1The Latin American continent has been considered a paradigm of dependent economies since the studies of the Dependency Theory in the sixties and seventies. That is why a study of this continent is a significant contribution to this particular issue of the Revue de la régulation. In an article written by Nölke and Vliegenthart, the concept of dependent market economies (DME) is used to describe the economies of East Central Europe. These authors situate the discussion of dependency in the context of the school of the Varieties of Capitalism (VOC), and consider that Hall and Soskice have a limited vision, as they merely consider only two types of capitalism, the well-known liberal and coordinated types. They consider that the post-socialist economies of East-Central Europe cannot be defined as either a hybrid or an intermediary form of these two types of capitalism. They write that in order to analyze this case, it is necessary to propose a specific type of capitalism, which they call dependent and describe as based on “…skilled but cheap labor, the transfer of technological innovations with transnational enterprises, and the provision of capital via foreign direct investment (FDI)” (Nölke and Vliegenthart, 2009, p. 672). This type of economy is based on an assembly platform for semi-standardized industrial goods, which depends on intra-firm hierarchies within transnational enterprises, and is controlled by the headquarters of these enterprises (ibid., p. 680). Furthermore, “…the TNC’s have fully integrated the Central and East European subsidiaries into their company networks” and they control them in a hierarchical manner; they are dependent on the decisions of their global empire, rather than of the particular country where they are installed. (ibid., p. 684). Furthermore, labor costs and a disciplined labor force are fundamental to assure the competitivity of these economies (ibid., p. 686), there is no incentive to invest on qualifying local labor, and competitivity slowly erodes due to the competition of ever cheaper countries further east (ibid., p. 687). As we will see below, although the Mexican economy could very well fit into this model, the economies of most other countries of Latin America, such as Brazil and Argentina, that in the last 15 years (before 2014) turned their economy towards the internal market, or the Andean countries, which depend absolutely on commodities, would hardly correspond. Although all these economies definitely depend on the world economy, their dependence is not the same as that of Mexico or the Central and East European countries.

2Other authors in this issue go in the same direction. Drahokupil and Myant likewise consider that there are different manners for countries to be dependent. And that these differences emerge when we consider that the “dependent variable” has to incorporate broader forms of integration into the world economy, “...the structure of the inherited industrial bases, natural-resource endowments, state capacity, [and] the rule of law.” (Drahokoupil and Myant, 2015, p. 4). When we consider these variables, we find that “different forms and different combinations may be compatible with the same form of international integration”. Then they go on to define five forms of capitalism in the transition economies of post-communist countries (Drahokoupil and Myant, 2015, p. 5).

3The literature that discusses Latin America from the VOC perspective considers that the principal characteristic defining this continent’s type of capitalism is the societal and economic heterogeneity (Schneider and Soskice, 2009; Schneider, 2014). Similarly to Nölke and Vliegenthart, Schneider proposes a unique type for Latin America: the hierarchical one; a unique variety that is in fact a deficient variant of the liberal market economy as defined by Hall and Soskice. Although Schneider interestingly emphasizes some of the most significant characteristics of the Latin American economies (Schneider, 2014), the idea that all the countries in the continent fit into this type is not adequate if one takes into consideration important structural and institutional differences between certain groups of Latin American countries.

4And in fact, the founding authors of the Latin American Dependency School, Cardoso and Faletto (1969), already discussed the diversity of the forms of dependency of Latin America countries. They differentiated between countries like Argentina and Brazil, in which local capital produced foodstuffs indispensable for the metropoles in large areas of the colonies, and countries like Mexico and Peru, which exported minerals that could be substituted by productions from other countries, were extracted in isolated enclaves, with little connection to the rest of the economy, and were owned by foreign capital. The first type of dependence led to the existence of a local bourgeoisie that made a relatively more stable transition from a colony to an independent country and facilitated the beginning of industrialization, while the second had a more difficult transition to independence, due to the characteristics of the enclave economy and the fact that it did not develop a local bourgeoisie.

5The global crisis of 2008 as well as the commodity super cycle that ended in 2014 revealed significant differences between Latin American countries in how they responded to the crisis, benefited from the super cycle, and are adjusting at present. Although both the global crisis and the super cycle of the commodities was a unique phenomenon that impacted most of the countries of the world, its consequences on a particular country differed considerably depending on its socio-political institutions and economic and social structures. One can approach these differences in two ways: an historical one, proposing that all countries are different, that any comparison flattens out long term differences that are crucial; an analytical perspective, meaning that although there are significant differences amid countries, there are also significant similarities that, while not being generalizable to define one single mode, can serve to theorize a number of ideal types. While the first perspective practically excludes any comparisons, the second one allows for constructive comparisons of groups or clusters of countries with similar structures and institutions, and which pursue growth and development in a similar manner.

