In: Journal of Development Studies, Vol. 33, No. 3, pp. 364- 91.
Brazil is one of many developing countries that are struggling to upgrade their industries to attain international competitiveness. Sustained industrial competitiveness, this article argues, rests not only on firms' capabilities (microlevel) and a stable economic framework (macrolevel) but also and in particular on a tissue of supporting, sector specific, and specialised institutions and targeted policies (mesolevel) and on governance structures that facilitate problem-solving between State and societal actors (metalevel). The concept of systemic competitiveness seeks to address the interrelationship between the four levels. The article analyses various obstacles to achieving systemic competitiveness in the particular case of Brazil.
In the recent past, a new kind of policy intervention to shape industrial change has emerged in industrialised countries. Policy formulation takes place in policy networks that involve several actors, including those affected by the respective policy. This approach is based on the notion that the State no longer enjoys a superior knowledge base, or strategic capability, but rather depends on the knowledge, creativity, support, and commitment of societal actors to formulate and, most importantly, successfully implement policies. The purpose of this paper is to investigate whether this kind of approach is relevant for advanced developing countries. More specifically, I will look into the case of Brazil,(1) as this country is in the process of transition from an inward-oriented development model with little competitive pressure to a new model where the demands on firms will be much higher. In Brazil there is thus a need for active measures to support and shape the industrial adjustment process and at the same time there are political structures which complicate any coherent policy formulation, let alone one that is based on something as demanding as policy networks.
Transformation is under way not only in Eastern Europe and the former Soviet Union but also in a number of developing countries. After decades of inward-oriented development, based on the concept of import- substituting industrialisation (ISI), these countries are now facing the challenge of opening to the world market. This challenge has emerged since the inward-oriented development model had exhausted its growth potential and had led to an accumulation of economic and political problems. Major economic problems were the low competitive potential of the industry that had been built up in closed markets, the explosion of foreign debt due to the deterioration of the terms of trade for primary products and the low export potential in manufactures, and runaway inflation due to monopolised market structures and irresponsible public spending behaviour. Major political problems were the persistence of traditional corporatist structures that deterred the build-up of modern systems of interest representation, the accumulation of entitlements by social groups that were already well off, the persistence of clientelist political systems, and altogether a low capability to establish societal learning processes that would have facilitated adjustment to changing internal and external conditions.
Brazil is among the countries which face the problems mentioned above. After fifty years of an impressive growth of manufacturing value-added inside a sheltered market,(2) the country experienced economic crisis and stagnation in the 1980s. It was only towards the end of the decade that this crisis began to be understood as a fundamental crisis of the ISI model. In fact, the industrial development of the 1980s did not reflect any even partial process of adjustment to new framework conditions and instead mirrored, primarily for lack of an adequate adjustment policy, the continuation of traditional behavioural attitudes as long as possible. Brazil's industrial policy continued, until the end of the 1980s, to operate with the same incentives that had been introduced in the previous decades as a means of building new industries, namely, tax breaks, favourable credits, investment licenses, import facilities, and export-promotion programmes. The only difference was that now the goal was no longer to build up new industries but to make it possible for existing industries to acquire rents, for the discretionary policy measures entailed a variety of market entry and exit barriers and therefore generated effects not conducive to competitiveness [Frischtak and Atiyas, 1990].
It was therefore only in exceptional cases that a constellation emerged that could have permitted functioning competition and thus led systematically to an efficiency close to the internationally usual level. In many sectors every producer along the value-added chain could reject criticism of his lack of competitiveness by pointing to the inefficiencies of the production stages upstream from him. In fact, however, the case was such that inefficiencies accumulated along the value-added chain and the possibility of shifting the blame to others provided for a situation in which no link in the chain felt responsible for undertaking efforts to improve its own competitiveness.
In some sectors the firms were inefficient because they were far removed from the minimum-efficient plant size once the low level of selectivity shown by industrialisation policy had led to the emergence of a number of firms too large for the narrow domestic market (e.g. in the capital goods industry) [Fritsch and Franco, 1989]. In other sectors (above all, heavy industry) highly concentrated supplier structures were prevalent and prices were formed not in markets but between monopolist suppliers or between a cartel and state price-control agencies. State and multinational firms played an important role, particularly in heavy industry, capital goods, and consumer durables.(3) Moreover, spatial concentration continued to be an outstanding feature of Brazilian industrialisation industry continued to be largely concentrated in the southeastern states, although there was a certain process of decentralisation which benefited the interior of the state of Sao Paulo, Minas Gerais, and the southern states [Diniz, 1993].
A continuing structural aspect of Brazil's industry was its high degree of heterogeneity as concerns firm size, technology levels, productivity, and competitiveness [Ferraz, 1989]. There were in many industries individual firms that were close to the international best practice. Their existence indicated that import protection, monopolisation, and deterrent framework conditions did not completely discourage dynamic entrepreneurship. The decline of competitiveness did, however, indicate that this state of affairs cannot last forever: in an environment inimical to innovation and competition, firms may go through learning processes that pave their way to international competitiveness if the management is determined to do so. But the closer they come to the top international level, the thinner the air becomes to be specific: all the more difficult it will be to be competitive in view of a lack of suitable suppliers and, accordingly, with a high level of vertical integration; all the more costly will be the efforts required to acquire or to train a sufficiently qualified workforce; all the more expensive will be R&D, which, for lack of any technological network, has to be operated on an inhouse basis. In fact, even relatively competitive Brazilian firms were way behind international best practice firms when non-financial performance indicators like defect ratios, turnover of stocks, or customer satisfaction were applied [Sequeira, 1990].
Industry's lack of competitiveness was in the end the outcome of a number of deficits plaguing Brazil's variant of the import-substitution strategy [Frischtak and Atiyas, 1990; Moreira, 1993; UNIDO, 1992]:
-- there was no long-term strategy, but also no five-year plan ΰ la East Asia (exception: the Plano de Metas in the 1950s, which, however, was not very differentiated and established few clear-cut industrial priorities);
-- there were no clear-cut priorities and was very little selectivity in decisions relating to industrial policy, for which reason no specialisation profiles developed;
-- there was hardly any co-ordination between sectoral policies, so that the linkages provided for in the ISI concept frequently failed to materialise;
-- for national firms there were no performance criteria and policy's only instrument was the carrot; there was no stick (the only exceptions being the procurement policy of Petrobras, the state oil company, which subjected its suppliers to quality controls, and the BEFIEX export subsidy programme, which was based on quantitative export obligations);
-- there were a variety of barriers to market entry and exit, so that the competitive pressure and the risk of entrepreneurial failure were low in many industries, especially capital-intensive ones (above all basic materials, capital goods, and durable consumer goods);
-- there was no systematic pressure to achieve international competitiveness;
-- the stop-and-go economic policy discouraged at the company level any strategic orientation and sustained export activity;
-- industrialisation was for the most part not financed via domestic savings (exception: compulsory saving to finance BNDES) but funded directly over the state budget (inflation tax as a variant of compulsory saving; the use of government bonds to attract savings) and through foreign savings;
-- there were depending on the market power and the price policies of the state firms a variety of distortions of relative prices.
It was only in 1990 that the central government set out to establish a fundamental reverse in the development model, gradually opening the domestic market for imports, selling State firms, and trying to deregulate the economy. Despite the persisting, and occasionally increasing, political turmoil of the early 1990s the government has sustained this course. Import competition is stinging firms that are not familiar with competition, a substantial number of State firms have been sold (among them icons of the import-substitution era like the Usiminas and CSN steel mills), and a deregulation of the economy is albeit slowly under way. Moreover, political actors today more or less share the basic orientation towards competition and an open economy.
