini himself.
The link between Cardoso's anti-Marxist dependency theory and contemporary expressions of neodevelopmentalism may appear elusive, and understandably so. After his eight years as president of Brazil, Cardoso is rightly identified with neoliberalism, the declared nemesis of neodevelopmentalism. Most of the Brazilian and Argentine contemporary neodevelopmentalist writers quote Cardoso only as a negative example of neoliberal, Washington Consensus-style policies. The 1978 controversy has been largely forgotten. Calling Cardoso's ideas "neodevelopmentalist" now sounds almost like a compliment. Nevertheless, the missing roots of neodevelopmentalism in Cardoso's bourgeois dependency theory can be traced by revisiting four defining moments in the Latin American debates about development.
The first moment is the already-mentioned splitting of paths in the dependency literature paradigmatically located in the controversy between Cardoso and Serra and Marini in 1978. The second is the Brazilian elections of 1994, which Cardoso won with a political platform announcing "a new development strategy" largely in line with his previous ideas. The third spin in the long transformation of Cardoso's antirevolutionary dependency theory into neodevelopmentalism happened during the collapse of the first phase of the Real Plan, in the aftermath of the 1998 crisis. Then, Cardoso abandoned the last socially progressive aspects of his old ideas to embrace full-fledged neoliberalism.
At that point, old collaborators of Cardoso such as Luiz Carlos Bresser-Pereira and Guido Mantega 3 took over the unfulfilled promises of Cardoso's 1994 program. With cosmetic changes to mark a difference vis-à-vis the Washington Consensus, Cardoso's "new development program" was rebranded as "new developmentalism." The final and defining moment in the rise of contemporary neodevelopmentalism was its adoption by the left. After the decentralized popular uprising in Argentina in December 2001 and the transformations undergone by the Brazilian Partido dos Trabalhadores (Workers' Party-PT) during the 2002 elections, neodevelopmentalism reappeared as a coherent, viable, and socially progressive alternative to neoliberalism, its old roots now totally erased.
Looking back, it is apparent that the consequences for the Brazilian and Argentine left of losing sight of a radical perspective on development were overwhelming. It allowed for the emergence of a rigid duopoly of development discourse centered on the false dichotomy between states and markets. The political economy debate in both countries is dominated by neoliberals arguing for continued market liberalization and neodevelopmentalists arguing for more state intervention. Not even the major upheaval represented by the recent election of Captain Bolsonaro in Brazil has changed this basic feature. Indeed, Bolsonaro may mark the ascent of peripheral fascism in Latin America (Antunes de Oliveira, 2019;Selwyn, 2018). Nevertheless, in terms of political economy, his program is fully neoliberal, in some characteristics resembling "Pinochetism." 4 In interview after interview, Paulo Guedes, a Chicago-trained economist and Bolsonaro's economic strongman, praised "markets" over the "state" and pledged to privatize, deregulate, and liberalize the Brazilian economy. 5 With the mainstream political and economic scene divided between neodevelopmentalists and neoliberals, the left finds itself between the devil and the deep blue sea. Marxist perspectives that directly challenge social privileges and denounce superexploitative labor relations and staggering wealth inequalities are dismissed as dated and unrealistic. Neodevelopmentalism often appears as the lesser of evils, particularly in the face of the sweeping neoliberal reforms implemented by Presidents Macri and Bolsonaro. Revisiting the four key moments in the genesis of neodevelopmentalism makes it possible to re-link what now appears as a moderate progressive program to an inherently elitist and reactionary theory crafted in explicit opposition to an alternative perspective of radical social change.

1978: The PrehisTory of NeodeveloPmeNTalism
Although the concept of neodevelopmentalism has been consistently incorporated into academic debates only in the past decade (see, for instance, Boito and Berringer, 2014;Katz, 2014;Morais and Saad-Filho, 2011;Webber, 2017), in reference to the rise of left-of-center administrations in Latin America (particularly in Brazil and Argentina), it was coined by the Marxist dependency theorist Ruy Mauro Marini in 1978. In his article "Las razones del neodesarrollismo" Marini answered the criticism leveled at his revolutionary dependency theory by José Serra and Fernando Henrique Cardoso and accused his critics of "neodevelopmentalism" (Marini, 1978: 102-103). What appeared in the late 1970s as a grave accusation of capitulation in the face of international capital and national ruling classes now sounds almost like praise for Cardoso and Serra, considering how much lower they would sink during the 1990s. 6 Dependency theory appeared in Latin America in the early 1960s as a response to both modernization theory and classical Latin American developmentalism. Exiled in Chile after the military coup in Brazil in 1964, a group of left-wing intellectuals including Vânia Bambirra, Theotônio Dos Santos, André Gunder Frank, Ruy Mauro Marini, José Serra, and Fernando Henrique Cardoso challenged Raúl Prebisch's state-centered theories and structural developmentalism at large. Although there was considerable diversity among this first generation of dependentistas, all of them recognized that the structural impediments to development in Latin America could not simply be corrected by virtuous state-led policies premised on a progressive class alliance between capital and labor. As Frank (1966: 21) famously put it, the "present underdevelopment of Latin America is the result of its centuries-long participation in the process of world capitalist development," and therefore, within global capitalism, Latin America could only experience "development of underdevelopment." Cardoso and Faletto's Dependencia y desarrollo en América Latina was from the start one of the most nuanced and elaborate versions of dependency theory. The authors placed great relevance on historical specificity and focused particularly on the role of the emerging middle classes, which appeared in Latin America in the first three decades of the twentieth century following the initial diversification of merely export-based economies. These new middle classes played a key role in the constitution of modern national economies. Contradicting the idea that capitalist development in peripheral countries was impossible, Cardoso and Faletto claimed that state-led industrialization based on the domestic market formed by this new middle class actually succeeded in reshaping the old colonial economic structures. This growing internal market allowed the bigger countries of Latin America-Brazil, Argentina, and Mexico-to overcome the condition of simply commodities-export economies to become important industrial clusters. Although this process generated some form of development, dependency was renewed in that the dynamics of industrial investment in Latin America were dictated by multinational companies whose decisions were made at their headquarters in the central countries.
