Introduction

1For late industrializers, their relationship to global capitalism is of crucial importance. Very generally, emerging economies have two options for their economic catch-up strategies: either opening up towards global capitalism or following a rather protectionist strategy focused on their domestic markets. While the issue of openness versus protection is one of the oldest discussions on economic development, it receives fresh attention in contemporary scholarship on Comparative Capitalism research. From this perspective, these two options relate to two ideal types of capitalism: dependent market economies (DME) versus state-permeated market economies (SME). Whereas DMEs focus on the attraction of foreign direct investments, mainly in Central and Eastern Europe, SMEs develop on the grounds of their large domestic markets, with China and India as leading examples.

2However, most economies do not fit neatly into one of these ideal typical models. They combine features of national control with substantial openness towards external capital. Brazil is one of the most interesting examples of such mixed cases. Due to its large domestic market it has better options for pursuing an SME strategy than many smaller emerging markets. At the same time, external capital has always played an important role in Brazilian economic development. Moreover, the question of external openness in Brazil is politically contested and varies over time. A phase of economic openness during the 1990s has been followed by a domestic focus during the 2002-2016 rule of the Labor Party (PT) which, in turn, is being dramatically reversed by the ensuing Temer government.

3This article sheds light on Brazil during the governments of the Labor Party. The episode is particularly interesting, because the PT pursued a consistent “social-democratic” approach over a period of more than ten years which shaped the institutional character of Brazilian capitalism. Macroeconomically and socially, this period has been relatively successful: Besides regular economic growth (as measured in GDP), capital formation (indicating economic development) and private consumption by households increased substantially (Fig. 1). Net national income per capita rose from little more than 2600 US$ in 2001 to almost 11.000 US$ in 2011, only to fall below 8000 US$ in 2015 (World Bank 2018). When the lower 60% of the population held only 19% of all domestic income, its share rose to 24.4% in 2015 – representing an increase of almost 28% from the perspective of these lower 60%.

4At the same time, however, Brazil has been depicted as a case of “neo-developmentalist neoliberalism” (Dierckx, 2015, p. 145), or as “liberal neo-developmentalism” (Ban, 2013). In order to move beyond these hybrid and vague labels, this contribution pursues a “Critical Comparative Capitalism (CCC)” perspective that combines some basic assumptions derived from the “Varieties of Capitalism (VoC)”-approach with core insights of Critical Political Economy. We identify three generations of Comparative Capitalism (CC) research within the last two decades (Nölke, 2017). Despite earlier and parallel developments (Jackson and Deeg, 2006), the “Varieties of Capitalism” (VoC) approach as formulated by Hall and Soskice (2001) clearly is the hallmark of the first CC generation. Inspired by microeconomics and rational choice institutionalism, it has constructed a binary juxtaposition of coordinated and liberal market economy types (CME/LME) which both constitute economic equilibria based on specific coordination mechanisms. Subsequent research has demonstrated many conceptual and empirical shortcomings of VoC (see Hancké et al., 2007; Hall and Gingerich, 2009; Bohle and Greskovits, 2012; Bruff et al., 2015). Although the immediate utility of the VoC approach for understanding emerging market capitalism remains limited, yet its conceptual instruments are still very influential for later CC applications. The shortcomings of the VoC research program have led to a flurry of extensions, in particular with a background in historical and sociological institutionalism (e.g. Amable, 2003; Jackson and Deeg, 2006; U. Becker, 2009; Schneider, 2013; Thelen, 2014; McNally, 2017). This “post-VoC” (Bruff et al., 2015, p. 34) research program (as a second generation) has particularly extended its geographical attention towards areas beyond the OECD. It has highlighted the importance of history and politics (instead of economic equilibria) for the emergence of capitalist institutions, of distributive struggles and inequality (not growth exclusively) and of the state for capitalist coordination. Still, the post-VoC agenda usually neither incorporates the demand side (unlike e.g. heterodox economics), nor the interaction of national capitalisms with each other. These shortcomings are important points of departure for a third wave of CC scholarship. This third generation is often also more critical about existing political structures and takes conflicts, power and income inequalities into account (Ebenau, 2015, p. 55-57). On the one side, these “Critical Comparative Capitalism” (CCC) scholars account for the interactions and problematic interdependencies between national varieties of capitalism. Thereby, they highlight unavoidable tensions within the global political economy (e.g., Nölke and Vliegenthart, 2009; Nölke et al., 2015; Nölke, 2016). On the other side, some CCC scholars find themselves in the tradition of the Regulation approach, Dependency theory and Post-Kaleckian Macroeconomics (Stockhammer et al., 2014; Baccaro and Pontusson, 2016; J. Becker and Jäger, 2013; J. Becker, 2014; Jessop, 2014; Gambarotti and Solari, 2014, Suau Arinci et al., 2015). Here, however, the focus is not only on supply-side institutions (and companies), but also on demand side institutions such as collective bargaining and unemployment insurance (and household indebtedness). With these conceptual extensions, especially its macro-economic and political dimensions as well as international scope, CCC moves in fact much closer to a regulationist research programme which ever since argued that the diversity of capitalisms cannot be explained exclusively from a firms' perspective but needs to take the institutional embeddedness of accumlation regimes and regulation modes more seriously (see Boyer, 1990; Boyer and Saillard, 2002; Amable, 2016).