The basic aim of the new, open-market strategy is to force industry to become more competitive by increasing efficiency and quality, thus revitalising the domestic market and expanding manufactured exports. Yet, the Brazilian strategy is not entirely in line with the structural adjustment orthodoxy of the 1980s that solely emphasised budget balancing, deregulation, and foreign trade liberalisation. From the onset, Brazilian policy- makers tried to introduce both carrot and stick. The opening of the market was the stick, while industrial and technology policy measures were meant to be the carrot. However, the carrot either did not materialise, or failed to stimulate the appetite of the firms. Some policy measures aimed at giving financial support for firm-level restructuring were never implemented, or were only implemented after long delays and with little effect (like fiscal incentives for corporate R&D). Other programmes, like the programme for quality and productivity, had only a limited impact. On the whole, the liberalisation of imports was by far the most important policy measure. The only reason it did not lead to a wholesale extinction of industries, as had occurred in other countries, was the gradual reduction of protection that provided firms with some lead time to prepare themselves for the things to come.
In fact, Brazilian industry on the whole proved to be able to adjust to changing conditions, and more so than might have been expected, for instance in view of the deterioration of the export performance in the second half of the 1980s [Batista and Fritsch, 1994]. Many sectors, especially those that benefit from favourable static comparative advantages, adjusted easily.(4) The number of sectors that entered fundamental crises was small; most prominent among them were the electronics industry (including the largely uncompetitive informatics industry) and the oversized capital goods industry.
However, it is questionable whether Brazilian firms will be able to sustain their competitiveness if the build-up of competitive advantages is left to them alone. It is questionable for two reasons. First, the surge in competitiveness of the early 1990s was probably to a large extent due to the existence of pent-up rationalisation potential that the firms did not have to realise as long as the competitive pressure on the domestic market was low. Thus, firms could achieve leaps in productivity by relatively easy means, for example, by introducing basic techniques like quality management or detailed cost accounting [Fleury, 1995]. Continuing the improvements in productivity and quality may prove difficult in the near future. This may jeopardise the achievements so far since on the global scale competitive pressure is constantly increasing, particularly with more Latin American and Asian countries trying to increase their exports.
Second, it is hard to fail to notice that sustained competitiveness is due not only to firm-level efforts but rather to a dense interaction between firms, and between firms and their supporting environment. Approaches which have been labelled 'The competitive advantage of nations' [Porter, 1990], 'Structural competitiveness' [OECD, 1992], or 'Systemic competitiveness' [Esser et al., 1996] address this issue. Their common finding is that competitive advantages are man-made. Environments that sustainably support the competitiveness of firms are often deliberately shaped by he firms themselves, their associations, and the State. The way this shaping is organised in particular has been addressed by the concept of 'systemic competitiveness'. In the following section I present the key elements of the concept. In the third section I discuss the considerable obstacles that hinder the emergence of systemic competitiveness in Brazil, while the final section considers why it may be emerging nevertheless.
2 NEW GOVERNANCE STRUCTURES TO FOSTER INDUSTRIAL COMPETITIVENESS GENERAL CONSIDERATIONS
On the global scale competitive pressure is rising. Some societies succeed in creating governance patterns that are capable of shaping an environment that stimulates and supports the firms' quest for competitiveness. I follow here the World Bank's rather broad definition of the term governance, as '... the manner in which power is exercised in the management of a country's economic and social resources for development' [World Bank, 1992: 1]. I refer to a pattern where State and societal actors are deliberately creating the conditions for successful industrial development as systemic competitiveness [Esser et al., 1996].
By using the term systemic I want to point out several factors. First, a firm will generally not become competitive on its own, that is without a supporting environment of suppliers and production-oriented services as well as the competitive pressure of local competitors [Porter, 1990]. Second, an environment that sustains competitiveness is rooted in the way a society is organising itself, that is in its general and specific institutions [Nelson, 1992]. Thus, systemic refers to externalities that often are deliberately created within specific governance structures.
Third, I maintain that the State has an important role to play in industrial development and restructuring. However, I take for granted that autocratic, hierarchical modes of governance are becoming obsolete. New forms of governance are emerging that are based on a new kind of interaction between State and societal actors, typically in horizontal networks [Mayntz, 1991].
Fourth, there are strong interrelationships between four different levels: the micro-, meso-, macro-, and metalevels. Mesolevel means the space between the microlevel of the firm and the macrolevel of the economy as a whole, that is organisations, institutions and policies that are specific to and necessary for certain segments of industry. The introduction of the meta-level refers to issues like the basic governance structure of a society and the key actors' ability to build a basic consensus and to formulate strategies.
We add, thus, two levels that were largely neglected in the framework of structural adjustment which dominated policy making form the 1980s, where macroeconomic reforms were perceived as not only a necessary but indeed a sufficient condition to stimulate microlevel restructuring, thus enhancing industrial competitiveness. Neoliberal theories of allocation and foreign trade emphasise the importance of properly functioning international markets for capital, technology, and products, and the optimal character of decentrally organised decision-making processes. They reject active, anticipatory and/or accompanying industrial and technology policies. It was only during the course of ill-fated adjustment programmes that the political dimension of structural adjustment received more attention [Haggard and Webb, 1993]. But again, emphasis was placed on the prospects of implementation and sequencing of macrolevel reforms rather than on mesolevel policies.
In fact, a number of developing countries succeeded within the framework of structural adjustment programmes in stabilising economic framework conditions, thus revealing the pivotal role played by the once neglected field of macroeconomic policy for the development of sustained competitiveness. But the expected reactivation of their economies often failed to materialise. The reason is that the support structures that distinguish competitive industrial sites had not been shaped and, to make things worse, in some cases important locational factors were further weakened (e.g. education and R&D) by the adjustment measures targeted on budget consolidation [Klitgaard, 1991]. Thus, instead of industrial development being stimulated, the conditions necessary for sustained competitiveness were further undermined.
2.1 RISING DEMANDS ON GOVERNANCE ON THE MESOLEVEL
The competitiveness of firms rests, first of all, on their internal capabilities. What makes the difference between more and less successful capable firms is the extent to which their can rely on a supporting environment. As competitive pressure rises in most industries, firms have to concentrate on their core competence and thus rely increasingly on supporting firms and institutions. The effectiveness and efficiency of industrial sites, or, in economic terms, the density of externalities, i.e. the intensity of interaction between firms and with universities, training institutions, R&D facilities, technology information systems, private consultancies, trading companies, financing institutions, and the like [Porter, 1990], becomes increasingly important. The demands on the local, regional and national level to create and support the business environment tend to grow; this applies to demands on business associations and other non-governmental actors as well as to demands on the State on all these levels.
Developed countries and NICs improved their position in the international economic hierarchy if they succeeded in structuring the institutional setting of industry specific services. Their experience has shown that a State which acts competently can correct market failures and shape the supply side of industry on the basis of:
-- a specific industry and trade policy which links support and protection for companies to clear performance targets;
-- a technology policy which encourages companies to engage in R&D and establishes technological institutes focusing not on international leading-edge developments but the most significant bottlenecks in the national technological system;
-- labour safety legislation and suitable framework conditions for the self-regulation of industrial relations to support the establishment of modern structures in companies; and
-- an education policy which provides both a broad basic education and further technical and scientific training tailored to the needs of society [Hillebrand, 1991; OECD, 1992].
Mesopolicies, in particular the development of a material and non-material industrial infrastructure, should focus on a number of main areas (concentration of existing resources) to speed up the process of world market-oriented specialisation. Michael Porter [1990: 78] comments on the significance of developing specialisation advantages as follows:
'Contrary to conventional wisdom, simply having a general work force that is high school or even college educated represents no competitive advantage in modern international competition. To support competitive advantage, a factor must be highly specialised to an industry's particular needs a scientific institute specialised in optics, a pool of venture capital to fund software companies. These factors are more scarce, more difficult for foreign competitors to imitate and they require sustained investment to create.'
