CIAO DATE: 12/99

Shifting States in Uneven Markets: Political Decentralization and Subnational Industrial Policy in Contemporary Brazil and Spain

Alfred P. Montero

Rights vs. Efficiency Paper #1
September 12–14, 1997

Institute for Latin American and Iberian Studies at Columbia University

 

Prepared for delivery at the Inaugural Meeting of the International Working Group on Subnational Economic Governance in Latin America and Southern Europe Columbia University in the City of New York, September 12-14, 1997. (c) Copyright by the author

The geographic unevenness of development in late and advanced industrializing countries is an old problem that does not seem to go away. Economists such as Alfred Marshall, Albert O. Hirschman and Gunnar Myrdal attempted to explain how “industrial agglomerations” develop and produce both economies and diseconomies within nations. ₁   While core economies develop propitious links between producers and consumers, secondary regions often stagnate due to poor infrastructure, few industrial producers and nascent markets. Such uneven patterns of industrial development were most apparent in the late-industrializing countries during the height of their state-led models of growth. The state-led model of industrialization in Latin America and some Southern European countries, for example, served to concentrate and develop industry in “core regions.” São Paulo was favored as the regional locomotive of President Juscelino Kubitschek’s push for “fifty years of development in five” through public firm investments, protectionism and official finance during the 1950’s. In a similar way, the dominance of industries in Madrid and Catalonia in Spain and Buenos Aires in Argentina was promoted by industrial policies that favored the most advanced segments of national industry invariably located in these core regions. Even when central state managers sought to remedy these spatial inequalities, the proposed cure often exacerbated the unevenness of development. The Brazilian military governments of the post-1964 period created massive development projects in the Amazon and some peripheral states, but these secondary economies worked only to bolster São Paulo’s dominance as this state continued to account for more than 55% of Brazilian industry. During the crisis of state-led development in these countries during the 1970’s and 1980’s, the problems created by uneven growth worsened. The regional distribution of public firms in Brazil, Spain, and other late-industrializers left underdeveloped regions with uncompetitive, burdensome wards of the national state. Productive “restructuring” of these firms would only add to the social misery of underdeveloped areas. The result of uneven growth was, to follow Richard Locke’s account of Italy, the creation of “national political economies that did not function as coherent systems but rather as incoherent composites of diverse subnational patterns." ₂

Recent political and economic changes in industrializing countries compel political economists, particularly those who focus on the state, to revisit these spatial distinctions in a meaningful way. During the 1980’s, the crisis of state-led industrialization and the transition to economic liberalism were accompanied by political transitions from authoritarian to democratic regimes in many developing countries. These “dual transitions” transformed the institutions, functions and authorities of states. Democratic transitions and their subsequent consolidation empowered subnational governments with new authorities and fiscal resources over development and social policy. Meanwhile, liberal economic reform shifted central state functions, particularly in economic policy, to the newly empowered regional and local governments.

These shifting states - politically decentralizing, economically composite and developing—represented an array of new challenges to the roles and domains of states and public policy. While democratization empowered subnational governments and political forces with new authorities over development, economic stabilization and structural reform rolled the national state back from several arenas of development policy, in effect, leaving these areas open to the policymaking of other levels of government. These changes created a host of opportunities and challenges to subnational regions in several late industrializers. Subnational governments, which at one time were relegated to playing a secondary role under state-led models of development, suddenly found that within newly shifting states they were empowered with new policy tools, political interests and fiscal resources for addressing the problems created by the uneven growth produced by state-led developmentalism. How did these subnational governments evolve their new policy capacities? Under what political, institutional and economic conditions could they promote industrial development?