Building on his previous work, Cardoso (1972: 89) explicitly rejected Frank's model for its incapacity to explain the capitalist development that actually happened in Latin America after the diversification of export-based economies: "The idea that there occurs a kind of development of underdevelopment, apart from the play on words, is not helpful." Given that dependency and development were not intrinsically incompatible, he argued, it would be more accurate to conceptualize the transformation of the economic and social structures in Latin America as a form of "dependent capitalist development." Cardoso's concept of dependent capitalist development stood on the brink of returning to the classical developmentalist perspective that capitalist development could be achieved by Latin American countries if only they had a suitable development strategy based on a progressive class alliance. The stageist view of development against which dependency theory arose in the first place is reflected here. As noted by Ramón Grosfoguel (2000: 366) "the propositions that an autonomous decision-making process at the nation-state level is possible for every country to achieve, that dependency is mainly an internal relation of forces in favor of foreign actors, and that underdevelopment is a backward stage of the productive system leads Cardoso to developmentalist premises." At that point, only the "dependent" character of the peripheral development process, a consequence of the association with foreign capital due to the structural lack of capital in Latin America, differentiated Cardoso's ideas from the developmentalist catch-up conceptions.
As he came closer to the original developmentalist thesis, the differences between Cardoso's reformist dependency theory and the stronger version advocated by Dos Santos, Frank, Bambirra, and Marini would become increasingly clear. Initially, Cardoso's dependent capitalist development appeared as an attenuated dependency thesis, perhaps more sophisticated than but not incompatible with the straightforward argument popularized by Frank. After the publication of Marini's "Dialéctica de la dependencia" (2009 [1973]), however, the abyss between the two forms of dependency theory would widen. In his most important book, Marini would substantiate the Marxist dependency thesis of Frank, Dos Santos, and Bambirra with a rigorous economic argument based on Marx's labor theory of value. For Marini (2009Marini ( [1973: 115), the "secret of the uneven exchange" that kept Latin America underdeveloped was that its incorporation into the world capitalist economy corresponded to a key transition in the way central capitalist economies operated. With the incorporation of the periphery into the world capitalist system, the supply of cheap consumption goods would make possible the shift in the dominant mode of exploitation of the working class in central countries from the extraction of absolute surplus value to the extraction of relative surplus value. In other words, cheap consumption goods produced in the periphery would allow for a decrease in the relative value of labor in the center, increasing its productivity. The other side of the coin was that surplus value in the periphery would be extracted mainly in its absolute form, often involving the superexploitation of labor-the payment of extremely low wages, even below the social cost of labor reproduction. The most visible effect of these dialectic dynamics of dependency was the appearance of large consumption markets in the center coupled with the permanent depression of consumption markets in the periphery (Marini, 2009(Marini, [1973: 124-135).
In the face of Marini's argument, Cardoso's notion that the recent industrialization in the largest Latin American countries potentially represented a break with underdevelopment was seriously undermined. Indeed, industrialization appeared as simply a new twist in the spiral of development of underdevelopment. This was the context of the attack launched by Cardoso and his life-long collaborator José Serra against Marini. For Cardoso and Serra (1978), Marini's theory was too rigid, implying the economic stagnation of peripheral countries. According to them, his economic argument was empirically wrong in ignoring the gains in productivity achieved in Latin America since the end of the nineteenth century. Most important, they accused him of economic determinism, leaving no room for politics to change the structures of dependency. In their words, "The economic reductionism that underpins the impossibility of capitalist expansion in the periphery within structural limits like those highlighted by Marini is not only false, as we have seen, but also kills the nerve of the political analysis, making it dependent on a catastrophism that is never realized" (Cardoso and Serra, 1978: 27).