Therefore, apart from and beyond general innovation-friendly framework conditions, the development of dynamic competitive advantages requires specific, selective mesopolicies. Going beyond 'generic' policies, selectivity in the mesodimension instead of the widespread pork-barrel approach to promotion aims at strengthening the strong with an eye to building, as rapidly as possible, dynamic industrial centres and efficient industrial sites, even though this may imply a certain degree of neglect of less developed areas.
2.2 GOVERNANCE FOR INDUSTRIAL COMPETITIVENESS: DEMANDS ON THE METALEVEL
The control and governance capacity of government and collective problem-solving arrangements, that is, well-developed structures on the metalevel [Esser et al., 1996], are a crucial precondition for optimising performance potentials at the micro-, macro-, and mesolevels. Until recently most developing (as well as, of course, socialist) countries were characterised by centralised political decision-making processes and a bureaucratic, inefficient government apparatus with a low level of governance capacity. Often, this was even overlaid with rentist-corporatist structures which allowed privileged groups effectively to realise their particularist interests [Kaufman, 1990; Cavarozzi, 1992]. These power structures corresponded with forms of social disintegration and fragmentation which were characterised by the exclusion of broad segments of the population as well as by political and social polarisation. Economic modernisation and the development of systemic competitiveness cannot succeed in the context of such social structures.
The structural adjustment programmes of the 1980s did not take into account that developing countries are by definition characterised by weak markets and weak firms, an omnipresent and at the same time weak government, and weak societal actors [Klitgaard, 1991]. The tendencies toward social disintegration are further exacerbated if macroeconomic reforms fail to establish regulatory and governance capacities (government reform, formation of complex linkages between strategic actors) and the requisite social structures. Systemic competitiveness cannot emerge without social integration. Building systemic competitiveness is thus a social transformation venture that goes far beyond correcting macroeconomic framework conditions.
Against this background, a basic social consensus on the direction of the changes aimed at is crucial to a reorientation. However, societies cannot choose directions randomly; the key actors have to accept the world market as a framework of reference. This does not necessarily imply a high export ratio; it rather implies that firms should aim to get close to international quality and efficiency standards. Moreover, medium-term orientations and visions are important for asserting future interests against current interests and generating stable expectations. If the effort to develop them fails, the required structural change will be deferred, as was the case in many Latin American countries in the 1980s, thereby prolonging the process of social disintegration; and weak societal actors who are unable to articulate their interests adequately will pay the bill. To overcome obstructive social structures, durable patterns of social organisation and values which the societal actors share in terms of concerted action and co-operative approaches to problem-solving, are needed in the medium term.
The process of social structural change involves safeguarding the autonomy of social institutions and organizations from encroachments on the part of government. The de-linking between State and trade unions, industrial associations, the scientific community and other societal actors establishes a crucial precondition for a profound internal restructuring of societal organizations from top-down to bottom-up. Restructuring organizations and newly-created intermediary institutions proceed along the lines of three complementary logics. First, they optimise their institutions or firms (inward-looking orientation) on their own responsibility. Second, they represent their interests vis-ΰ-vis government or other societal actors (competition). Third, they shape their own environments through co-operation and networking with public or private institutions (co- operative competition). Increasing social self-organisational competence on the one hand, and clustering and channelling creativity potentials on the other, are complementary tasks.
Major groups in society must learn that safeguarding government from influential, privileged groups may establish a positive-sum game. Only a relatively autonomous government is able to orient its activities toward overall social and economic interests. Transparency and accountability are crucial. Autonomous functional subsystems are based on a clear-cut separation of government, industry, and societal actors. They may then be further developed by intrinsic learning processes, flexibility and responsiveness, and by dialogue and efforts to search co-operatively for all optimal solutions involving government and societal actors. This may occur on the national as well as on the regional and local level. In fact, traditional governance structures often implied a high level of centralisation of political decision-making. Redefining the role of the State may include a revision of the division of responsibilities between central, regional and local governments, and re-distributing tasks and funds to the latter. Systemic competitiveness requires fundamental changes in the national environment, especially in fields like trade policy and exchange rate policy which are crucial in determining the incentives that economic actors face. Yet, it will often be created on the local level where the actors involved find it easier to identify common interests and to devise measures to improve both firm competitiveness and social welfare.
2.3 NEW FORMS OF GOVERNANCE
Although the dogma which stipulates that government is obliged to assume a strictly subsidiary role vis- ΰ-vis market processes is inadequate, the neo-liberal critique of the traditional ways of government intervention is basically correct. The idea that government alone, as a kind of central control centre of a society, can selectively direct technological and economic processes presupposes that government bureaucrats are more capable and better informed than other actors in society, including firms. This has been the case in some latecomer countries, most impressively in East Asia [Cumings, 1984]. However, as societies get ever more differentiated, and firms as well as other actors undergo learning processes, it becomes the other way around in OECD countries as well as industrially advanced developing countries. Moreover, top-down approaches are unsuitable in the sphere of industrial location policy and the development of mesopolicies because the capacity to act towards a certain goal, the know-how requisite to formulate long-term policies, and the implementation capacities are distributed across a variety of governmental, private, and intermediary agencies. One-dimensional, etatist and centralist patterns of governance are doomed to failure when the development and support of complex entrepreneurial networks and specialised institutional landscapes are called for [Best, 1990].
Still, the conclusion that the State has no role to play at all is not well founded since this proposition ignores the indisputable fact that new forms of governance have emerged, initially in a number of OECD countries where government policy no longer follows the pattern of a traditional interventionist State. Rather, government acts as a co-ordinator, moderator and communicator in policy networks with firms and their associations, science, intermediary institutions, and trade unions. It aims at collecting and disseminating relevant information and working out joint medium- and long-term visions that can serve as points of reference for government mesopolicies as well as private initiatives. This makes it possible to relieve the government's burden by shifting decision-making processes into intermediary arenas, to ensure a higher degree of information availability, to heighten the legitimacy of government decisions, and to mobilise the creativity available among societal actors by involving them and their respective problem-solving capacities in a strategic fashion. They do, however, presuppose on the part of societal actors a capacity to compromise, to perform and learn, and to accept transformation [Scharpf, 1991; Mayntz, 1991; Heritier, 1993; Atkinson and Coleman, 1989]. These new location policy strategies, which have emerged in a number of European [Juergens and Krumbein, 1991; Abromeit and Juergens, 1992; Cooke and Morgan, 1993] as well as US regions [Sabel, 1993], differ fundamentally from the top-down approaches of traditional industrial policy, industrial planning, or investment guidance.
This is also true for late industrialisers, like Japan where MITI had to redefine its role after the completion of industrial catch-up [Vestal, 1993], or like South Korea where the role of the State and the type of interaction between the State bureaucracy and the private sector is undergoing profound changes. During the phase of catch-up industrialisation in Korea the State clearly played a guiding role, often ruling into the firms even when it came to day-to-day decisions [Hillebrand, 1991]. This was based on the bureaucrats' superior planning and evaluation capacity. Over time, however, private firms (especially the big conglomerates) have increasingly developed a capability for strategic planning. A painful process started recently with bureaucrats resisting giving private actors more autonomy in their decision making, let alone a bigger role in strategy formulation. Fierce struggles between private firms and the bureaucracy have emerged.(5)
Thus, aside from the forms of governance already prevalent in societies organised along the lines of market economies hierarchic co-ordination and steering in firms and public institutions, market-like co-ordination among firms, and hierarchic governance of society by government network-like forms of organisation are emerging [Powell, 1990]. This is happening in particular at the mesolevel, where the emerging forms are characterised neither by simple market allocation (competition and price) nor by centralist governance mechanisms (hierarchic control and State interventionism).