7To analyze other less developed countries, where non-complementarity between the institutional forms configure a non-coherent or disarticulated form of capitalism, this article borrows its perspective from Cardoso and Faletto (1969), who considered that the different countries developed in distinct manners according to what they produced, their insertion to the world economy and the different coalitions that led the development strategy. I am also closely related to a long tradition of structuralist analysis, among others those of Barrington Moore, Theda Skockpol, Esping Andersen, Colin Crouch, as well as that of the Regulation Theory, centered both on structures (class structures, the structure of production, the State) and institutions (workers’ and employers’ organizations, social protection systems, the wage relation). In addition, this article builds upon several of our previous analyses (Bizberg and Théret, 2012; Bizberg, 2011; 2014; 2015) as well as those of Boyer (2014), and Marques-Pereira and Théret (2004). With respect to my own studies, while the articles cited above (2011; 2012; 2014) dealt mainly with four countries —Argentina, Brazil, Mexico, and Chile—, in this article we extend our study to the commodities-producing countries, in order to develop the idea of four rather than two types of capitalism in Latin America. Furthermore, while the other articles were more descriptive, the present one adopts a more conceptual perspective. Finally, the present article is a tight synthesis of a book that will be published by Palgrave-Macmillan beginning 2019.

8On this basis, I outline six analytical dimensions in order to define the different types of capitalism existing in Latin America:

1. The accumulation regime.

2. The mode of integration to the world economy

3. The role of the State.

4. Social actors and coalitions behind one or another mode of accumulation.

5. The political system

6. The social pact/wage regime.

9This article is divided in three sections: in the first one I will define the four distinct types of capitalism, in the second section I will analyze each of the varieties according to the six dimensions we have just defined, and in the third and concluding one, I will discuss the frailties of each of the types of capitalism.

1. Four types of capitalism in Latin America: dimensions and definitions

10In our previous studies, we found significant differences between Mexico and Brazil, which we formalized into two ideal types of capitalism:

1. • 2 A situation that resembles the dependent capitalisms described in other articles featured in this i (...)

   The first one, a stylization or formalization of Mexico, which shares characteristics of the central American countries and the Dominican Republic, is an Outward-looking Liberal Peripheral Capitalism, which can be better characterized as an International outsourcing capitalism. It is a disarticulated form of capitalism that depends entirely on the demand of parent companies situated in the United States or other central countries (mainly Germany, Japan, and Korea), and that assembles imported spare parts.2 It produces manufactures that can have a relatively high technological content, although as the main process occurring in this country is the assembly of spare parts that come from all over the world, the aggregate value added is very low. The mode of accumulation is extensive, as productivity gains are low. Production is disconnected from the rest of the productive structure; there are few, if any, backwards linkages to national producers. The production chains are disconnected between the external market and the internal final stages of production. The country is a platform for the last stage of the production process. It thus depends on low labor costs and high labor market flexibility, and the repression of internal demand. The mode of consumption is thus undeniably profit led; salaries are maintained as low as possible.

The socio-political situation that favors this type of capitalism is one of weak social actors, a dominant coalition made up of financial capital, large domestic and international companies. The State is weak, having been more or less dismantled after adopting the orthodox recipes of neo-liberalism; in fact, it functions as an agent of the neoliberal model. The State is unable/unwilling to impulse a developmentist industrial policy like the one implemented in Korea and Taiwan, which resulted in the upgrading of the industry on the basis of national suppliers of industrial enterprises.

The wage relation is based upon the repression of wages and the productivity growth of exporting industries, such as automobile, steel, and electronics (Palma, 2005), together with low salaries and low productivity growth in the rest of the economy.

1. • 3 A mode that has its roots in past economic trajectories, as we tried to show in an article (Bizberg (...)

   A second capitalist variety, a stylization of the mode of development that was followed by Brazil from the beginning of the year 2000 to 20143, an Inward-looking State-led Peripheral Capitalism, or a neo or Socio-developmentism. This capitalistic form produces both commodities for the external market and manufactures for the internal one. It is thus a mode of accumulation that is both based on extensive production of commodities (in some respects also with an ever greater productivity) and on the intensive production of some manufactured products, such as steel, weapons, planes, biofuels, etc. The State is a fundamental actor that tries to arbitrate between the external dependence of a peripheral commodities-producing economy, linked to financial capital, and an industrial production destined to the internal market. The mode of consumption is also a settlement between both: the search for foreign markets and attraction of foreign capitals and the inducement of both domestic industry and internal demand through increasing wages and a generous social protection system. It is a capitalist form that depends on both the exports of commodities and the input of financial capital as well as internal market growth.

The State is strong, interventionist, tries to find a balance between a wage growth model and a profit led model; an internal market-led growth with an export-oriented growth. It uses external resources, emanating from the export of commodities, income of foreign portfolio investments, FDI, and debt, in order to develop industry. It attempts to direct the economy through industrial policy and finance it with a large development bank (the BNDES). All of which defines a neo-developmental strategy.