It is not difficult to see that this critique purposefully misrepresented Marini's ideas to reconstruct the prospect of capitalist catch-up development and justify a reformist political strategy premised on the possibility of progressive class alliances. Marini's immediate answer restated points that were clear in "Dialéctica de la dependencia" but had been ignored by his critics. First of all, his argument did not preclude growth, technical change, or even an increase in productivity in the periphery (Marini, 1978: 58). It simply explained why and how, despite those changes, underdevelopment reproduced itself. His point was that the incorporation of technology and increase in productivity were compatible with underdevelopment as long as the periphery's role as producer of cheap consumer goods for the center did not change and wages in the periphery remained depressed. The recent appearance of highly productive maquiladoras in Mexico shows that Marini's argument remains relevant. Finally, the critique that Marini's dependency theory leaves no room for politics could not be more misleading. Marini's argument-just like Frank's-is, in the end, a call for social revolution. Is there any better example of faith in the political agency of the subordinated classes?
Cardoso and Serra's argument represented Marini's ideas as "economic reductionist" by ignoring their political content. Taking for granted the impossibility of radical political action, it was they who, despite prizing politics over economics (a dichotomy that did not exist in Marini's work), effectively limited the potentiality of politics. 7 The contradictory point made by Cardoso and Serra-stressing the creative powers of politics but discarding revolutionary change-would only make sense retrospectively as self-justification. Indeed, within a couple of years, both would become mainstream politicians, starting successful careers in the aftermath of the political opening in Brazil in 1979.
With the perspective of self-emancipation of the working class excluded from the agenda, the old promise of developmentalism had to be restored as the natural goal of "backward" peripheral societies. Thus, the old and discredited promise of catch-up capitalist development became once again the holy grail for Latin American countries. Following Cardoso's creed, the path to it included renewed alliances with putatively progressive fractions of the ruling class in association with international capital.
Although largely ignored at the time, the neodevelopmentalist seeds already present in Cardoso's bourgeois dependency theory were spotted and vividly denounced by Marini. Making the first recorded academic use of the term "neodevelopmentalism," Marini (1978: 102-103) strikingly foresaw the future degeneration of Cardoso's ideas: "The new ideologists of the Brazilian bourgeoisie [Serra and Cardoso] find themselves obliged to reinstate this tradition [developmentalism] and try to give credibility to an American-or Europeanstyle Brazilian capitalist development. In a nutshell, we are facing a neodevelopmentalism still ashamed of itself, but soon to lose its inhibitions." As was later noted by Perry Anderson (2016), "in pursuit of office Cardoso had sacrificed not only his early convictions, which were Marxist and socialist, but over time his intellectual standards." The "sacrifice" of his convictions corresponds exactly to the loss of Cardoso's neodevelopmentalist "inhibitions" foreseen by Marini. What appeared in 1978 as a mainly negative thesis aiming basically at disputing the intellectual space that Marxist dependency theory had conquered in the 1960s and 1970s would fully evolve into a bourgeois development program in the subsequent years. In spite of or perhaps exactly because of their theoretical shortcomings, Cardoso's ideas easily completed their metamorphosis into contemporary forms of developmentalism. Based on the subordination of the working classes' priorities-such as increase in wages, wealth distribution, and socialization of the means of production-to the priorities of national elites in association with international capital, the neodevelopmentalist program would make its political debut in the contested 1994 Brazilian elections.

1994: The PoliTical BirTh of NeodeveloPmeNTalism
Contrary to what the growing literature on neodevelopmentalism suggests, the first appearance of neodevelopmentalist political discourse was not in Lula's 2002 presidential campaign or in Néstor Kirchner's dispute with Carlos Menem in 2003. The key features of the neodelopmentalist model-faith in the developmental powers of the state and the renewal of the grand promises of capitalist catch-up development typical of the old developmentalist cycle of the 1950s and 1960s-were articulated by Cardoso in the Brazilian 1994 presidential elections. Looking directly at the camera in a television spot broadcast to millions of Brazilians, Cardoso (1994) abandoned the nuances of his previous writings and fully embraced the promise of catch-up development: Great nations of the world are not built by better people than ours. They are built by people like me and you. It so happens that in those places, the governments knew how to work side by side with their people and built fairer societies.
And this is what we are going to do in the next four years. Brazil can be one of the great powers in the world. This is no exaggeration. This is an idea as big as Brazil itself.
Much can be read between these lines. They are not simply a call to national greatness similar to those seen during the military dictatorship and, more recently, in the presidential campaign of Jair Bolsonaro. In fact, they translate into straightforward language ideas already present in Cardoso's argument with Marini 16 years earlier, representing the political culmination of his bourgeois dependency theory. Cardoso is calling for a cohesive society led by a competent and well-intentioned government. The union of the "people" and the "government" can potentially make Brazil "one of the great powers of the world." The "Brazil" to which Cardoso addresses his speech is not fundamentally divided into social classes with conflicting interests. Furthermore, it does not face fundamental international constraints to becoming a "great power." Cardoso's discourse identifies no enemies, no social conflicts. The only challenge facing Brazil is getting the government and the people to work "side by side." Here the line between Cardoso's bourgeois dependency theory and the developmentalist tradition finally melts into thin air. The limits he previously identified to associated or "dependent" development are transformed into opportunities for association with international capital. In a text published just after his being elected, Cardoso (1995: 151; see also Cardoso and Font, 2001) revisited his old ideas and denied that peripheral countries are doomed to develop in "distorted" ways: Today we know that it is not true. Countries that were able to manage their economies sensibly to the transformation of modes of production within capitalism, as well as to social issues, have had more favorable trajectories than others. The case of the Asian Tigers is well-known. What remained of "determinism" in dependency theory, maybe a Marxist trait-and I always criticized determinism-certainly must be fundamentally reformulated. The level of influence of political choice over the economic structure is greater than it appeared to us in the 1960s.