The predominant discussion in the 1980s of market vs. government overlooked these innovative forms that were involved in the shaping of social structures. They are based on a combination of market, government and a variety of forms of self-coordination, these operating in the shadow of the market, the shadow of hierarchies, and in self-organising networks. This view of increasingly differentiated forms of social organisation and governance surmounts the classical dichotomies of market versus government and of total autonomy of decentral actors (liberalism) versus totally integrated society (socialism).
Successful policy networks are based on six core elements [Mayntz, 1991: 16]:
-- autonomous collective actors capable of internal conflict resolution;
-- trust and commitment to fair exchange;
-- orientation towards a substantial outcome (problem-solving, beyond minimum consensus);
-- joint decision-making based on information-sharing;
-- reciprocity, or a just distribution of the costs and benefits of a joint decision (or a given problem solution); and
-- voluntary restriction of each actor's freedom of action because it is accepted that each actor has a legitimate claim that his interests be respected.
Policy networks differ from traditional corporatism in that the role of the State has changed: rather than the State's organising private interest and arbitrating between corporatist groups which hardly interact among each other, representatives of associations interact with each other and with State bureaucrats on an equal basis. Policy networks are also different from the 1970s brand of European neo-corporatism which basically involved the central State government and the peak organisations of capital and labour. Yet it is difficult to find a profound difference between 'mesocorporatism' and policy networks: both terms describe arrangements that used to have a certain level of institutional (albeit often informal) stability and a set of (albeit often tacit) rules that govern the interaction process.
Policy networks tend to be organised on a sectoral basis, dealing with fields like science policy, technology policy, or health policy; and they tend to be embedded in political structures where there is some higher level that may intervene in case a policy network runs astray. The State can, for instance, stimulate the build-up of local or regional policy networks that set out to formulate an industrial strategy. It can make sense to support such efforts financially as long as this support is linked to performance criteria. On other levels, for instance a network for technology policy on the national level, the threat that the State might unilaterally devise measures that run counter to the interests of the parties involved is an important stimulus for the proper functioning of policy networks.
The conditions for establishing this new form of governance for an effective mesopolicy are difficult in many countries. More often than not, polarisation between societal actors is very pronounced; there is no experience in communication and interaction between private and public actors; and corporatist structures with a rent-seeking orientation block any attempt of joint problem solving, that is policies that go beyond the lowest common denominator. Therefore, structural change towards competitiveness based on a new model of governance can only take place stage by stage. Experience shows that, in the long term, more complex forms of interaction and intervention at mesolevel are successful only if radical change on the basis of sound macropolicies has already been introduced. Social frustration can only be avoided if key actors explain the complex nature of the change to society, something that was not done in Latin America until the late 1980s. On the other hand, various experiences show that overcoming minimalism is the basic prerequisite for sustained competitiveness and the improvement of the welfare of the population.
3 SYSTEMIC COMPETITIVENESS: WHAT PERSPECTIVE FOR BRAZIL?
At present, the governance capability appears severely limited in Brazil. Therefore, it would be out of place to formulate a broad set of policy measures that would have to be implemented to support industry in its struggle for more competitiveness. What is even more dramatic is that, at first sight, a total lack of systemic competitiveness seems to mark the situation in Brazil. It appears that not just a few factors are missing but rather a number of factors interact in inhibiting the emergence of systemic competitiveness:
-- the disarticulation between the executive and the legislative branch of central government which complicates the formulation of macro- and mesopolicies;
-- the persistence of traditional corporatist structures which curtails the development of collective actors which are firmly rooted in their respective constituencies;
-- the mutual mistrust between State and societal actors which makes joint problem-solving in policy networks difficult;
-- the absence of anything like a national system of innovation, which establishes a major deficit at the mesolevel;
-- the high level of disorganisation in the public bureaucracy; and
-- the low propensity to develop creativity in policy-making on the local and State level.
3.1 THE DISARTICULATION BETWEEN THE EXECUTIVE AND THE LEGISLATIVE BRANCH OF CENTRAL GOVERNMENT
The president appears to be the central figure in Brazil's political system. He is popularly elected, and presidential elections do not necessarily coincide with parliamentary elections. Lamounier  has described the system as a plebiscitarian presidency where the president receives a mandate to pursue fundamental political changes, and the electorate expects him to implement them without delay. However, the Brazilian political system has accumulated various mechanisms of checks and balances that curtail the president's capacity to govern autonomously. In the 'Old Republic' (before 1930), the leaders of some federal States (Sao Paulo, Minas Gerais, Rio Grande do Sul) were powerful enough to challenge the president, and the situation has changed only partially since then. After the end of Getulio Vargas' dictatorship in 1945, the congress rapidly dismantled certain institutions that had contributed to strengthening the executive power [Skidmore, 1967: 34 ff]. The military had always played an important role in politics, even in the democratic phase from 1945 to 1964, and it did not really come as a surprise that it toppled the civilian president in 1964. But even the military presidents saw their governing autonomy circumscribed by the necessity to mediate between conflicting factions of the military and to get political support from regional oligarchies [Skidmore, 1988]. After the smooth transition to civilian rule in 1985, the congress again succeeded in extending its responsibilities. Therefore, the president has to strike bargains with the congress, the governors of federal States and the military in order to get his policies implemented.
Over the recent years this has proved to be particularly difficult with regard to the congress. In fact, a system that was meant to establish checks and balances has largely led to a blockade of the political process. This is due to the distorted representative structure of the congress, the large number of parties and low party discipline, and the prevalence of clientelistic behaviour, all of which make it difficult for the executive to organise a reliable political base in the legislative.
In the congress, the North-eastern and Northern States are clearly overrepresented. This is most obvious in the Senate where each State is represented by three senators be it the State of Sao Paulo with half the industrial product and a fifth of the population, or Amapa with hardly any economic importance and population. But it is also true for the Chamber of Deputies [Rosenblatt and Novaes, 1993].
The political behaviour of congressmen from the Northeast and the North is largely shaped by their efforts to channel government funds to their local constituencies. Congressmen from the Southeast and South may take an interest in industrial and technological development. Yet, their legitimacy is also based on successful pork- barrelling. This has to do with the structure of the electoral system: There are open, State-wide lists, and the citizens can elect whichever candidate they find appropriate [Mainwearing, 1992]. Typically, this means that a candidate will campaign within a certain region inside of a given State in order to raise the necessary votes there. Therefore he is dependent on satisfying the local clientele.
This has two implications. Firstly, party coherence is low, as the parties do not have to set up ranked lists and most parties have a limited capacity to support their candidate. In fact, politicians frequently change parties or set up new ones. Consequently, there is no ex-ante-reason why congressmen should follow any fraction discipline. Secondly, the executive finds it difficult to build a reliable base in Congress. Even trying to build an ad hoc coalition between fractions to support a given policy initiative is often no viable option. Rather, the executive (or its 'leader' in Congress) has to find support from a sufficient number of groups of congressmen which are typically led by one of the leading political figures in Congress.
The mutual blockade between the executive and the legislative branch of government has been a crucial factor in delaying(6) the macroeconomic stabilisation of the Brazilian economy which is a key prerequisite for successful mesopolicies. The government literally had to buy support in congress, and the congressmen had to channel funds to their constituencies. Both factors contributed to the rising public deficit that became the single most important reason for the explosion of the inflation in the second half of the 1980s and the early 1990s.