5While CCC looks at the classical VoC institutional spheres and coordination mechanisms, it also takes power relations, distributional issues and the related demand side as well international economic system into account. The next section builds upon recent developments within Comparative Capitalism scholarship, in order to give a theoretical basis for the conceptual juxtaposition of the DME and SME models. The core of the article focuses on Brazil, by first highlighting advances towards state-permeated capitalism during the decade of Labor Party rule in Brazil, before then depicting the residual dependency of Brazilian capitalism and related weaknesses. The main contribution of the article is in the field of conceptual development. It proposes the notion of “residual dependency” in order to depict the substantial problems caused by remaining economic openness in the transition towards a more state-centered development model in large emerging economies. As we shall see, the core problem of residual dependency is more in the political, than in the economic realm.

1. Dependent versus state-permeated capitalism: the ideal types

6In order to demonstrate the usefulness of CCC for the analysis of emerging market capitalism, this section presents the two fundamental alternatives of dependent versus state-permeated capitalism. It aims to explain emerging market capitalism through a set of analytical tools from both recent CCC (domestic market demand, the mode of integration into the international economic system and the nature of the state) and “traditional” VoC approaches (institutional complementarities and overall coordination mechanism). Theoretically, these tools stem from abstract reasoning about different modes of social coordination and the way institutions shape the behaviour of firms and governments, thereby overcoming the preoccupation with economic policy analysis to explain different development trajectories. At the same time, CCC attempts not to deduce macroeconomic structures from the patterns of firm behaviour exclusively (see Boyer, 2005, p. 47). Empirically, these have been validated through matched comparisons for the first decade of the new millennium between China and India (for the SME model) and Poland, Hungary, the Czech Republic and Slovakia (for DMEs).

7With regard to the role of domestic market and the international insertion of national capitalisms, DMEs aim at full integration into the global economy. There are few measures for the protection of national companies and countries are very open for imports. The economic catch-up strategy builds on the attraction of foreign direct investments and the integration into global value chains, even if this implies a high dependence on the economic decision-making of external actors. This gives Western (or “Northern”) transnational capital a powerful position, especially for quick relocation of production to other economies. SMEs, in turn, use foreign direct investments and selective privatizations as modernizing instruments, but only as long as these do not undermine the preponderance of national capital.