By abandoning the last "Marxist traits" of his former dependency theory, Cardoso came back full circle to the developmentalism in opposition to which dependency theory had arisen in the first place. Nevertheless, the conjuncture of the 1990s was substantially different from that of the 1950s and 1960s. The discursive reframing of developmentalist ideas in that new scenario, defined by what Cardoso increasingly called "globalization" (Batista Jr., 1999), marks a crucial jump from classical developmentalism to neodevelopmentalism through the hoops of his dependency theory. The example of the Asian Tigers, which "managed their economies sensibly" and took advantage of the changes in international capitalism to "develop," would become a constant in neodevelopmentalist discourse. The questionable democratic records and the historical specificities of countries such as South Korea are normally omitted (see Kvangraven, 2021). Tellingly, the same argument that appeared in the controversy against Marini-the priority of "political choice" over "economic structure"-was used again here. "Political choice," of course, does not mean the autonomous decision making of the subordinated classes. It appears here as a surreptitious justification of the alliances Cardoso made with representatives of big capital and traditional oligarchies in order to come to power. "A New Development Project" is the title of the first chapter of President Cardoso's first administrative program, the most detailed document of its kind produced in Brazil since the end of the civil-military dictatorship (Cardoso, 2008(Cardoso, [1994). This crucial text, totally ignored by both contemporary neodevelopmentalist writers and critics of neodevelopmentalism, is the real birth certificate of neodevelopmentalism. It marks the political consolidation of Cardoso's early ideas, translates them into concrete policies, and spells out a development strategy that aims explicitly at updating Latin American traditional developmentalism.
The core message of the program was clear. The traditional Brazilian development strategy had to be reframed. Accordingly, classical themes of the developmentalist tradition (protection of infant industry, import substitutive industrialization) were replaced by a more positive view of the potentialities of international trade. The implementation of the "new development project" involved redesigning the role of the state, creating investment partnerships with the private sector, and promoting the attraction of foreign investment. The fundamental importance of international capital for financing development-a direct insight from Cardoso's dependent capitalist development-is fully incorporated: "For Brazil to take the best possible advantage of the external conditions of resources supply . . . it is necessary to guarantee an attractive and trustworthy environment for foreign investors" (Cardoso, 2008(Cardoso, [1994: 9).
Notwithstanding eventual updates clearly influenced by the context of the 1990s, Cardoso's first administrative program maintained the hallmark of developmentalism-faith in the state's capacity to transcend class struggle and direct national development to the benefit of all. Indeed, Cardoso (2008Cardoso ( [1994: 127) keenly insisted that his "new development" program was not to be confused with neoliberalism or the Washington Consensus. Instead, it was an adaptation of the quintessential developmentalist goal-capitalist catch-upto new international circumstances. The human face of his program was revealed through compensatory social policies aiming at correcting historical distortions and neutralizing potentially negative side effects of capitalism. As summarized by Cardoso (2008Cardoso ( [1994: 4), "It is necessary to define and implement a new model of development that fights extreme poverty, improves income distribution, guarantees for the Brazilian economy an intelligent integration into the world, and reorganizes the state." This early expression of neodevelopmentalism puzzled contemporary critics amid the political disputes of 1994. Many on the left, siding with the more radical perspective of social change then represented by Lula da Silva, accused Cardoso of abandoning his old ideas. José Luis Fiori (1994), one of the few to remember the old dependency-theory debates, was the first analyst to unveil the continuity between Cardoso's old and new ideas, but he missed the point by supposing a direct line between Cardoso's reformist dependency theory and the Washington Consensus. A close reading of Cardoso's first administrative plan and an analysis of his political discourses do not show any evidence of his intellectual adherence to neoliberalism. Cardoso never praised the putative efficiency of free markets or advocated the retreat of the state-as President Menem (1998) was doing at the time in Argentina. Instead, his "new development" project was largely based on an active state, which was supposed to lead a broad social alliance for capitalist development.

1998: The "New develoPmeNT Program" versus The washiNgToN coNseNsus
While, at a discursive and ideological level, Cardoso cannot be accurately classified as neoliberal, the macroeconomic policies his administration adopted, particularly after the 1998-1989 crisis, in fact came close to the Washington Consensus model. In other words, Cardoso's administration never lived up to the neodevelopmentalist promises contained in his 1994 administrative program, which was gradually abandoned in favor of full-fledged neoliberalism.