This capitalistic form is based on a strong civil society that exerts pressure on the State to redistribute. The State is thus forced to arbitrate between the international capital, large domestic enterprises, financial interests and the poor, the workers, the low middle classes; all sectors that constitute the social pact. The wage relation is characterized by expansive wage policies and an expansive social protection system oriented to universality, which tends to reduce inequality. This scheme expands internal demand, all of which defines a socio-developmentist strategy.

11These two types, nonetheless, do not encompass all Latin American economies, most notably the ones that are essentially based on commodities, the rentier types characteristic of Andean countries. For these countries we have constructed two other varieties, liberal rentier and redistributive rentier capitalism.

1. The first mode is a stylization of Peru, Colombia, and partly Chile. The first two countries share an extremely open economy, a very weak State, weak unions, deregulation of the labor market and a reduced, assistance oriented, social security system. The dominant social coalition is constituted by large multinational and domestic companies, the financial sectors, and the middle classes, excluding the popular classes. While Chile shares most of these characteristics with these two countries, the Chilean State has a capacity that other liberal States, such as the Peruvian, Colombian or even the Mexican do not have. The Chilean State owns Codelco, which controls about 30% of the country’s copper exports. Part of its resources are kept in a reserve fund that is used as a counter-cyclic instrument in case of a crisis; something none of the liberal countries have. On the other hand, the State performs effectively its functions: taxes, police, and certain aspects of economic regulation. In these three cases, nevertheless, the dominant social coalition is equally formed by foreign and domestic large companies, as civil society and social organizations are weak. The mode of consumption is oriented towards profits. Wages grow below productivity gains.

2. Finally, the fourth capitalist variety is a redistributive rentier capitalism. It is equally dependent on the international commodities market, but the existence of strong social actors exerts pressure and drives the State to intervene in the economy and redistribute the rent. Bolivia and Ecuador have a relatively strong State, which exploits or lends mineral or oil concessions in exchange for royalties and taxes. The State has established a compromise between the interests of the mobilized popular classes and the multinational companies, the financial interests and the local entrepreneurs, in order to distribute the rent of the commodities exports between profits and wages. The popular classes are part of the dominant social coalition, due to their organizational and mobilizational capacity.

We can summarize the differences in the next table, which we will analyze more in depth in the second part of this paper.

2. The political economy of the four types of capitalism in Latin America: a regulationist perspective.

12In this second section, I will analyze the four varieties of capitalism according to the six dimensions we have previously defined.

2.1. The mode of accumulation.

13As I mentioned above, this dimension includes what a country produces, how it produces it, and the manner in which it redistributes wealth between profits and wages. Table 2 synthesizes the discussion of this section.

Table 2. Mode of Accumulation

	International Outsourcing	Socio-developmentism	Rentier/Liberal	Rentier/ Redistributive
Productive Structure	Low added value manufactures (assembly)	Commodities and manufactures	Commodities	Commodities
Mode of accumulation	Mainly extensive/intensive in some specific sectors	Extensive/Intensive	Extensive	Extensive
Mode of Consumption	Profit led growth	Wage/profit led growth	Profit led growth	Wage/profit led growth
Character	Disarticulated productive structure	Intent to articulate external and internal market	Articulated upon the external market	Articulated upon the external market

Source: Own elaboration

14The literature analyzing the consequences of what a country produces on its socio-political organization is significant: Cardoso and Faletto (1969) associate what a country produced in colonial times with the rise of a national bourgeoisie; Engerman and Sokolof (1997) link climate and socio-economic organization with productive structure; Evans (1995) and Kay (2002) correlate rentier productive structure with the character of the national State; Boyer (2015), Bresser-Pereira (2015), and Salama (2012) analyze the negative relation that exists between different types of rentier economy (commodities, land, housing, financial), “Dutch disease”, and industrialization.

15In the central economies, Fordism was an economic mode based on the production of manufactured goods, where production was increased through productivity growths (thus intensively) and through a mode of consumption that compromised among profits and wages. Even though the Latin American countries, or peripheral capitalism in general, have industrialized to a certain degree, a profit led mode, which depends on the appropriation of rent on the part of an oligarchy, has almost always prevailed. During the import substitution industrialization (ISI) period (1945-1980), an intensive mode of accumulation was implemented in the largest countries of the continent, accompanied by a form of consumption based upon redistribution through wages and social protection. Starting with the lost decade, when import substitution was abandoned, a bifurcation of the trajectory of the different countries in the continent began. Some countries abandoned ISI to return to the production of commodities; a situation that included countries that had basically never industrialized (Bolivia and Ecuador) or had de-industrialized (Colombia, Peru and Chile). Other countries became out-sourcing platforms assembling manufactures for export: Mexico and the smaller countries in central America. Finally, some countries tried to continue developing industry: Brazil.