Facing such challenges would not be easy for central and regional governments in late-industrializers. The shifting of states in the developing world evolved very differently from composite economies in the developed world. Unlike Italy and established federalisms such as Germany and the United States, the economic and political crises of these late industrializers during the last three decades were sharper and more traumatic. While composite economies in the developed world could experiment with an array of economic reforms given relatively mild economic dislocations and under the stability of consolidated democratic regimes, composite economies in the late-industrializing world struggled with intense economic and political change that occurred concurrently. The crisis of state-led industrialization in the late-industrializing composite economies generated more dramatic fiscal crises and more rapid transitions to market-oriented development during the 1980’s and 1990’s. State-led policy systems of industrialization collapsed with little to take their place. Additionally, composite economies in the late-industrializing world faced these severe economic transformations while they engaged in political transitions from authoritarian to democratic regimes. In this way, “dual transitions” emerged between state-led/authoritarian models of development to market-oriented/democratic models. In late industrializers such as Brazil, Spain, Argentina, Mexico, Peru, Chile, Portugal, Uruguay (to name a few), these dual transitions transformed economic and political life in unprecedented ways.

As these new political realities took shape, political economists were without the analytical tools to understand them. The economics profession largely abandoned its earlier concern with the spatial dimensions of development. ₃   Political economists working on industrial development employed comparative political models following East Asian contrasts with Latin America. ₄   These authors argued that “strong states” were politically centralized, bureaucratically developed actors capable of “governing the market." ₅   Applications of the East Asian “strong state” model to other industrializing countries retained an emphasis on political centralization and executive-level relationships between the state and business. ₆   Such arguments even influenced political economists’ view of democratization. These scholars would conclude that organized accountability was best maintained by two- and dominant-party systems; a politically centralized system of “stable political rule.” ₇   The analytical framework of the strong state and centralized accountability was poorly suited for understanding the shifting of development questions to the policy domain of newly empowered, subnational governments. Despite academic prescriptions for political centralization, it appeared that shifting states were moving in an entirely different direction; one that could not be sufficiently understood with references to “strong states.”

Decentralization studies, however, fared no better in shedding light on shifting states. Approaches to “fiscal federalism” and the principle of “subsidiarity” in the European Union (EU) improved on “strong state” arguments by pointing to the potential efficiency gains from the distribution of intergovernmental authorities and resources that decentralization could provide. ₈   These arguments, however, were weak on the politics as they tended to embrace static and functionalist models. Federalism scholars failed to highlight the distributional conflicts that punctuated the political decentralization process in developing countries. Studies of the socio-political associations governing economic change in advanced capitalist regions also failed to create applicable, analytical frameworks. Scholars of the propitious social and economic conditions (“economic governance”) that explained cases of successful economic adjustment in the Northern and Central Italian industrial districts, ₉   Baden-Württemberg in Germany, ₁₀   and Silicon Valley and Route 128 in the United States, ₁₁   supplied few workable theories for developing countries. These studies assumed either the preexistence of a political order amenable to economic restructuring, or they posited the emergence of coherent networks of well-organized political, economic and societal actors. Analysts looking for the same propitious social and political factors found in the dynamic industrial districts of the “Third Italy,” military spending and research support of Silicon Valley and MIT, and the preexistence of a “climate of social consensus and strong credibility” could find none of these conditions in the secondary or most of the core regions of developing countries. ₁₂

None of these approaches sufficiently examines the factors that allow subnational government in shifting states to promote industrial development. The true richness of shifting states in uneven markets is ignored analytically when, in fact, these dimensions are essential for investigation of how public policy affects economic life in these cases. More important, an examination of the factors that explain the organization and performance of subnational industrial policy systems in shifting states is more likely to tell us more about the conditions necessary for building and maintaining such networks in the regions of more developed economies. Given the challenges facing shifting states, the factors affecting the evolution of subnational policy systems in these cases are even more likely to be effective in cases where the mercurial nature of national political institutions and regional economies is not as great.