Before the decisive neoliberal turn marked by the appointment of Armínio Fraga as president of the Central Bank and the adoption of the macroeconomic tripod in 1999, 8 a political cleavage appeared within Cardoso's administration: on one side a neodevelopmentalist group of old collaborators committed to Cardoso's moderately reformist "new development" program (including Serra and Bresser-Pereira) and on the other side a neoliberal group of orthodox economists from the Pontifícia Universidade Católica do Rio de Janeiro, best represented by Gustavo Franco and Pedro Malan. The second group made no secret of its disdain for the developmentalist tradition, never having had any affinity with the neodevelopmentalist program. 9 Above all, committed to curbing inflation, the neoliberal faction was willing to follow an orthodox approach to macroeconomic management, in line with International Monetary Fund recommendations. From the start, the neoliberal faction occupied key economic positions in the Ministry of Finance (Pedro Malan) and the Central Bank (Gustavo Franco), while the neodevelopmentalist faction gained the Planning Ministry (José Serra) and the State Reform Ministry (Luiz Carlos Bresser-Pereira), as well as influence over the social area. The monetary stabilization plan that catapulted Cardoso to power in 1994-known as Real Plan-combined structuralist elements derived from the Latin American school of heterodox economics with a strong austerity package aligned with economic orthodoxy.
During the four years of Cardoso's first presidency, the ascendancy of the neoliberal group and the adoption of Washington Consensus-style reforms were discursively justified by the imperative of "tightening up" the house in the aftermath of the debt crisis and controlling inflation. The socially progressive aspects of the 1994 program were indefinitely postponed but never openly abandoned. The balance-of-payments crisis of 1998 marked the end of the tense coexistence between the neoliberal and the neodevelopmentalist faction in Cardoso's administration. To obtain access to up to US$41 billion in international credits under International Monetary Fund supervision, the government agreed to pursue ambitious fiscal surpluses and promote sweeping structural reforms. The neoliberal faction was decisively empowered, while the neodevelopmentalist faction either left the government or was cornered in the social ministries. In 1999 Armínio Fraga, a self-declared follower of the Washington Consensus (Fraga, 2004), joined the government as an allpowerful and virtually independent president of the Central Bank, while Serra was moved to the Health Ministry and Bresser-Pereira left the government for a visiting position at Oxford.
The underlying reason for the failure of Cardoso's administration to implement the neodevelopmentalist program as proposed in 1994 was the inherent limitations of the political alliance it managed to form. Since Cardoso's "new development" program sought to combine outward-oriented economic growth with gradual gains for the working class, it depended on the government's ability to strike a balance between the contradictory demands of labor and national and international capital. Nevertheless, from the start of his administration Cardoso governed with fierce opposition from the PT whose leaders in Congress went so far as to demand his impeachment amidst the 1999 crisis.
As a consequence, instead of a class alliance between capital and labor under a neodevelopmentalist compromise, the political base of Cardoso's government was overwhelmingly dependent on the support of international capital, big national capital, and co-opted traditional oligarchies represented by the president of the Senate, Antonio Carlos Magalhães, and Cardoso's vice president, Marco Maciel. The political conditions for wage increases or potentially disruptive social reforms were absent, and, as a result, apart from the implementation of pioneering but very limited social programs, the policies effectively adopted during Cardoso's administration did not consistently incorporate gains for the working class, falling far short of the promises of the 1994 program.
Given the nature of the social alliance he led and the macroeconomic policies he adopted, it is not surprising that most critical writers do not hesitate to identify Cardoso's administration with neoliberalism. At that point, Theotônio dos Santos was one of the few intellectuals to recognize that, at least at the theoretical and discursive level, Cardoso had not renounced his ideas. In an illuminating article published in Latin American Perspectives, he challenged his old dependentista colleague to prove his old theoretical claims: "He [Cardoso] will prove his thesis only if he is able to advance social reforms-postponed interminably-in his moderate-conservative government" (Dos Santos, 1998: 67). Cardoso's undeniable failure to bridge social inequality and bring about catch-up capitalist development should, conversely, prove his neodevelopmentalist perspective wrong.
The Latin American developmentalist dream of capitalist catch-up died for the first time in the 1960s under the heavy boots of putschist generals. Cardoso revived it only to let it die again in the capable hands of Washington Consensus technocrats, but this was not the end of the story. When the left finally came to power in Brazil and Argentina in 2003, developmentalism would be brought back from the grave once again.

2002: losT aNd fouNd: The lefT iN search of a Program
At the beginning of this century, the left came to power under very different circumstances in Argentina and in Brazil. In Argentina, neoliberalism was defeated by a popular uprising. In Brazil, the PT won a disputed election, making significant concessions in the process. In both cases, left-wing administrations took office without having clear economic programs or development strategies except for a broad rejection of the Washington Consensus. In that context, neodevelopmentalism was adopted as a ready-made, viable alternative to neoliberalism.
In Argentina, the decentralized character of the protests that emerged against neoliberal policies at the end of Menem's and throughout de la Rua's administration did not converge in a coherent program led by any established political force. Instead, the political slogan widely adopted by social movements-¡Que se vayan todos! (Every one of them must go!)-revealed a radical rejection of liberal democracy and traditional political parties. Exactly because of its fragmented and horizontal character, the popular revolt produced many inspiring examples of grassroots activism (see, for instance, Sitrin, 2006). Nevertheless, it could not offer an alternative development strategy capable of reorganizing the Argentine productive structures after the meltdown of neoliberalism in 2001 (Sader, 2008).