8Unlike DMEs, SMEs are generally dominated by national capitalists and a state which is cooperating closely with these business elites. Because of the strong position of national capital, the state does not automatically have to give in to the demands of multinational enterprises but is able to negotiate conditions in exchange for access to its large domestic market. Yet, we do not deal with a “strong”, centralized state, a dominance of public enterprises and centralized economic planning. Rather, we speak of 'state permeation' because state-business alliances often follow their particular interests. This helps to pursue coherent national development strategies and domestic capital has a long-term strategy for its domestic investments, whereas transnational capital is mobile and short-term oriented. The foundation of these strategies are large domestic markets which are much more protected than in DMEs in order to nurture national companies. Financial integration is even more cautious: capital controls and the accumulation of large currency reserves are typical strategies to limit the impact of turbulences on global capital markets.

9DMEs show strong institutional complementarities (or compatibilities) between a system of corporate governance which is controlled by the headquarters of multinational corporations and a system of investment finance that is based on foreign direct investments and foreign-owned banks. Moreover, competitive advantages are based on fairly low labor costs that are a result of an industrial relations regime centering around the appeasement of skilled workers through company-based agreements (instead of sectoral or even national-wide agreements) and the transfer of innovations within multinational corporations.

10The symbiosis of national capital and public authorities and the selective stance towards foreign direct investments enables us to specify the central institutional complementarities of the SME model. Major SME companies are organized by national capital, owned by families and/or the state. This makes SME companies independent from volatile global capital markets and demands by international investors. Moreover, it frequently allows SME companies to enjoy public support (e.g. subsidized credit by state banks). A close cooperation between major companies and public authorities also impinges on industrial relations, as public regulations (and their selective implementation) often help keeping labour costs low. Finally, public interventions also shapes the transfer of innovations and competition policy. A soft protection of intellectual property rights facilitates reverse engineering and subsequent adaption of foreign technologies. At the same time, competition policies are geared towards the temporary protection of individual companies.

11The institutional coherence of capitalist models is crucially determined by their central mechanisms of coordination. In DMEs, these are dominated by the interfirm hierarchies of multinational corporations. The main firm decisions in these economies are not being taken in host countries but rather in TNC headquarters. State-permeated capitalism, in contrast, is dominated by a close cooperation between public and business actors within public-private alliances characterized by interpersonal reciprocity and loyalty. Strong reciprocal norms and, obviously, family ties form the backbone of a close-knit network of mutual obligations and trust that makes economic coordination highly effective.

12These features enable DMEs to be particularly competitive as an assembly platform for semi-standardized industrial goods with a fairly high level of technology. The competitive advantage of companies from state-permeated capitalism, in contrast, consists of serving the domestic market with products and services on a medium level of technology.

2. Advances towards state-permeated capitalism during the decade of Labor Party rule in Brazil

13This theoretical framework enables us to study in how far the administrations led by the Brazilian labor party have had been able to institutionalize a state-permeated form of capitalism. At the same time, it allows us to highlight where Brazil has remained wedded to the more open model of dependent capitalism that had dominated the economy under previous liberal governments (see section 5).

2.1. Domestic market and international insertion

14Compared to other Latin American countries, the large and growing domestic market, in particular mass production for private consumption, has been the backbone for the rise of many well-established Brazilian companies. Production for the domestic market – and the regional markets in Latin America – has historically given rise to a number of Brazilian multinational enterprises such as AmBev (beverages, now part of InBev), meatpacking companies (JBS, Marfrig and BRF), textile producers (e.g. Camargo Correia), cosmetics (Natura), pharamaceuticals (EMS), construction companies and the corresponding engineering bureaus (Norberto Odebrecht, Andrade Gutierrez), IT services (Stefanini, Bematech and TOTVS), autoparts (Weg and Lupatech) and air transport (VARIG and TAM). Among the large, but less internationally oriented companies, are also the major private banks, such as Bradesco, Banco do Brasil and Itaú Unibanco. As mentioned earlier, government support has been crucial for the emergence of these companies, e.g. by stabilizing corporate control, providing investment finance and mobilizing human capital. In some sectors, companies also enjoyed direct state support, such as in government procurement that was essential for the development of large business groups in construction and engineering, or restrictions on foreign carriers in air transport (Schneider, 2013, p. 51).