16A fundamental distinction between the Latin American economies is whether they mainly produce commodities or manufactures, or a combination of both. While the first type of economy, the rentier, bases its growth on the mere extension of production, the other types depend on its intensity, on the rise of productivity. To a certain degree, all Latin American economies depend on extension, as they are all partly rentier: the case of oil and other mining products in Mexico, of agricultural, mining, and oil in Brazil. Finance can also be considered a rentier activity, and this is the case in both Mexico and Brazil (Boyer, 2015; Bresser Pereira, 2017). On the other hand, international outsourcing depends more on extension of investment than on innovation and productivity hikes, as neither the State, nor the capital intend to modify the organization and upgrading of production. As Palma (2005) has proven, outsourcing in Mexico depends less on productivity increases than on the differential between levels of productivity similar to those in advanced countries and salaries in peripheral countries.

17Fordism, where wages follow closely and sometimes surpass increases of productivity, is a wage-led growth, and is thus based on the increase of demand. The liberal mode of development, which has been implemented since the demise of Fordism, is a profit-led growth, based on offer (Boyer, 2015, Stockhammer, 2011). Socio-developmentism is likewise a wage-led growth mode, a mode that, in an open economy, has to be careful to balance internal demand with domestic offer, in order to prevent the growth impulse from being transferred to the exterior through imports (Bresser-Pereira, 2015). The international out-sourcing capitalism is based on profits, on the gap between productivity similar to that of the advanced economies and the salaries of the peripheral economies; where the State’s function is to repress salaries and social protection costs. The liberal rentier economy channels rents towards profits, while the redistributive rentier regime redistributes part of them to wages. None of these two latter models is a sustainable growth model, as it depends wholly on the price of commodities determined by the world market; the physical resources are depleted and there is no effort to develop alternative economic sources. In the case of the redistributive type, redistribution does not lead to a wage growth regime, but basically to internal demand funneled towards imports: resources obtained by the export of commodities (directly by state enterprises or through taxes and royalties) mostly lead to an increase of imports, as the economy is subject to the Dutch disease and there is practically no industrial policy.

18In figure 1 we define the types of capitalism depending on whether the mode of accumulation is based on rent or productivity, and whether the mode of consumption is led by profits or wages.

19The mode of consumption depends on the strength and character of the social pact between State and social actors. While in countries where civil society is strongly organized the mode of consumption has been redistributive (Brazil, Argentina, Uruguay, Ecuador and Bolivia), in countries where civil society is weak, they have been profit led (Mexico, Chile, Peru, Colombia). In addition, in Mexico, and in general in the outsourcing economies, both accumulation and consumption are disarticulated. While the mode of accumulation is disarticulated as it is dependent on a productive structure that lies between countries (in the case of Mexico and central America, mainly between the United States and the home country), the mode of consumption depends heavily on remittances sent by a significant proportion of the population that has migrated, as well as other resources coming from all kinds of illegal activities, including drug smuggling. This is complemented by an offer of cheap consumer products distributed by informal commerce, some of which are smuggled into the country, do not pay taxes, salaries, or rent as they are sold in the streets.

2.2. The international insertion.

20The mode of integration to the world economy is fundamental in order to analyze the peripheral or dependent economies, as they are in one way or another dependent of the international market. Table 3 synthesizes the discussion of this section.

21The international outsourcing capitalism and the two types of rentier economies share the external orientation of their economies, and although the socio-developmentist may produce and export commodities, it is fundamentally oriented towards the internal market. Data concerning the weight of exports in both groups of countries confirms this: while in Brazil and Argentina the aggregate demand is balanced between the external and the internal market, in the case of Mexico the external market is much more significant. The impact of exports on the growth of GDP in Mexico was 58% between 2000 and 2008, and 67% in the previous decade (1990-2000), for Chile the percentages are 48% and 39%, respectively. In contrast, for Brazil and Argentina the numbers are 27% and 29% respectively between 2000 and 2008 (Bensusán and Moreno-Brid, 2012).

24The socio-developmentist economy is more protectionist, it applies safeguards, subsidies, loans, and implements industrial policies. Its relation to the exterior is much more defensive; It has, at some moment, imposed customs tariffs or proscribed the import of certain products. The financial sector is not as open, most banks are still in the hands of the State or of national capital. Brazil has not signed any free trade agreements that would oblige it to be much more liberal; especially not with the United States. Although most analysts have mentioned the failure of the Brazilian government to control the overvaluation of its currency, in 2014, during the second presidency of Dilma, the government intended to accomplish it by lowering the central bank interest rates (Prates, Fritz, and de Paula, 2017).

2.3. The intervention of the State.

25According to the regulation theory, the State is a fundamental actor which can be defined by its autonomy, its capacity, as well as by its embeddedness in the socio-economic context of a country. Table 4 synthesizes the discussion of this section.

Table 4. State Intervention

	International Outsourcing	Socio-developmentism	Rentier/Liberal	Rentier/ Redistributive
Character	Neoliberal State/agent of neo-liberalism	Intent of an embedded State	Subsidiary State	Tendency towards a clientelist/corporatist State
Autonomy	Weak	Strong	Weak	Strong
Embeddedness	Weak	Relatively Strong	Weak	Weak
Stateness	Weak	Strong	Weak (Chile-relatively strong)	Strong

Source: Own elaboration.