At the same time, on the other side of the Uruguay River, Cardoso's administration was about to leave office with extremely low popularity ratings. According to one survey, 36 percent considered Cardoso's administration bad or very bad while only 26 percent classified it as good or very good (the remainder of the interviewees considering it only middling or failing to answer). Perhaps more revealing, for 33 percent of the interviewees politicians were the social group most privileged under Cardoso's presidency, closely followed by bankers (29 percent), and 49 percent believed that the working class was the social group that suffered the most under Cardoso (Datafolha, 2002).
By associating Cardoso's administration with the Washington Consensus, the PT successfully presented itself as an anti-neoliberal alternative to the continuity represented by Cardoso's candidate in the 2002 elections, his old partner José Serra. At that point, after eight years of neoliberal reforms, the elusive neodevelopmentalist promises of Serra's campaign lacked any credibility. To take power, however, Lula had to compromise. The prospect of the election of a working-class president for the first time in Brazil's history spread panic among the ruling class. As the 2002 elections approached, capital flight increased, the stock exchange collapsed, inflation rose, and the value of the Brazilian real quickly deteriorated. To appease capital owners, Lula signed the "Letter to the Brazilian People," a document directly negotiated with important representatives of the Brazilian bourgeoisie, including the Marinho family, the owners of Globo, Brazil's most powerful media empire. 10 While maintaining traces of anti-neoliberal rhetoric, the letter renounced the most radical aspects of the PT program and pledged commitment to stability and the macroeconomic policies implemented by Cardoso after 1999. For one reason or another-the lack of a clear alternative program or capitulation to the ruling classes' economic blackmail-the Brazilian and Argentine lefts ended up turning to neodevelopmentalism. The consolidation of neodevelopmentalism as the only viable alternative to neoliberalism was not immediate, however. In Argentina, the neodevelopmentalist macroeconomic cycle started with the break with the convertibility law and a sharp devaluation of the Argentine peso. The change in exchange-rate policy was not so much a choice as a necessity given the shortage of international capital and the consequent lack of dollars to maintain exchange-rate parity (Frenkel, 2002: 41-59;Kicillof, Nahon, and Costa, 2006). Nevertheless, Roberto Lavagna, the original architect of Argentine neodevelopmentalism, 11 seized the opportunity to "transform this particular conjuncture into a macroeconomic policy regime" (Kulfas, 2016: 109). After winning a convoluted election against Carlos Menem-who dropped out of the race during the second round, realizing that defeat was inevitable-Néstor Kirchner effectively established a neodevelopmentalist alliance between big national capital and organized labor, stabilizing economic policy for the next decade. The development strategy adopted by Lavagna and Kirchner has been praised by Bresser-Pereira as an at least partially successful attempt at "new developmentalism" (Bresser-Pereira and Theuer, 2012; see also Bresser-Pereira, 2013).
In Brazil, the cleavage between neoliberals and neodevelopmentalists that appeared in Cardoso's first administration was reproduced after Lula's inauguration. Reaching beyond the PT cadres, Lula recruited his Central Bank president directly from Cardoso's Partido da Social Democracia Brasileira (Brazilian Social Democracy Party-PSDB). 12 The Finance Ministry was entrusted to Antonio Palocci, a Trotskyist converted to neoliberalism and one of the minds behind the "Letter to the Brazilian People." The opposing neodevelopmentalist faction was led by Planning Minister Guido Mantega, a former collaborator of Cardoso's,13 and Carlos Lessa, an old collaborator of Celso Furtado and a direct representative of old-school developmentalism, named president of the powerful National Development Bank.
Contrary to what had happened in Cardoso's administration, after a couple of years of hesitation the neodevelopmentalist faction ended up winning the internal dispute for power. Pressed by the sluggish growth rates, the poor results achieved in the local 2004 elections, and the first corruption allegations against his government in 2005, Lula boosted social programs and empowered the neodevelopmentalist faction. In 2006 Guido Mantega replaced Palocci as finance minister, starting a cycle of decided fiscal expansion and state-led investment in strategic sectors. As in Argentina, the implementation of neodevelopmentalist policies was premised on an alliance between capital and labor skillfully articulated by Lula. At a theoretical and discursive level, the turn to neodevelopmentalism in Brazil and in Argentina went hand in hand with new expressions in the neodevelopmentalist literature. Economists close to the PT proposed a form of social developmentalism, emphasizing the role of the state in reducing poverty and inequality (see, for instance, Mercadante, 2010;Neri, 2010;Pochmann, 2010). In Argentina, beyond the writings of Lavagna himself, a new generation of developmentalist economists close to Kirchnerism highlighted the importance of industrial policies and "fine-tuned" economic management to bridge the technological gap characteristic of underdevelopment (Kicillof, 2016). Nevertheless, the best-articulated expression of neodevelopmentalism is to be found in the recent works of one of Cardoso's closest collaborators and his former state reform minister, Luiz Carlos Bresser-Pereira. After leaving Cardoso's administration in 1999, Bresser-Pereira (20062012;Bresser-Pereira, Oreiro, and Marconi, 2014) developed a full "new developmentalist" macroeconomic theory with great emphasis on the role of the exchange rate.