3.2 THE PERSISTENCE OF TRADITIONAL CORPORATIST STRUCTURES
The internal reorganisation of interest groups which is an important pre-condition for enabling their representatives to make commitments in policy networks has only started. So far, the relationship between the State and major societal groups is largely organised in a way that makes it difficult for the State to forge alliances and build networks. In this respect, sorting political actors along lines like left versus right does not help much. The opening, deregulation and destatisation of the economy faced resistance on the left and on the right, and probably even more on the left than on the right. This has to do with the fact that the largest federation of trade unions, CUT, which may represent as much as 70 per cent of formal sector employees,(7) is particularly strong in sectors that represent the core of the traditional etatist development model, like Petrobras and the petrochemical industry, and in sectors that gained from the crisis of the 1980s; that is the public service and the banking sector, which both expanded enormously.(8) Therefore, despite its leftist rhetoric CUT is actually one of those forces that at least want to slow down opening and liberalisation and also have a somewhat ambiguous position regarding economic stabilisation. Also, CUT is in a difficult position regarding the structure of the system of representation itself. On the one hand, it would like to see the corporatist system, where 'Labour Courts' settle wage disputes, dismantled. On the other hand, CUT's member unions gladly accept the money they receive through the compulsory contributions of the workers (imposto sindical).
On the antagonistic side, industry has also found it difficult to develop clear political positions. This has to do with the fact that interests diverge vastly. Many firms hope to benefit from an opening of the market (especially in terms of cheaper and better quality inputs). Others oppose it fiercely. The structure of the system of business associations has not helped industry in defining a common position. The corporatist system which was introduced in the 1930s is still in place in this sphere as well. Within each municipality firms are compulsory members of the association of their branch (sindicato). All the sindicatos of a given State form the State Federation of Industries which again form the National Federation of Industries (CNI). CNI thus represents anybody from car manufacturers to hammock producers, which makes it correspondingly difficult to define a political standpoint on any given issue [Mathieu, 1991]. Moreover, since membership is compulsory there is little incentive for sindicatos and federations to take care of the interests of their members. In the end, their orientation and type of action depends on their respective presidents who may use the entity as a means of increasing their personal political influence, but who may also, provided they feel the mission, try to stimulate a modernisation process in the entity as well as the business community as a whole. Both kinds of behaviour have been observed in the recent past.(9) The orientation of any sindicato and federation depends to a high degree on the view of its president rather than being based on transparent internal processes of decision-making. This situation makes it difficult for the government to build reliable relationships with the business community: it can never be sure whose opinion and interests a 'business representative' actually represents.
So far, there have been only a few attempts to establish new business associations according to common interests, for example, PNBE (Pensamento Nacional de Bases Empresariais), a federation of modern small and medium-sized firms, and IEDI (Instituto Economico de Desenvolvimento Industrial), which was organised by some 40 leading private firms and has thus far served above all as a think tank. This is due to the fact that the costs are high, since their members have to pay both the compulsory contribution to the corporatist association and the fees of the new association, whereas benefits are uncertain and the temptation to free-ride is big.
3.3 MISTRUST BETWEEN FIRMS AND GOVERNMENT
Government-business relationships suffer from a high level of mistrust on the side of the latter. This is due to various experiences:
-- the change in the foreign trade regime that was unilaterally imposed by the government in 1990,
-- arbitrary and frequent changes in the more specific rules that govern the business environment (in taxation, regulation for foreign exchange or financial transactions, etc.) and cause high transaction costs;
-- arbitrary interventions in factor markets, e.g. regarding the level of real interest rates or nominal or real wages;
-- the limited and deteriorating technical capacity of the government bureaucracy which at times leads to unreasonable or peculiar policy measures;
-- the unpredictable outcome of the bargaining between government and legislative;
-- the low level of transparency of the budget process; and
Therefore, industry has been increasingly sceptical regarding any kind of government intervention, even towards those interventions that were meant to support the firms' adjustment.(10) This was one reason why policy initiatives that were meant to add the carrot to the stick by and large had a limited impact:
-- A programme on industrial competitiveness (PCI) announced in 1991 never got off the ground.
-- A programme for technological capability building (PCT) announced in 1990 was passed by Congress only in 1993, and its main instrument, fiscal incentives for R&D, had little effect in a situation where company profitability depended mainly on the details of the design of macroeconomic stabilisation efforts. Moreover, as firms rely to a large extent on creative accounting there is not much to be expected from fiscal incentives anyway.
-- A programme to raise consciousness regarding quality (PBQB) seems to have had some (albeit apparently limited) effect, mainly contributing to a trend that was already under way [PBQP, 1992: 17].
These programmes appear to suffer from having been formulated in an isolated way and based on the experiences of advanced industrial countries. Their formulation was not based on a broad consultation process with the prospective recipients of incentives; rather, it replicated the traditional top-down approach to industrial policy.
3.4 THE HIGH LEVEL OF DISORGANISATION IN THE PUBLIC BUREAUCRACY
The Brazilian public service has always been very heterogeneous in terms of effectiveness, efficiency, and accountability. The absence of transparent recruitment procedures, predictable career paths and personnel stability, and the political mechanism of patronage, reinforced each other, thus minimising the longer-term orientation and the personal commitment of the employees [Evans, 1989]. Certain sectoral agencies and institutions like the National Development Bank (BNDES) were the exception to the rule.
Things became worse after 1990 [Holanda, 1993]. The Collor administration tried to implement a public sector reform that amounted to little more than arbitrarily closing down some State agencies and firing employees. In fact, the government failed to achieve the latter since the constitution had established a job-for-life guarantee for all public servants employed for more than five years. What the government actually achieved was to cut the real wages, undermine the morale of the employees and stimulate many of the more capable among them to look for better paid, more stable jobs elsewhere (for instance, as professional staff of the Congress). Therefore, the technical capability of the public service to formulate, let alone implement, policies has decreased. This is particularly true at the middle level of the public service.
3.5 THE ABSENCE OF A NATIONAL SYSTEM OF INNOVATION
In the past, the prevailing incentive structure did not stimulate innovative behaviour, at least not in the way it is normally understood in industrialised countries [Meyer-Stamer, 1995]. This is not to say that Brazil was a stationary economy with no technical change at all; the contrary has been documented. However, the larger part of technical change followed the typical pattern of latecomer industrialisation where firms try to master technologies that have been developed elsewhere, try to adapt them to their needs and possibly try to improve them so that they fit better into the local environment. 'Typical R&D efforts would be determined by the need to use different raw materials, scale-down (to smaller) plant size, diversify the product mix, change the product design, use simpler, more universal, less automated and lower capacity machinery, stretch out the capacity of existing equipment, etc.' [Teitel, 1987: 109]. This kind of activity does not mainly require intra-company R&D but rather process engineering activities.
In the closed market environment, firms felt little pressure to extend their technological effort beyond this kind of activity [Dahlman and Frischtak, 1993]. With the degree of competition remaining low, they had little incentive to introduce truly innovative products or to look systematically for radical improvements in the production process. There were only a few exceptions, like the aircraft industry where the Brazilian firm tried to compete at the leading edge of one segment of the world market for commuter planes, or in oil exploration where the internationally available know-how on deep water exploration was apparently limited, or in the production of alcohol fuel, which simply did not exist anywhere else. In these cases, firms have set up a systematic R&D effort and have also co-operated with universities and research institutes. In most other industries, however, firms did not have R&D departments. They are an important feature in the organisation of R&D co-operation with external agents since they form the major recipient structure for the internalisation of research results from external sources. In fact, as a rule firms did not confront problems, challenges or opportunities that would have made it necessary or advisable to look for research results or external research support.