27We can redefine the types of capitalism based on whether the State plays a central role in pursuing capitalist development or the economy is left to the market. We can combine this taxonomy with our prior binary classification of the accumulation regime based on the emphasis given to either profits or wages. Our four types of capitalism in Latin America can be arranged accordingly (figure 2).

Source: Own elaboration

28We can identify two types of capitalism where the State has a significant role. In the first one, it orients capitalism towards the internal market through labor and social policies that increase the internal demand though the redistribution of profits towards the workers, and through industrial policies. In this type of capitalism, the State exerts a strong fiscal pressure on the employers and consumers, and tries to achieve an active integration to the world economy. It also regulates and defends national capital and the internal market, by implementing counter-cyclical measures in times of crisis. In general, facing a strong State, there are strong unions and business organizations that exert pressure on the government in order to preserve their particular or common interests, although there are exceptions where the developmentist State is authoritarian and represses or controls social organizations, such as Korea and Taiwan in the 50’s and 60’s, and more recently China.

29In the case of the redistributive rentier mode, the role of the State does not consist in boosting an economy oriented towards the internal market through industrial policies, investment, or innovation, but it is almost “purely” redistributive. Political and social relationships as well as the economic ones are defined by the fact that the State owns plants that exploit natural resources, or that it extracts taxes and royalties from private enterprises. In many cases, its abundant financial resources are distributed without any productive goals, they are expended in a clientelist, or state-corporatist logic. These resources may either be used to foster organizations serving the State as political bases (Venezuela), or, when pressures from autonomous social organizations or movements exist, they may foster a participatory democracy (Bolivia, Ecuador).

Source: Own elaboration, based on Cepalstat

32On the other hand, the Brazilian federal government has a development bank unparalleled in Latin America and most countries of the world, the BNDES, which handles one fifth of all the finances of the Brazilian private sector (Santana, 2011; Hochstetler and Montero, 2012). From 1999 to 2009, its disbursements as percentage of GDP grew from 3% to an impressive 8.5%; although it was reduced to 5.5% by 2011. In contrast, the Mexican development banks have dramatically reduced their intervention and limited their character. The Mexican State’s main development bank, Nafin, has reduced its role since the 80’s. Its credit operations for productive investment were reduced by 70% between 1996 and 2004. It no longer finances companies directly, but serves as a guarantor for the loans offered by commercial banks to small and medium enterprises, and can also advance financial resources for the bills due by suppliers or clients (Manrique, 2007, p. 111-113).

2.4. The dominant social coalition and the type of political regime.

33In many respects, what the country produces and exports, how it does it and the character of the international insertion of an economy are determined by the orientation that the State and the dominant social coalition imparts them. If this were not so, we would be living in a perfect Ricardian world where every country would produce what it is best endowed to produce (Haggard, 1990). That is why one of our most fundamental assumptions is that the type of capitalism is in fine defined by the character of the dominant social coalition.

34This idea has a long tradition. Cardoso and Faletto, as well as Haggard, show how a strong and cohesive industrializing coalition (including the State) explains the depth of industrialization (Cardoso and Faletto, 1969; Haggard, 1990). Esping Andersen (1990) and Théret (2011) emphasize the role of the coalitions and the State in shaping the character of the national social protection systems. Crouch explains the cooperative or contentious character of unionism by analyzing the relationship and coalitions around the State, the church and the guilds during the formation of the nation-states (Crouch, 1993). Amable (2005) and Boyer (2015) consider the participation of the salaried working class and the State as crucial to define the types of capitalism. Finally, Bresser Pereira considers that whether a social coalition includes rentiers and financiers, or rather industrialists and a developmental State, is crucial to define the development possibilities of peripheral countries (Bresser-Pereira, 2017).

35Where social actors are strong enough to form part of the pact, and the State is capable of creating a broad coalition including industrialists, middle class and the working class in a compromise with agro-exporters and financial sectors, we have either a socio-developmentist type of capitalism —if based on productivity gains— or a rentier redistributive regime —if based on rents—. On the other hand, where social actors are weak and the social pact is essentially constituted by the State, multinationals, large national and foreign entrepreneurs and financial capitals, gains in either productivity or rent are mostly oriented towards profits.