15After the macro-economic stabilization of the 1990s, the domestic market took over from the export sector as core engine of growth. Between 2003 and 2009, private consumption increased continuously (see Fig. 1). This increase was based on a stable labor market, the creation of new formal jobs, substantial social transfers and wage increases (Melleiro and Steinhilber, 2012, p. 219). Doubling of the minimum wage set the standards for a third of the employed population and provided the basis for the majority of wage negotiations and some important social security standards. Arguably, the increase of the minimum wage was more important than the famous “Bolsa Família” cash transfer program for poor households and the related social programs “Bolsa Escola” and Bolsa Alimentacao” which together only amounted to less than 1% of GDP (Casanova and Kassum, 2014, p. 30-31; Leubolt, 2015, p. 241-242). Brazil's basic social security and pensions system, however, certainly helped with stimulating domestic demand (Jaeger, 2013). Average real income has risen during the last years and has increased the demand for local products and services. According to the IMF, “consumption – both public and private – has accounted for almost all the growth in Brazil” (Ibid., 2013, p. 15).

16Brazil protects its home markets in order to make sure that the domestic industry benefits from increasing demand. It is a widely shared misconception that Brazil’s economic fortunes primarily are dependent on commodity exports in general, and to China in particular. In 2011, Brazil ranked lowest on the export share in GDP (12%) among the top 10 global economies, even after the similarly closed economies of Japan (15%) and the US (14%), and much behind export-led economies such as Mexico (32%) and Germany (50%): “Growth in Brazil is still largely made in Brazil” (Canuto and Schellekens, 2014, p. 3; see also Bizberg, 2014, p. 14). With an average applied tariff rate of 14.1% on non-agricultural goods in 2010, Brazil has a much higher level of protection than the major OECD economies (OECD, 2011). Moreover, in terms of trade-GDP ratio, Brazil is less open than all other Latin American economies (Gaitán and Boschi, 2015, p. 176). The protection of domestic industry has been a strong objective for the major initiative Plano Brasil Maior (“Greater Brazil Plan”) launched in 2011, including preferential treatment in procurement, more active use of anti-dumping and a decrease of import licenses (Arnold, 2011, p. 28; Casanova and Kassum, 2014, p. 39). A special tax on cars that have been built with less than two thirds of locally sourced components has led several foreign car manufacturers to open new plants in Brazil (Casanova and Kassum, 2014, p. 32).

2.2. Institutional spheres and complementarities1

18Besides retained earnings, domestic bank loans represent the most important source of finance for Brazilian companies (World Bank, 2014). However, while in DMEs and emerging countries such as Mexico foreign banks account for more than 75% of the domestic banking sector, they play a minor role in Brazil where they only hold about 20% of the total banking assets. This is much less than in DMEs, but still higher than in China and India (Deutsche Bank Research, 2011). Finally, while most ‘private’ banks are actually listed on local stock markets, their shares are controlled by public bodies.

20BNDES credit approximates 7.5% of the Brazilian GDP and is four times higher than loans by the World Bank (Lazzarini et al., 2012, p. 32). The objective behind BNDES lending is not to provide credit for small firms facing constraints on the private credit market, but to boost the investments of already profitable firms (Lazzarini et al., 2012). As a consequence, large enterprises receive 70 to 80% of BNDES loans (BNDES, 2012, p. 35). This is part of a ‘national champions’ strategy assisting Brazilian multinationals on the world market. The provision of credit is one arm of this strategy, the acquisition (or retaining) of minority stakes in these companies by its daughter BNDESPAR is another. This idea of “Leviathan as a minority investor” is arguably a distinctive feature of state capitalism in Brazil (Lazzarini and Musacchio, 2014).