Despite differences of emphasis, all the variants of neodevelopmentalism share a renewed faith in the state as a privileged vector of capitalist catch-up development. Furthermore, all neodevelopmentalist writers appreciate the achievements of the 1950s-1960s' Latin American developmentalist tradition but recognize that, in the contemporary context, industrial production should be geared to the international market instead of aiming primarily at import substitution. Finally, all expressions of neodevelopmentalism depend on a political alliance between capital and labor that generates gradual gains for workers in terms of wages but does not directly challenge social-property relations. Even when potentially conflicting social interests are recognized, it is supposed that the state can rise above social classes and promote national development.
The continuity with Cardoso's ideas is striking. Although none of the contemporary neodevelopmentalist writers are willing to recognize it, all the key elements of neodevelopmentalism were already present in Cardoso's "new development program" of 1994. Furthermore, the underlying sociological premises of neodevelopmentalism stretch all the way back to Cardoso's bourgeois dependency theory. As Marini saw as early as 1978, the core purpose of neodevelopmentalism was lending credibility to the possibility of "American-or European-style Brazilian [or Argentine] capitalist development" (Marini, 1978: 102-103).
Among contemporary neodevelopmentalists, only Bresser-Pereira (2010) explicitly addresses Cardoso's dependency theory. Notwithstanding his engagement's being intended as a critique, it ends up revealing exactly how much contemporary neodevelopmentalism owes to Cardoso's ideas. Because Bresser-Pereira takes for granted the possibility of capitalist catch-up development in peripheral countries-exactly the bone of contention of the 1978 debate-he fails to recognize how close Cardoso's thesis comes to the developmentalist tradition. Cardoso's ideas are thus misrepresented as being primarily a defense of some form of association with international capital as opposed to the putative nationalism associated with the developmentalist tradition.
Although Cardoso did indeed believe in the potentialities of association with international capital, his thesis, as we have seen, is primarily about the possibility of engendering capitalist catch-up development against a revolutionary socialist alternative represented by Marini. Modernizing national elites are perceived as the privileged actors in Cardoso's theory, just as in Bresser-Pereira's. The marginal disagreement between the two is reduced to the weight the relative participation of international capital should have in the national development process.
Contrastingly, from a Marxist perspective akin to Marini's dependency theory, Cardoso's and Bresser-Pereira's ideas appear at best as different shades of neodevelopmentalism. Defying hundreds of years of historical evidence, both still have faith in an economic convergence between peripheral and central countries. Both believe that national elites will eventually be able to lead processes of catch-up capitalist development. Both ignore structural challenges such as racism, patriarchy, and the superexploitation of labor. Both underestimate the agency of the working class, which appears solely as the putative beneficiary of top-down, state-led development initiatives rather than as the active agent of its own history.

2021: fifTy years laTer, Back To mariNi
The social and economic limitations of the recent neodevelopmentalist cycle and the retreat of the so-called pink tide invites the Latin American left to rethink its development discourses and political strategies. A closer look at the genealogy of neodevelopmentalism should suffice to remind us that it never really represented a socially emancipatory development alternative. Neodevelopmentalism is just an updated strategy of the historically privileged ruling classes to mobilize the state in defense of their interests in a context of increasing international competition.
The theoretical critique of neodevelopmentalism and the denunciation of its antirevolutionary and elitist roots in Cardoso's dependency theory do not amount to a denial of the very concrete social achievements of recent neodevelopmentalist administrations. The first decade of the twenty-first century can only be compared to the heyday of classical developmentalism in the 1950s in terms of gains for the working class in Brazil and Argentina. While most of the world remained under neoliberal rule, the rise of neodevelopmentalism in Latin America allowed for a rare, if short, period of substantial economic growth and poverty reduction. The contrast with the mounting losses for the working class under Presidents Macri, Temer, and Bolsonaro could not be starker.
Although undoubtedly better than neoliberalism, neodevelopmentalism represents the subordination of the priorities of the working classes to a topdown, elite-led strategy of capital accumulation. Fortunately, however, it is possible to devise an alternative to both neoliberalism and neodevelopmentalism. The contemporary conjuncture, marked by social regression in Brazil and continued hardship in Argentina, offers the Latin American left an opportunity to break free of neodevelopmentalism and rediscover a revolutionary development discourse. Five decades after his pioneering denunciation of neodevelopmentalism, two crucial insights of Marini's dependency theory can be recovered as guidelines for a truly emancipatory development strategy that shifts to the forefront the priorities of the working class and other historically oppressed groups.
First, it is necessary to reaffirm the link between the superexploitation of labor and Latin America's persistent economic underachievement. As long as wages are depressed, development necessarily appears as short-lived periods of economic growth boosted by international demand, as in the recently exhausted commodities supercycle. A substantive increase in wages as part of a radical strategy of wealth and income redistribution must be a cornerstone of the left's program. Instead of a long-term perspective of gradual gains to avoid social unrest-as proposed by neodevelopmentalists-a radical development strategy must start with a demand for social change based on the valorization of labor over capital.