Thus, researchers in universities and research institutes had to have the impression that there was little potential demand for their results in industry. Yet, this was only one reason for the clear separation between research and industry. Other reasons were [Castro, 1989]:
-- the easy availability, particularly in the 1970s, of research funds from public financing institutions which gave the emerging scientific community a lot of space to define research priorities according to their personal interests;
-- the research ideals that PhDs returning from abroad brought with them, leading them to strive for academic reputation rather than application of their research results (they often also oriented their research so that it fitted into the line of activity of their foreign alma mater rather than domestic priorities); and
-- a certain degree of introspection that emerged inside the science system from the early 1970s, where communication mainly took place inside a circle made up of people from research institutes, universities and public funding agencies; reputation and career perspectives were dependent on the peer evaluation inside this circle, not on things like successful co-operative links with firms.
3.6 DEFICIENCIES IN POLICY-MAKING ON THE LOCAL AND STATE LEVEL
Industry tends to be clustered in Brazil. That is, at a reasonable level of disaggregation in terms of industry classification a substantial number of companies of a particular industrial branch appear to be located in a restricted geographical area, mostly within a part of a federal State. For instance, the production of women's shoes is concentrated in the Sinos valley in Rio Grande do Sul, the production of men's and children's shoes in the interior of the State of Sao Paulo (around Franca and Biriguei, respectively), the production of ceramic tiles in the south of Santa Catarina, and the processing of chicken meat in the west of Santa Catarina. That is not to mention the concentration of automotive and capital goods industry around the city of Sao Paulo (or, for specific political reasons, the consumer electronics industry in Manaus and, for technical as well as political reasons, the petrochemical industry in Cubatao, Camacari, and Triunfo). As has been shown in the case of the shoe industry in the Sinos Valley, the characteristic feature of clustering is not only the geographical concentration of firms but especially the emergence of a network of specialised suppliers, capital goods firms, and services, as well as a certain level of political organisation [Schmitz, 1993]. Therefore, in terms of mesopolicies, industry does not necessarily rely on the central government, and the disorganisation on the federal level may even stimulate joint action and policy initiatives on the local or regional level. Thus, systemic competitiveness may emerge on the local or regional level long before any such development on the national level.
Regional initiatives are facing two serious problems, however. First, policy approaches that seriously aim at problem-solving do not form part of the traditional role definition of either politicians and public servants or business associations on the local and regional levels. This is not only true for the less-developed North-eastern and Northern regions, but especially for the Southeast and the South. Local politics are strongly based on clientelism. Rivalry between competing local oligarchies is strong, and mutual trust tends to be low. Policy initiatives that are being launched by one key figure often face fierce resistance on the grounds that his rivals presume that the main goal of any given initiative is to advance political ambitions. This logic is one element which explains the extreme orientation towards approaches that produce immediately visible results. Any other approach that would require a certain degree of consultation between different political actors simply is not viable.
Second, as long as the division of responsibilities between central, regional, and local government is not clearly defined, policy initiatives on the regional or local level are under the permanent threat of being thwarted by central government interventions be it that the central government changes its mind and starts to pursue a competing initiative on its own, be it that the central government launches policies of some other kind that undermine the regional or local policy formulation potential, for example, by abruptly changing the rules of the game in foreign trade. Since political actors on the regional and local level know about these possibilities, it is probable that they are going to anticipate this in their activities, and that means that their initiatives are going to be very limited in their ambitions, in the profile they keep, and in terms of the financial resources involved.
4 WHY SYSTEMIC COMPETITIVENESS MAY EMERGE NEVERTHELESS
It seems that the prospects of systemic competitiveness emerging in Brazil, at least anytime soon, are bleak. However, there are also a few examples that may indicate that profound changes are actually under way. First of all, there is the voluntarist shock of the Collor government that pushed through an opening of the market against various political actors in 1990, thus changing profoundly the incentive structure for firms. Today, Brazilian industrial firms by and large operate under totally different conditions than in the six decades of import substitution before, having had to increase their competitiveness rapidly. Surprisingly, many have so far proved to be reasonably adaptive [Coutinho and Ferraz, 1994]. More precisely, they have started to adapt after a certain period of hope that they might be able to revert the political decision.
Second, there is the new type of approach to industrial policy emerging from the so-called sectoral chambers (camaras setoriais). In the Sarney era (1985-90), sectoral chambers existed as a forum for bargaining to control prices. The early Collor government tried to set up Executive Groups for Sectoral Policy (GEPS), but this attempt to articulate policy initiatives with private industry bore little fruit due to mutual distrust. In spring 1991, sectoral chambers were set up again. Initially, this was with the aim of containing inflation in a period of phasing- out the general price and wage freeze [Salgado, 1993]. The chamber for the car industry, however, turned eventually into a fully fledged attempt not only to restructure the industry but also to revitalise the whole economy. It was remarkably successful.
The chamber was made up of State representatives, the associations of the car and car parts industry as well as the dealers, and the apex organisations of the trade unions. The deal they struck aimed at dynamising the sector, mainly by reducing the price of cars (through tax reductions, rationalisation, and smaller profit margins) and stabilising industrial relations. Although this does not exactly qualify as an advanced version of a policy network (because the distribution of costs and benefits was uneven since it appeared as doubtful that industry and dealers would actually forego profits), it comes close to such an experience.
This was certainly facilitated by the fact that the sector is highly concentrated (in terms of companies as well as geography) so that the number of actors involved was limited, and the effects of certain measures were predictable; that is, the actual outcomes came close to the intended ones. It is therefore not really surprising that other sectoral chambers which dealt with more differentiated and dispersed industries have not been so successful. Nevertheless, given the past experience of conflict and mutual mistrust between all the actors involved, it was not necessarily to be expected that the venture would succeed in the car industry. It followed the pattern of a macroeconomic turnaround (opening of the market) which enforced microlevel adjustment, which in turn contributed to developing a basic consensus at metalevel, thus creating a crucial precondition for joint problem-solving at mesolevel.
Third, some federal States and city governments have recently shown a substantial capability for stabilising the local economic environment and stimulating local production activities, thus contradicting the traditional logic of action outlined above. This has been most prominent in the North-eastern State of Ceara where the State government (having fired a lot of redundant employees who had got their jobs due to the patronage system rather than personal qualification, and having rehabilitated public finance) started public purchasing programmes to simulate local industry [Amorim, 1994]. Another example is the city of Curitiba where a well-developed local infrastructure and a solid financial position serve as a base for local initiatives to stimulate new industrial activity, inter alia in fields like the computer software business.
Fourth, there is some evidence (albeit so far undocumented) that business associations are changing in profile, venturing to offer services to their member firms. In the State of Santa Catarina, for instance, the Federation of Industries has started to organise the participation of local firms in international fairs. It also engages in awareness building and information dissemination, on such matters as new manufacturing practices or ecological product requirements in the export market.(11)
The factors and experiences described above indicate that changes are under way in Brazil. Societal actors are experimenting with new kinds of governance, and the results seem to be encouraging. Nevertheless, it will take quite some time, not only for structures to change, especially to reshape the relationship between the executive and the legislative branch of government and to strengthen the technical capacity of the public service, but also for actors on different levels to learn how actually to use the space that opens for policy initiatives. It will also take some time, and a much effort and goodwill, to overcome the mutual mistrust among actors. It may happen that different trends reinforce each other in this regard: Changes in the electoral system may stimulate a greater coherence of political parties; this may make recruitment processes more transparent and predictable, thus reducing the suspicion among actors regarding their respective political ambitions. Changes in the political parties, again, should facilitate the construction of a reliable base of the executive in the Congress, thus making it easier to get beyond ad hoc deal-striking to the formulation of medium-term projects. A changing role definition by business associations and trade unions may facilitate the shift from clientelism and traditional corporatism to pluralist lobbying efforts and the set-up of policy networks.