21Apart from BNDES, state-controlled banks such as Banco do Brasil and Caixa Economica Federal, whose credit thus represent loans by the state, are important funders for domestic companies. Not only does this permit the state a certain 'grip' on national companies, it also gives borrowing companies a strategic advantage due to lower interest rates, longer credit maturities and fewer conditions for creditworthiness in contrast to global credit markets (Masiero and Caseiro, 2012). Another state lever for long-term investments in Brazil are large pension funds of state-controlled enterprises, such as Previ (linked to Banco do Brasil), Petros (Petrobrás) and Funcep (Caixa Economica Federal). The role of these private funds – with domestic assets approximating 20% of GDP – has become so massive that some observers speak of “pension fund developmentalism” in Brazil (Datz, 2013).

22Thus, compared to companies from liberal economies, Brazilian enterprises are well protected against short-term financial turbulences, as could be seen during the global financial crisis (Ocampo, 2009, p. 19). Domestic banks had no incentives for risky speculation because they earned well on high domestic interest rates.

2.3. Coordination mechanism

24In comparison to other Latin American economies with far more adverse-business-government relations (such as Argentina or Venezuela), the smoothness and reliability of economic decision-making in Brazil stands out. It is based on compromises and piecemeal changes, even radical course changes towards privatization, opening or the reestablishment of state dominance are less pronounced than in most neighbor countries (Busch, 2010, p. 155), at least before the ongoing reforms of the Temer administration. Particularly the Lula administrations have further cultivated the building of consensus and gradual reformism (Casanova and Kassum, 2014, p. 32). Strikingly, the informal cooperation between Brazilian business and government works better than cooperation between organized business associations and the state in other Latin American countries (Schneider, 2015b, p. 48-50). Cooperation between business and state, as well as other social forces, is supported by a developmentalist ideology that highlights common national interests (in contrast to class conflicts).

25These thick ties are, however, are not only beneficial but can also be a liability for economic development, as indicated by many episodes of corruption that are under judicial investigation. Corruption is not a unique to the PT administrations but has been present under Cardoso (Anderson, 2011, p. 4) and the Temer government as well (The Guardian, 2017). However, the often weak parliamentary majorities for the PT governments, together with the individualistic election system and private financing of electoral campaigns, made a certain degree of corruption 'necessary' in order to bring laws through the Congress. Legislators are generally important targets for business lobbying, given their often very powerful role on specific subjects (Schneider, 2013, p. 143-144). Governments thus have to offer strong incentives to convince legislators to support specific legislature and not use their veto power.

3. Residual dependency of Brazilian capitalism and related weaknesses

3.1. Domestic market and international insertion

29Alternatively, domestic demand could (temporarily) be stimulated by increasing private indebtedness. In Brazil, however, a process of modest banking inclusion – for example based on easier access to credit for poor households – also seems to be exhausted during the mid-2010s, due to rising interest rates (Serrano and Summa, 2015, p. 14-16). Easier access to consumer credit for lower and medium income groups has temporarily boosted demand for medium-tech industrial goods (such as fridges and television sets) that can be manufactured domestically during the early 2010s, but heavily increasing private indebtedness has halted this boost of domestic demand. In 2011, the consumer debt service ratio in Brazil has reached 23%, a significantly larger value than in other Latin American countries or even the US (IMF, 2012, p. 5). In the meantime, the negative consequences of this “financialization of social policy” (Lavinas, 2017), where social security benefits often served as collateral for credit to marginalized groups, have become more than obvious. Equally, mortgage credit, which strongly expanded after a series of subsidized mortgage schemes during the early 2010s, still has been kept within certain limits due to fiscal constraints and cannot expand much stronger (López Vicente and Serena Garralda, 2014, p. 21). Also company indebtedness has increased fairly strongly, leading an increase of total household and company indebtedness from 107% of GDP in 2007 to 139% of GDP in 2014 (FAZ, 2015b).