36During the seventies and eighties, the strength of civil society depended greatly on its participation in the democratization process, and the sequence between liberalization and democratization. In Mexico and Chile, civil society was not crucial in the democratization process, and thus liberalization occurred prior to democratization. This meant a more orthodox economic liberalization, which in turn undermined the social actors, especially unions. In this case, the dominant socio-political coalition that emerged during the nineties and 2000s did not include organized civil society. In addition, since civil society was basically absent of the process of democratization, this process gave rise to a liberal, purely electoral democracy, with strong particratic tendencies, impervious to the interests of the working class and the lowest revenues. In some cases, as in Peru and, at least partially, in Colombia, both the civil society and the political system were de-structured and therefore the political system now tends to support a “pure” technocratic government, equally unreceptive to popular interests (Bizberg, 2011; Aziz, 2015)

37Where the civil society has been a central actor in the process of democratization, this process preceded liberalization and the society strengthened through democratization and by its resistance against the liberalization process. It was reinforced by both processes and imposed itself as part of the socio-political coalition that emerged during the nineties and 2000s. This was also the case in Brazil, where trade unions and the civil society were central in the political transition, and contributed to the creation of a socio-democratic party, the PT. This was also the case of Argentina after the emergence of a myriad of social movements in the wake of the deep socio-political crisis of 2001-2002. In Argentina and Brazil, unions as well as social movements have been very active: in the first, this was the case for the piqueteros and the human rights movements; in the second, a myriad of different social organizations that pushed forward the democratization process and the drafting of the 1988 Constitution have maintained their intervention on social policies through informal and formal (the councils) channels (IzunzaVera and Gurza Lavalle, 2012.)

38In the rentier distributive economies (Bolivia and Ecuador), these movements, more than the unions, have been especially significant. These countries are less industrialized and, in the case of Bolivia, the very strong mining Confederación Obrera Boliviana lost influence in the face of the new social movements: the indigenous and coca producers of Chaparé in Bolivia, which are the basis of the Movimiento al Socialismo (MAS) that led Evo Morales to the presidency (Mayorga, 2011). In the case of Ecuador, the indigenous movement, the CONAI, has been a central social and political actor. In both countries, they have been a very significant mobilizing force that has imposed a popular alliance on the State. Social movements emerged in the struggle against liberalism during the nineties, a process that led to the arrival of a national-popular government in the 2000s (Silva, 2009); in Bolivia, Evo Morales and the MAS, in Ecuador, Rafael Correa. In these cases, the political system is very open to the civil society, the demand for redistribution comes from below, from autonomous social movements. These political regimes can be defined as either participatory or movementist democracies (Bizberg, 2010). They can also derive towards a delegative democracy if and when the State uses the resources obtained by the exports of commodities in order to control social organizations in a clientelist or corporatist manner.

Figure 4 illustrates the four types of capitalism according to the variables we have been discussing in this article, adding the strength of civil society and the openness of the political system. Table 5 summarizes the discussion of this section.

Table 5. Dominant social coalition and type of Political Regime

	International outsourcing	Socio-developmentist	Rentier/Liberal	Rentier/Redistributive
Dominant Social Coalition	Oligarchic: State, multinationals, large national entrepreneurs, financial capital, middle classes	Compromise between oligarchies and popular classes: State, middle classes, unions in a compromise with agro-exporters and financial capitalism	Oligarchic: State, multinationals, large national entrepreneurs, financial capital, middle classes	State / Civil Society Alliance: State, Popular sectors, lower middle classes
Civil Society	Weak	Strong	Weak	Strong
Type of democracy	Weak representativeness Particracy	Strong representativeness: Participatory democracy	Weak representativeness; Particracy or de-structured political system	Strong representativeness: Movementist/Delegative democracy

Source: Own elaboration

2.5. The wage relation; the social pact

39The wage relation does not comprise only wages, but also indirect forms of remunerations, labor market regulations and the social security system (health, pensions, unemployment benefits, etc.). State intervention, as well as the strength and organization of the labor movement and of the employers organizations and their relation, are central to define the character of the wage relation, which reposes upon a social pact between these three central actors.

40We have basically two situations: while in the outsourcing and in the liberal rentier types (Chile, Peru, and Colombia) the industrial relations have been radically deregulated and made more flexible, social protection systems have been transformed into assistantship; in the socio-developmentist type (Brazil, Argentina, Uruguay) although industrial relations were also liberalized, the force of the labor movement and the fact that the government has had, at some moments, a close relationship to worker’s and other social organizations, contributed to the fact that social protection systems have better endured in their traditional forms and industrial relations continue to be regulated.

42Although industrial relations have also been liberalized in Brazil, unions have managed to retain an important degree of autonomy and capacity of action. This is partly due to the fact that the labor movement in Brazil was, together with numerous other social movements, a central actor in the democratization process, but also because it never lost its character as an interlocutor of the successive governments, even with the more liberal ones. During the presidency of Cardoso, for example, the government promoted negotiations between employers and labor (the tripartite Cámaras sectoriais) in order to set conditions for the modernization and growth of several branches of the economy. More recently, under the presidency of the PT with Lula, the government implemented negotiating bodies to discuss the labor law, and the tripartite Social-Economic Council. Since the eighties, unions have managed to impose local representation, through delegates, in some of the largest companies (Cardoso and Gindin, 2009).