Two main conclusions emerge regarding the relevance of new patterns of governance and the concept of systemic competitiveness for Brazil. First, the traditional, interventionist, hierarchical State is definitely dead and gone. The prospects for a comprehensive, Korean-style industrial policy that tackles various problems in several industries at the same time, remain dim. The single most important policy instrument since 1990 has been foreign trade policy. Whenever the government felt that a certain industry did not make a sufficient effort in restructuring, it reduced the respective customs tariffs in order to put more pressure on firms. Future initiatives to also support industrial restructuring will have to differ from the traditional top-down approach; they will have to be based on new role definitions by the actors involved. These will have to be more specific; and they should be based on the thorough separation of roles and tasks between different levels, that is the local, regional and national levels. Thus, the concept of systemic competitiveness signals an interesting option for Brazil to go beyond the alternative of traditional governance vs. free market capitalism without any collective action, in the development of competitive advantages.
Second, however, the transition from hierarchical to network-based governance is a difficult task, and it certainly is the opposite of a quick-fix to solve the shortcomings of a weak government. Shifting responsibility for policy- making into the society can only work if the government is sufficiently strong to prevent policy networks from running amok; that is, introducing discriminations, abolishing competition in the market, or redesigning clientelist links. If this can be avoided, policy networks on the mesolevel can help to overcome the problem of persistent overburdening of the government. Efforts to reform government may then focus on establishing a capacity to perform context control, that is supervising the functioning of policy networks, rather than trying to re-establish a capacity to guide other societal actors. These are demanding tasks, and it would be a big step forward if the Brazilian State could fulfil them.
Abromeit, H., and U. Juergens (eds.), 1992, Die politische Logik wirtschaftlichen Handelns, Berlin: Edition Sigma
Amorim, M.A., 1994, 'Lessons on demand', Technology Review, Jan.
Atkinson, M.M., and W.D. Coleman, 1989, 'Strong States and weak States: sectoral policy networks in advanced capitalist countries', British Journal of Political Science, Vol.19, pp.47-67
Batista, J.C, 1992, Debt and Adjustment Policies in Brazil, Boulder etc.: Westview
Batista, J. C., and W. Fritsch, 1994, 'Dinamica recente das exportacoes brasileiras (1979-90)', in J.P. dos Reis Velloso and W. Fritsch (eds.), A nova insercao internacional do Brasil (pp.49-82), Rio de Janeiro: Jose Olympio
Best, M.H., 1990, The New Competition. Institutions of Industrial Restructuring, Cambridge: Polity Press
Bielschowsky, R., 1991, 'Ideology and development: Brazil, 1930-1964', CEPAL Review, No.45, pp.145- 67
BNDES, 1988, Strategic Plan 1988/90, Rio de Janeiro
Brunner, H.P., 1993, The Recreation of Entrepreneurship in Eastern Europe: Neither Magic nor Mirage, Berlin: Wissenschaftszentrum
Castro, C.M, 1989, 'What is happening in Brazilian education', in E.L. Bacha, H.S. Klein (eds.), Social Change in Brazil, 1945-1985. The Incomplete Transition (pp.263-309), Albuquerque: University of New Mexico Press
Cavarozzi, M., 1992, 'Beyond transitions to democracy in Latin America', Journal of Latin American Studies, Vol.24, No.3, pp.665-84
Cooke, P., and K. Morgan, 1993, 'The network paradigm: new departures in corporate and regional development', Environment and Planning D: Society and Space, Vol.11, pp.543-64
Cumings, B, 1984, 'The origins and development of the Northeast Asian political economy: sectors, product cycles, and political consequences', International Organization, Vol.38, No.1, pp.1-40
Dahlman, C.J., and C.R. Frischtak, 1993, 'National systems supporting technical advance in industry: the Brazilian experience', in R.R. Nelson (ed.), National Innovation Systems: A Comparative Analysis (pp. 414- 50), New York, Oxford: Oxford University Press
Diniz, C.C., 1993, Competitividade industrial e desenvolvimento regional no Brasil, Estudo sobre a Competitividade da Industria Brasileira, Nota Tecnica Extra-Blocos Tematicos, Campinas (mimeo)
Esser, K. et al., 1993, Systemic competitiveness: New governance patterns for industrial development, London: Frank Cass
Evans, P.B., 1989, 'Predatory, developmental, and other apparatuses: a comparative political economy perspective on the Third World State', Sociological Forum, Vol.4, No.4, pp.561-87
Ferraz, J.C. 1989, A heterogeneidade tecnologica da industria brasileira: perspectivas e implicacoes para politica, Rio de Janeiro: UFRJ, Instituto de Economia Industrial
Fleury, A., 1995, 'Quality and productivity in the competitive strategies of Brazilian industrial enterprises', World Development, Vol.23, No.1
Frischtak, C. and I. Atiyas, 1990, Industrial regulatory policy and investment incentives in Brazil, Report No. 7843-BR, Washington: World Bank
Fritsch, W., and G.H.B. Franco, 1989, The quest for efficient industrialisation in a technologically dependent economy: the current Brazilian debate, paper presented to symposium on Competition and Economic Development, Paris, OECD, 17-18 Oct.
Furtado, C., 1984, 'Rescuing Brazil, Reversing Recession', Third World Quarterly, Vol.6, No.3, pp.604- 23
Goldin, I., 1990, 'Agriculture and economic crisis. Lessons from Brazil', OECD-Observer, No.161
Haggard, S. and S.B. Webb, 1993, 'What do we know about the political economy of economic policy reform?', The World Bank Research Observer, Vol.8, No.2, pp.143-68
Heritier, A., 1993, 'Einleitung. Policy-Analyse. Elemente der Kritik und Perspektiven der Neuorientierung', in idem. (ed.), Politische Vierteljahresschrift, Sonderheft 24, Policy-Analyse. Kritik und Neuorientierung (pp. 9-36), Opladen: Westdeutscher Verlag
Hillebrand, W., 1991, Industrielle und technologische Anschlussstrategien in teilindustrialisierten Laendern. Bewertung der allokationstheoretischen Kontroverse und Schlussfolgerungen aus der Fallstudie Republik Korea, Berlin: Deutsches Institut fuer Entwicklungspolitik
Holanda, N., 1993, A crise gerencial do Estado brasileiro, in J.P. dos Reis Velloso (ed.), Brasil: a superacao da crise (pp.163-176), Sao Paulo: Livraria Nobel
Coutinho, L., and J.C. Ferraz (eds.), 1993, Estudo da Competitividade da Industria Brasileira, Campinas: Papirus
Juergens, U., and W. Krumbein (eds.), 1991, Industriepolitische Strategien. Bundeslaender im Vergleich. Berlin: Edition Sigma
Kaufman, R.R., 1990, 'How societies change developmental models or keep them: reflections on the Latin American experience in the 1930s and the postwar world', in G. Gereffi and D.L. Wyman (eds.), Manufacturing Miracles: Paths of Industrialization in Latin America and East Asia (pp.110-38), Princeton: Princeton University Press
Killick, T., 1994, 'East Asian miracles and development ideology', Development Policy Review, Vol.12, No.1, pp.69-79
Klitgaard, R., 1991, Adjusting to Reality. Beyond 'State versus Market' in Economic Development, San Francisco: ICS Press
Lamounier, B., 1991, 'Unternehmer, Parteien und Demokratisierung in Brasilien 1974-1990', Lateinamerika. Analysen Daten Dokumentation, No.16, pp.7-19
Lamounier, B., 1993, 'Institutional structure and governability in the 1990s', in M. D'Alva G. Kinzo (eds.), Brazil The Challenges of the 1990s (pp.117-137), London: British Academic Press
Mainwearing, S., 1992, 'Brazilian party underdevelopment in comparative perspective', Political Science Quarterly, Vol.107, No.4, pp.677-707
Mathieu, H., 1991, Transnational corporations, the State, and class formation in industrial policy development: the Brazilian automotive, chemical, and electronics industries in comparative perspective, draft of a Ph.D. thesis for submission to the London School of Economics and Political Science
Mayntz, R., 1991, Modernization and the logic of interorganizational networks, Koeln: Max-Planck- Institut fuer Gesellschaftsforschung
Messner, D., 1993, 'Shaping industrial competitiveness in Chile. The case of the Chilean wood-processing industry', in K. Esser et al., International Competitiveness in Latin America and East Asia (pp.33-66), London: Frank Cass
Meyer-Stamer, J., 1992, 'The End of Brazil's Informatics Policy', Science and Public Policy, Vol.19, No.2.