31However, this exhaustion of domestic demand-driven growth happened at the same time as export demand-driven growth was weakened as well, in particular due to decreasing demand from (and increasing competition by) China. Although Brazil's export dependency is far less than for the “extractivist” export-driven models such as Chile and Venezuela (Jäger et al., 2014) which are far more affected by decreasing demand from China, the decrease of external demand did have an additional effect. Yet, the small share of manufacturing production for export (about 20%) limits the negative effect of this external factor (Jenkins and Freitas Barbosa, 2012, p. 77). Comprehensive macro-economic analyses confirm that the slowdown of Brazilian growth in the 2011-2014 period has overwhelmingly been triggered by a sharp decline of domestic demand and not by lowern exports or external financial conditions (Serrano and Summa, 2015).

32Yet, because of the weakness of both international and domestic demand, growth rates are sluggish since 2011. However, the mid-2010s stagnation of the Brazilian (semi-)state-led model also has supply-side causes that relate to the haphazard process to establish institutions that are necessary for a successful state-driven economy. Most notably, labor productivity stagnates because increased wages have not been met by higher capital accumulation, investments and improvements in education (Canuto and Schellekens, 2014, p. 3-4). Again, financialization plays a negative role, in particular where companies prefer generating income from financial income instead of investing for increasing productivity (Jibril et al., 2018). Correspondingly, domestic manufacturing industry is continuously losing weight vis-à-vis large-scale agricultural exporters as well as services (e.g. finance), thereby weakening the political case for a massive state-led industrialization strategy (Bresser-Pereira, 2015, p. 6-7).

3.1. Institutional spheres and complementarities

33In the following, we try to sketch how recent developments in the currency and capital regime shake the institutional complementarity between corporate governance and investment finance, thereby unleashing what we call residual dependency. Particularly since the privatizations of the 1990s, foreign multinationals are bringing a certain measure of volatility and external dependency into play. 20 out of the 50 biggest companies in Brazil are foreign-owned. Still, these companies are less relevant than in more liberal Latin American countries such as Chile, or even the dependent market economies of East Central Europe (Nölke and Vliegenthart, 2009). As Brazil is far more open to foreign ownership than e.g. China or India (see Nölke et al., 2019), this increases the potential contestation of economic policy-making. Since the 1990s, MNCs have drastically increased their prominence in the Brazilian economy in terms of sales, share of manufacturing or revenues (Schneider, 2013, p. 76). This had the effect “to box domestic firms out of most dynamic manufacturing sectors like electronics and automobiles” (Schneider, 2013, p. 82), thereby leading to a suboptimal sharing of tasks between national and foreign capital in manufacturing.

34However, the capital controls introduced in 2009-2011, aimed at limiting the influence of FDI, are “half-hearted” and very limited in their impact (Dierckx, 2015, p. 154). Moreover, in order to remedy an increasing current account deficit and decreasing capital inflows, these capital controls have gradually been lifted since December 2011. In line with the semi-closed nature of the Brazilian economy, many non-financial companies have made increasing use of corporate bonds, including foreign currency bonds, in order to escape the limitations of the state-controlled Brazilian credit system. By 2014, nearly 20% of total corporate debt financing were issued in bonds (IIF, 2015, p. 2). In addition, many companies turned to a multitude of short-term alternative assets (Kaltenbrunner and Painceira, 2015). With it comes a certain degree of market-mediated vulnerability, in particular when interest rates in the North start to rise again after their low level in the aftermath of the subprime crisis. These currency developments are very detrimental to the investment process in manufacturing industry as “sustained periods of exchange rate appreciation, interrupted by sudden and large depreciations, can have detrimental effects on the tradeable goods industrial sector, which may face competitive pressures and difficulties in forming long-term expectations” (Kaltenbrunner and Paincera, 2015, p. 1301). A more comprehensive developmental state strategy – with lower interest rates, a lower exchange rate and a more progressive taxation – would require far more thorough capital and exchange controls as well as limits of foreign investors on domestic financial markets, in order to prevent massive capital flight and to reduce the problem of sudden stops (Dierckx, 2015, p. 155; Kaltenbrunner and Painceira, 2015, p. 1300).