Second, as Marini, Bambirra, and the first generation of dependentistas made clear, it is necessary to give up any faith in the progressive and democratic character of national bourgeoisies in Latin America. When confronted with the prospect of leading national development projects that could potentially undermine their own privileges in the future, Latin American ruling classes have chosen to retreat to their internationally subordinated position to cement their power. The traumatic removal of Dilma Rousseff's administration in Brazil is only the most recent example of this century-old ruling-class reproduction strategy. After obtaining endless concessions for more than a decade-including tax breaks, subsidized credits from the National Development Bank, and debt amnesty-the Brazilian bourgeoisie did not hesitate to abandon the PT, break the social pact with organized labor, and actively campaign to overthrow an elected government. The subsequent election of a former army captain who openly defended the dictatorship, praised torturers, and refused to take meaningful measures to protect the population during the COVID-19 pandemic shows just how far capital owners are willing to go to guarantee their class privileges. The current tragedy in Brazil demonstrates that Latin American ruling classes are more tolerant of genocide 14 than of the prospect of moderate social change. The sooner the left realizes that, the better. Because of the intrinsically conservative character of Latin American national bourgeoisies-the result of centuries of entrenched privilege, galvanized by racial and social prejudices against the lower classes-nothing can be expected from putative progressive alliances between labor and capital. As a consequence, a truly emancipatory development strategy can prosper only if led by the subordinated classes themselves.
It has been five decades since the Latin American left could propose a coherent development discourse capable of posing a real challenge to the mainstream liberal and the development traditions. Since then, heroic social movements and fringe left-wing parties have offered real political resistance to neoliberalism and neodevelopmentalism but have not been able to propose a consistent alternative. For this alternative to be built, the left must abandon once and for all the false promises of neodevelopmentalism and refuse to be once again junior partners in social coalitions led by national bourgeoisies. It is time to impose our own agenda and reclaim a radical anticapitalist perspective on development.
NoTes 1. A collection of some the most representative writings on Latin American classical developmentalism can be found in Bielschowsky (2000). Furtado (1976) is the most iconic work of the Latin American classical developmentalist tradition translated into English.
2. The classical developmentalist literature became increasingly aware of political and sociological obstacles to capitalist development in Latin America after the deceleration of growth in the 1960s (see, for instance, Furtado, 1965a;1965b). Cardoso and Serra (1978) mistakenly read Marini and other Marxist dependentistas as making the same stagnationist argument as Furtado. 3. Luiz Carlos Bresser-Pereira is today the most prolific advocate of "new developmentalism." He helped to found Cardoso's Partido da Social Democracia Brasileira (Brazilian Social Democracy Party-PSDB) in 1988 and served as minister of state reform under Cardoso for four years. Guido Mantega did his Ph.D. in sociology at the Universidade de São Paulo under Cardoso's direct influence and worked at the Centro Brasileiro de Análise e Planejamento (Brazilian Center for Analysis and Planning-CEBRAP), Cardoso's think-tank. Both are professors at the Getúlio Vargas Foundation in São Paulo, the hotbed of Brazilian neodevelopmentalism.
4. I thank an anonymous reviewer for Latin American Perspectives for raising this point. 5. Unfortunately, Paulo Guedes does not publish peer-reviewed articles, which would allow for a more serious engagement with his ideas. A sample of them can be found in the many interviews he gave during the 2018 elections in Brazil (see, for instance, Guedes, 2018).
6. As explained below, Cardoso would be instrumental in the implementation of the Washington Consensus agenda in Brazil, and Serra would be one of his key collaborators.
7. Here Marini comes the closest to Marx and anticipates part of the argument highlighted by political Marxists (for instance, Wood, 1995).
8. The macroeconomic "tripod" refers to primary fiscal surplus targets, floating exchange rates, and inflation targets. The alignment of these three key macroeconomic variables was expected to produce a favorable environment for the market to work its magic and generate the best possible development of the economic potentialities of any given country. For a defense of the model by some of its proposers, see Fraga, Goldfajn, and Minella (2003). For a critique, see Arestis, Ferrari-Filho, and de Paula (2011). 9. For this group's not very edifying views on the developmentalist tradition, see Franco (1998).
11. This article has largely focused on the Brazilian side of the genealogy of neodevelopmentalism. In Argentina, a parallel genealogy of neodevelopmentalism could probably be traced. This genealogy would certainly have to include the work of Eduardo Basualdo and the FLACSO school in Buenos Aires (among others, Arceo et al., 2010;Basualdo, 2010;Basualdo, Nahon, and Nochteff, 2007), as well as the already mentioned writings of the finance minister Roberto Lavagna. Although a complete intellectual history of Argentine neodevelopmentalism has yet to be written, I would suggest that, similarly to what has happened in Brazil and across Latin America, the political and epistemological space for the emergence of neodevelopmentalism in Argentina appeared in the 1980s, following the crisis of radical Marxist alternatives such as dependency theory.
12. The reference here is to Henrique Meirelles, a former top executive of BankBoston and later finance minister under President Temer. In 2018 he ran for president, getting fewer than 1.5 percent of the votes.
14. The term "genocide" is not used lightly or rhetorically here. Bolsonaro's government is actually being accused of perpetrating genocidal policies by indigenous communities and leading international genocide scholars (see State Crime Initiative, 2021)