Meyer-Stamer, J., 1995, 'New departures for technology policy in Brazil', Science and Public Policy, Vol.22, No.5
Meyer-Stamer, J., C. Rauh, H. Riad, S. Schmitt, and T. Welte, 1991, Comprehensive Modernization on the Shop Floor: a Case Study on the Brazilian Machinery Industry, Berlin: German Development Institute.
Moreira, M.M, 1993, Industrialisation and Interventions: The Role of Governments in Developing Countries: Brazil, Rio de Janeiro: UFRJ, Instituto de Economia Industrial
Nam, C.H, 1994, 'Industrial clientage in democratic reforms', Pacific Focus, Vol.9, pp.153-70
Nelson, R.R., 1992, 'National Innovation Systems: A Retrospective on a Study', Industrial and Corporate Change, Vol.1, No.2.
O'Donnell, G., 1988, 'Challenges to democratization in Brazil', World Policy Journal, Vol.5, No.2, pp.281- 300
O'Donnell, G., 1993, 'On the State, democratization and some conceptual problems: a Latin American view with glances at some postcommunist countries', World Development, Vol.21, No.8, pp.1355-69
OECD, 1992, Technology and the Economy The Key Relationships, Paris
PBQP, 1992, Pesquisa qualidade e produtividade no meio empresarial, Brasilia
Perez, C. and L. Soete, 1988, 'Catching up in technology: entry barriers and windows of opportunity', in G.Dosi et al. (eds.), Technical Change and Economic Theory, London, New York: Pinter
Pinheiro, A.C., et al., 1993, 'Composicao setorial dos incentivos ΰs exportacoes brasileiras', Revista Brasileira de Economia, Vol.47, No.4, pp.473-501
Porter, M.E., 1990, The Competitive Advantage of Nations, New York: The Free Press
Powell, W.W., 1990, 'Neither market nor hierarchy: network forms of organization', Research in Organizational Behaviour, Vol.12
Rosenblatt, D. and A.D. Novaes, 1993, 'O poder regional do Congresso uma atualizacao', Revista Brasileira de Economia, Vol.47, No.2, pp.305-12
Sabel, C.F., 1993, 'Constitutional ordering in historical context', in F.W. Scharpf (ed.), Games in Hierarchies and Networks (pp.65-124), Frankfurt: Campus
Salgado, L.H., 1993, 'Politica de concorrencia e estrategias empresariais: um estudo da industria automobilistica', Perspectivas da economia Brasileira 1994 (pp.297-321), Rio de Janeiro: IPEA,
Scharpf, F.W., 1991, 'Die Handlungsfaehigkeit des Staates am Ende des zwanzigsten Jahrhunderts', Politische Vierteljahresschrift, Vol.32, No.4
Schmitz, H., 1993, 'Small shoemakers and fordist giants: tale of a supercluster', IDS Discussion Paper, No.331, Brighton: IDS, University of Sussex
Schwartz, G., 1990, 'Stabilization Policies under Political Transition: Reform versus Adjustment in Brazil', 1985-89, The Developing Economies, Vol.28, No.1
Sequeira, J.H., 1990, World-Class Manufacturing in Brazil. A Study of Competitive Position, Sao Paulo: American Chamber of Commerce Publications Division
Shapiro, H., 1994, Engines of Growth: The State and Transnational Auto Companies in Brazil, Cambridge: Cambridge University Press
Skidmore, T.E, 1967, Politics in Brazil, 1930-1964: An Experiment in Democracy, New York: Oxford University Press
Skidmore, T.E., 1988, The Politics of Military Rule in Brazil, 1964-85, New York, Oxford: Oxford University Press
Soto, B.F., 1993, Da industria do papel ao complexo florestal no Brasil: o caminho do corporatismo tradicional ao neocorporatismo, Campinas: IE/Unicamp
Teitel, S., 1987, 'Towards an understanding of technical change in semi-industrialized countries', in J.M. Katz (ed.), Technology Generation in Latin American Manufacturing Industries (pp.94-115), Basingstoke, London: Macmillan
UNIDO, 1992, Brazil's Industrial Policy: An Assessment in the Light of the International Experience, Vienna (PPD.226)
Vestal, J, 1993, Planning for Change: Industrial Policy and Japanese Economic Development, 1945-1990, Oxford: Clarendon Press
Weyland, K., 1993, 'The rise and fall of President Collor and its impact on Brazilian democracy', Journal of Interamerican Studies and World Affairs, Vol.35, No.1, pp.1-37
World Bank, 1991, Brazil. economic stabilization with structural reforms, (Report No. 8371-BR): Washington
World Bank, 1992, Governance and Development, Washington
World Bank, 1994, Brazil. an assessment of the private sector, (Report No. 11775-BR): Washington
Notes (1) The paper is based on interviews with Brazilian policy-makers and analysts that took place in May 1993, March and June 1994, and August and October 1995. It furthermore builds on earlier research that focused on industrial change on the micro-level [Meyer-Stamer et al., 1991] and traditional sectoral industrial policy [Meyer-Stamer, 1992]. (2) The contribution of industry to GDP grew from 20% in 1929 to 35.8% in 1960 and 51.2% in 1985. The share of manufactured products in total exports grew from 1.2% in 1950 to 37.7% in 1980 [Moreira, 1993, Table A-2 and A-4]. (3) On the whole, in 1985 State-owned and transnational firms contributed 10,7 % and 10,0 % of manufacturing value-added, respectively [World Bank, 1994, Table SA.4]. (4) A comprehensive research project on the competitiveness of Brazilian industry identified various resource- based industries to be particularly competitive, namely soybean oil, coffee, concentrated orange juice, petrol and petrochemical, iron and steel, aluminium, and pulp and paper [Coutinho and Ferraz, 1994]. (5) In the case of South Korea this took the shape of Chung Ju Yung, founder of Hyundai, one of the four leading conglomerates, standing for presidency in 1992. The State reacted to this by arbitrarily excluding Hyundai firms from incentives, subjecting Hyundai firms to unusually strict tax examinations and arresting Hyundai executives for practices that used to be tolerated [Nam, 1994]. (6) To say 'delaying' rather than 'inhibiting' implies that the 'Plano Real', the stabilisation programme that was launched in December 1993 and gradually introduced until July 1994, is going to succeed in reducing inflation. This appeared to be a reasonable proposition, although not a safe bet, at the time this article was completed (May 1996). (7) There are no reliable data regarding this question. (8) Over the 1980s, the number of central government employees grew 'by abut 150 000 or roughly one third of its level at the beginning of the decade' [World Bank, 1991: 123]. The share of the banking sector in GDP grew from five percent in 1970 'to an average of 13 percent by the end of the 1980s' [World Bank, 1994: xiv]. (9) For instance, the long-time president of CNI, Albano Franco, is current governor of his home-state Sergipe. (10) Being asked what kind of industrial policy he would prefer, a typical statement of a firm-owner or manager would be 'I would rather prefer to be left alone. The best thing government could do would be just to stop creating confusion'. (11) Verbal communication in interviews with business association executives in Florianσpolis and Blumenau, May 1993 and March 1994.