3.2. Coordination mechanism

36Although some industry associations temporarily have supported the labor party governments, this has proven to be an unstable alliance, particularly under economic stress. Under these conditions, the tension between the issues pursued by the labor party – such as rising wages, higher taxation to increase public expenditures, an increased role of public banks and political guidance on investment – and the interests of parts of industry became more obvious. Traditionally, the labor party is not part of the densely-knit network between domestic family capitalists and the state. The rise of the labor party thus somewhat complicated the established coordination processes in the Brazilian political economy, even if fractions of the party have become part of the establishment during the PT administrations.

37While Brazil shares some important state-business-coordination features with other SMEs, it thus features some special properties that make economic coordination more contested. These features include (1) the role of foreign capital as a “third” party, (2) the divisions between domestic capitalist class fractions as well as between these fractions and the new middle class, (3) the working of Congress based on individualism and local constituencies and (4) the additional fragmentation brought about by the rise of the labor party and its allies.

38(1) State-led capitalisms are characterized by the close coordination between representatives of the state (or state-owned enterprises) and domestic capitalists. In Brazil, however, there are traditionally three parties involved. The third party in this tripé (Evans, 1979) is foreign capital – and the domestic capital fractions that are closely linked to it. In Brazil, the role of this third party fluctuates. Whereas the military government made sure that trilateral cooperation serves the needs of national economic development, the liberal reforms of the Cardoso era encouraged the ”fusion of the local bourgeoisie with international capital” (Abu-el-Haj, 2007, p. 92) and the integration of Brazilian capital in global capitalist networks (Flynn, 2007). While the Lula/Dilma governments tried to strengthen the role of domestic authorities and private capitalists, they were unable to completely reverse the expanded role of internationalized capital. The majority of the population considers a prominent role of foreign MNCs as a normal feature of economic life (Schneider, 2013, p. 182). Correspondingly, economic coordination has to take the voice of international capital into account. This leads to a fragmentation of decision-making between business and the state in Brazil and to a more short-term perspective on the side of business, since MNCs - in contrast to many domestic capitalists – “are less interested in national development strategies and more interventionist states” (Schneider, 2013, p. 86).

39(2) The pursuit of coherent national development strategies is also complicated by the various internal divisions in the Brazilian capitalist class and the difficult relations of the latter to the lower middle classes. Quarrels within the Brazilian bourgeoisie already were a pretext to military interventions in the 1960s (Faucher, 1981, p. 13). The Cardoso governments, however, were temporarily able to unify the Brazilian bourgeoisie (Schmalz, 2012, p. 268). This unity has eroded steadily since then. Today, the main dividing line within the Brazilian bourgeoisie on economic policies is between outward-looking factions (agribusiness, allies of foreign capital, rentiers and the financial sector) and factions with a focus on the domestic market (domestic industrialists, trade sector) on the other (Diniz and Bresser Pereira, 2012, p. 195, p. 201). The main battleground between the latter is on macroeconomic policy issues, with the former, more powerful faction preferring a highly valued currency and high interest rates, due to their preference for financial over industrial investments (Bresser-Pereira, 2015, p. 6-7; Dierckx, 2015, p. 147; Leubolt, 2015, p. 199-204). These conflicts, however, are being played out in a very civilized way, in contrast to the more aggressive conflicts between the upper/higher middle classes on the side and the new/lower middle class and the PT government on the other (Anderson, 2011; Saad Filho, 2014; Nowak, 2015, p. 87). During the mid-2010s, the former have taken these conflicts to the streets, massively supported by the mainstream private media.