44A very direct indicator of the character of the social pact is the data on the rise of wages, especially minimum salaries, which have an impact both on active workers and pensioners. They are also a good sign of whether the mode of economic growth is redistributive or liberal, and on the weight that is given to either profits or wages. In figure 5, we can clearly see that in the liberal rentier type (Peru, Colombia, and Chile), salaries have grown moderately, while in the outsourcing type (Mexico) they have stagnated; in both, the State represses salaries, especially in the outsourcing mode, as they constitute its principal competitive advantage. In contrast, in both the socio-developmentist (Brazil and Argentina) and the redistributive rentier countries (Bolivia and Ecuador), minimum salaries have grown strongly.

Source: own elaboration based on Cepalstat

45The character of the social protection system also shows a contrast between the different types of capitalism. Social protection policies have a short-term impact on domestic demand through pensions, unemployment compensations, health investment and expenditure, and a medium and long-term impact through productivity growth. The expansion of public resources dedicated to social policies and health is very significant in the three countries (Argentina, Brazil, and Uruguay) coined “socio-developmentist”. Other countries have all lagged behind (see figure 6).

a. Central government; b. general government; c. central government

Source: Own elaboration based on Cepastat

46Nonetheless, as Esping Andersen (1990) has written, to know how much a country spends on welfare is not enough to understand its characteristics, it is necessary also to understand the way resources are used. One of the main policies where social resources are used is social security pensions. If one looks at pension coverage (see figure 7), high levels of coverage coincide with high union density and a redistributive economic model. The case of Chile, as an exception, is interesting: it was one of the countries with the most widespread coverage before the military regime, when the inclusive pension system was privatized and converted into a capitalization scheme. Although this country has a high coverage, the Lagos government (2000-2006) instituted a minimum salary pension for workers who did not reach this level through capitalization; this has as a consequence a high coverage, but a low replacement rate for higher salaries.

47In general terms, in countries with a strong labor movement or strong social movements, pensions and health coverage are more extended and more generous than in countries with weaker unions. We can see this difference in the data on coverage of figure 7. These divergences are directly related to the governmental policy aimed at the formalization of workers. While in Mexico and Peru practically nothing has been done to reduce informality (one could hypothesize that informality is functional to the type of capitalism that exists in these countries) in Argentina and Brazil, tax incentives and stricter work inspections have resulted in a decrease of informality (Maurizio, 2014;Berg, 2011). Although Chile has been able to reduce informality to around 22% of GDP (Quenan and Velut, 2011, p. 52), Bensusán (2006) considers that this has been achieved at the expense of lowering considerably the conditions of formalization. Colombia has an almost universal basic health system, similar to the Mexican Seguro Popular, which includes a basket of diseases, but in contrast with Mexico, in Colombia, it is a compulsory, subsidized, contributory system.

48In Mexico, the welfare regime was closely linked to the needs of the authoritarian PRI; it was the way in which the State exchanged benefits for control of the social organizations in a corporatist scheme. Since the 1980’s, the Welfare State has evolved towards a more universal, albeit minimalist scheme. Since the mid 90’s, social programs have decidedly shifted to assistance (Valencia Lomelí, 2008). The main social program Prospera focalizes on the poorest part of the population.

49Brazil and Argentina (since 2003) stand in sharp contrast to Mexico: in the first place, the welfare regime was not radically modified, especially the case of Brazil. It did not abandon the “pay as you go” pension system or even institute a mixed one. Non-contributive pensions to the rural workers were expanded: 12.8 million people get a minimum salary. Another social assistance program, called the BPC (Beneficio de Prestação Continuada), covers about 4.27 million old (over 65) or incapacitated individuals which live in a home with a revenue lower than one fourth of a minimum salary (Lautier, 2007, p. 60-62). In Argentina, in 2008 the Fernandez government renationalized the pension funds, which had been partially privatized by the Menem presidency. The government unified the system under a State controlled solidary regime, eliminating the segment of capitalization administered by the AFJP (Administradoras de Fondos de Jubilaciones y Pensiones) (CEPAL, 2010, p. 8-9).

50The impact of a socio-developmentist mode of growth is clear in the case of Brazil, Argentina, and Uruguay, both in terms of decreasing inequality, reducing extreme poverty and expanding the middle class. The significant increase in salaries, especially minimum salaries, as well as a process of formalization of the working force, contributed to this aim (Saboia and Neto, 2017 for Brazil). In addition, in Brazil, there was an intensive, pro-poor housing program, financed wholly or partly by the State, and an increased access to credit of the working glass, which allowed them to acquire durable consumer products. The countries closest to the liberal rentier type have also shown a very positive performance in reducing inequality and extreme poverty, although this was less due to an increase in wages, work formalization, or social security expenditure (as the accumulation regime is profit oriented), than to a “mechanic” effect of economic growth. Nevertheless, the reduction of inequality has been less significant in these countries, with the exception of Peru. The outsourcing mode, typified by Mexico, is exemplary for its stability in its inability to reduce poverty or inequality (figres 8 and 9).