40(3) Not only the existence of class fractions with very different economic preferences complicates economic decision-making, but also the nature of the Brazilian political system, which is not only fragmented between the federal and the state level, but also features some particularities in the party system. Members of Congress usually are not bound to a major political party, but rather follow individual career incentives. Pork barrel politics are usually more important than party coherence and financial donations by individual companies are crucial for financing elections campaigns (Bazuchi et al., 2013). As an illustration, more than a third of the deputies in Congress switched parties during Lula’s first term and more than a quarter of both Congress and Senate were involved in judicial procedures by the end of his second term (Anderson, 2011, p. 4). Within this difficult legislative system, the Lula and Dilma administrations were only able to pursue their reform agenda in a piecemeal way (Schmalz, 2012, p. 265). This problem has become worse since the 2014 election that resulted in the record number of 28 parties in the Brazilian National Congress. Many powerful positions are occupied by oppositional politicians, many important economic reform projects are therefore blocked in Congress. Thus, disaffection with coalitional presidentialism runs deep in Brazil (Chaisty et al., 2015). Arguably, certain forms of corruption are “necessary” for an effective economic coordination in this context. The mensalão scandal thus was symptomatic for the constant need of the government to mobilize sufficient support in Congress. Similarly, donations to the PT to finance a massive election campaign against heavy opposition by large shares of the upper middle classes and the mainstream mediawere an important element of the 2015 Petrobras scandal.

Conclusions

43In contrast to China and India, Brazil deviates significantly from the model of a state-permeated economy. While it shares a focus on a large domestic market, some crucial institutional complementarities and a substantial ability for coordination between state and business, it still suffers from the remaining residual dependency on Western capitalism. Although Brazilian capitalism is much less controlled by Western (or Northern) multinational corporations than Central Eastern Europe, it still shares some common features and weaknesses that we have termed “residual dependency”.

44From the perspective developed in this contribution, Brazil moved into a promising direction towards macroeconomic autonomy during the rule of the Labour Party. However, PT governments were unable to fully mobilize the institutional complementarities inherent in state-permeated capitalism in order to make the Brazilian economy more 'self-sufficient'. They suffered from a limited ability to use the large domestic market in Brazil for the support of domestic Brazilian companies, from a still too open and volatile financial system, and from numerous frictions for domestic economic coordination due to the forceful presence of foreign multinationals in economic sectors that are key for productivity development. A move towards a domestically oriened accumulation regime has been initiated but insufficiently supported by a coherent mode of regulation, so to speak. Both the close personal alliances between state and business as well as the crucial state support for such an industrial strategy have not been strong enough to serve as an effective guide for economic action. As a result, firms and bureaucracies face uncertainty, leading to inadequate or even contradictory policy and management decisions.

45Given the favourable structural preconditions of the Brazilian economy with regard to the SME model and the beneficial economic policies pursued by the Labor Party governments, we need to conclude that the problem of residual dependency has to be addressed in order to develop a stable catch-up process for Brazil (and other countries that are stuck in a via media between autonomy and full dependency). Instead of the current return to the liberal economic model, Brazil should rather further pursue its path towards state-permeated capitalism. This would however require more thorough limitations on capital flows (more stringent capital controls), the imposition of a mechanism to combat the Dutch disease (for example through an export tax) and the mobilization of higher taxes on the rich, with the receipts to be used for investments into education, innovation and further expansion of domestic demand. Obviously, this would meet fierce resistance by those classes that benefit most from a return to a liberal form of capitalism: rentiers and export-oriented fractions of the Brazilian economy.

46The main problem of Brazilian capitalism today, thus, is political. A successful move towards state-permeated capitalism requires broad political support by the middle classes. The current political system in Brazil, particularly with regard to the status of the legislative, does not allow for a broad coalition for growth based on domestic consumption and production. It is therefore a misconception to assume that the solution would lie in ever more smart economic policies. The problem of residual dependency is not primarily a macroeconomic one that could be tackled by twisting knobs in the well-known repertoire of economic governance. Overcoming this political system is an important precondition for overcoming residual dependency in Brazilian capitalism.