*Mexico data based on Cepalstat

Table 6 synthesizes the discussion of this section

Table 6. The Wage relation

	International outsourcing	Socio-developmentist	Rentier / Liberal	Rentier/Redistributive
Wage policy	Wage repression	Strong increase of minimum wages	Increase in salaries according to productivity	Strong increase in salaries
Labor market policy	Non-formalisation	Strong formalisation	Non-formalisation (except Chile)	Formalisation
Union density / Civil society	Low	Neo-corporatism	Low	Strong social movements
Pensions	Low coverage of contributory and social pensions	Increasing coverage of contributory and social pensions	Low coverage of contributory (except Chile) and social pensions	Low coverage, increasing social pensions
Health system	Social security/public/private: strong segmentation/non-coverage	Public/private (Brazil); plus social security (Uruguay, Argentina)	Social security/public/private/strong segmentation and low coverage (except Colombia)	Low social security, high public, low coverage
Character of the Social protection regime	Assistance	Uncertain tendency to universalisation	Assistance	Tension between universalization and assistance
Inequality	No reduction	Strong reduction	Moderate reduction	Strong reduction
Poverty	Growth	Strong reduction	Moderate reduction	Strong reduction

Source: Own elaboration

3. Concluding remarks: The frailties of the diverse types of capitalism

51As we discussed in the introduction, the four ideal types of capitalism do not coincide with any particular country, as they are formalizations/stylizations of the trajectory of a country or of a group of countries. The references to a particular country have served to stress a specific characteristic of the type of capitalism. All real countries are, in fact, hybrids of the ideal types. In this manner, when we analyze particular countries we are able to stress some of the weaknesses of the ideal types.

52The weakness of both rentier types is evident: it is their dependence on the demand and price of the products they export, which are extremely volatile and subject to external economic conditions. The liberal rentier capitalism depends on economic growth to generate jobs, increase wages and consumption, and reduce poverty; in this manner, when it does not grow sufficiently, it accumulates redistributive demands that at one point or another may result in an open social or political crisis. The redistributive rentier capitalism is even more unstable, because the decline of resources coming from exports of commodities may lead to a political crisis. As the political support to the government depends on redistribution through corporatist or clientelist mechanisms, it is prone to a double crisis: economic and political, as we can at present witness in Venezuela.

53The fragility of the international outsourcing model is also its dependence on the external economy. The case of Mexico is extremely close to this model, although it is also partly a rentier country, as government finances depend heavily on crude oil exports. The Mexican economy is a platform that integrates labor at the last stages, and lowest aggregate value, of the manufacturing process. This implies that its level of competitivity depends on low labor costs (both wages and indirect social security costs), high flexibility of the labor market, low taxes and low environmental regulations. This results in the repression of internal demand, which together with low productivity gains results in low job creation, an increase of poverty, and a social protection system that is merely a safety net that assists the population unable to enter the labor market. This situation entails the accumulation of demands, a permanent or rampant crisis, especially in a large country such as Mexico, where part of the population suffering from low job creation and low salaries is prone to all sorts of illegal, and even criminal, activities.

54Finally, as can be exemplified best by the Brazilian case, socio-developmentism did not achieve developing the country with social equity, as it de-industrialized prematurely and culminated in a profound economic and political crisis (Salama, 2012).

55This does not prove that this model is unsustainable, but rather that a hybrid between a socio-developmentist and a rentier model such as the one adopted by Brazil and Argentina is unsustainable. In fact, Brazil shared many of the characteristics of the rentier model: increasing exports of commodities, overvaluation of the national currency due to the entry of foreign currency resulting from increased exports and foreign investments, and an increase of imports (Bresser-Pereira and Gala, 2012 and 2015). This situation, accompanied by wage-led growth, resulted in a domestic demand expansion considerably faster than that of the domestic supply, and thus a multiplier that benefited the external market and de-industrialization. (Bresser-Pereira and Gala, 2012 and 2015; Aguila and Lo Vuolo, 2016; Salama, 2012).

56The political economy of the rentier character of the Brazilian economy resulted from a contradictory compromise of the State with both the working class and the commodities and financial interests; a situation which also exists in the other redistributive capitalisms and explains their fragility. While the former benefited from redistribution, the latter profited from investment in State bonds with very high interest rates and the rise of the exchange rate in favor of the real (Bresser-Pereira and Gala, 2012; 2015; Marques Pereira and Bruno, 2015; Salama, 2012). The political compromise broke up because of two factors : on the one hand, the reduction of the government’s capacity to continue redistribution as a consequence of the decrease of its resources; on the other hand, with respect to the financial interests, the need of the government to reduce the interest rate, to control the overvaluation of the real and the circulation of capital, led to the end of the government’s capacity of arbitration between the population’s revenues and the financial and commodity exporting interests. The institutional and political crisis that ensued from Lava Jato resulted in the breaking up of the coalition between the PT and the PMDB that sustained the Dilma presidency, and deepened the economic crisis. (Costa Pinto, 2017).