 

The Argument

The present study explains how the organization and performance of subnational development policy systems are determined by a set of “structural” (national institutional and socio-economic) factors and more proximate political and institutional associations. My argument assumes that the effects of political decentralization on the organization and performance of subnational governments is neither inevitable nor automatic. Specifically I argue that subnational industrial policy systems are built due to the identification of the political interests of subnational leaders with technocratic solutions to regional economic problems. However, once these institutions are created, their maintenance and performance depend upon the development of an array of horizontal ties between public policy agencies and organizations in political society (unions, parties, and business associations) and vertical associations between development agencies and private firms. The diversity of horizontal ties allows subnational development agencies to accumulate political and technical resources needed to maintain the policy system even when the political support of regional leaders wanes. Through vertical ties public agencies develop an understanding of the economic problems of private firms. Vertical ties also enhance accountability of both firms and public agencies by creating monitoring functions that prevent mutual shirking. ₁₃   Industrial policy that results from such a rich network of associations is more likely to successfully promote industrial development—that is, improvements in levels and quality of industrial investment and employment—than subnational policy systems with none or few of these characteristics.

Although industrial firms are not the only, or even the chief employers and investors, in developing countries, they are still central to any political strategy for promoting economic growth. As in some advanced capitalist countries, a significant proportion of service and even agricultural production are linked to industrial manufacturing and extraction. ₁₄   My focus on industrial policy is not intended to exclude other subnational policy areas relevant to regional economies. Rather, my choice of industrial policy highlights what may be an unlikely occurrence: the shifting of an area of economic policy previously dominated by central states in developed and developing countries to subnational government. The improbability of this shift adds to the theoretical significance of the findings. The study is also empirically significant outside the area of industrial policy. It is likely that my arguments concerning subnational political capacities apply to other areas such as social, education and technology policy.

The treatment of the state in this study does not differ greatly from that of the ‘strong state’ authors. My analysis embraces Peter Evans’ “comparative institutional approach” which eschews reducing the state to an aggregation of individual interests, treating it instead as a “historically rooted institution” affected by its interactions with the society and economy in which it is embedded. ₁₅   The approach in this study differs slightly from Evans’ in the degree to which the state’s institutional embeddedness is conceived. Rather than treating the state as a centralized entity and as the basic unit of analysis, the approach here focuses on decentralized state structures and their interactions with society. Decentralization and its effects are treated not in the functionalist manner of fiscal federalist studies, but as the result of a historically and politically contingent process. ₁₆

The present study is not primarily concerned with the causes of political decentralization, although these issues are treated empirically in the cases. Rather, the argument focuses on a series of structural, ₁₇   political and institutional factors that determine the capacities of subnational governments in shifting states to promote industrial development. I do assume, however, that political decentralization is not a one-dimensional process. National states rarely have the power to decide the scope and domain of the decentralization of functions based on their own interests or under the criteria of an abstract model of efficient federalism. Distributional conflicts between national and subnational interests always intervene. ₁₈   As a result, many different processes of political decentralization are conceivable. Even distinct regions within shifted states may maintain different levels of political and resource autonomy from their central states. Indeed, one of the essential conclusions of my study is that they often do.

In this study I posit four structural factors which constrain how political decentralization affects the economic policymaking capacities of subnational governments: 1) national democratization, 2) fiscal constraints, 3) distinct levels of development among subnational regions, and 4) firm strategies in global markets. These factors are often interrelated. They are not discrete, insulated categories.

In developing countries where democratization leads to national institutional changes that shift authority and resources to subnational government, the nature of the institutionalization of democracy during transition and consolidation moderates these shifts of intergovernmental capacity. In new democracies where procedural and institutional assets such as electoral systems, political parties and constitutional rules are well organized, the question of intergovernmental capacity is gradually negotiated within a strengthening “political society." ₁₉   The result is a relatively stable and flexible system of distribution of policymaking capacities. Spain is often considered the archetypal empirical case of this model. ₂₀  

In new democracies where the procedural and institutional assets of democracy are poorly developed or fraught with organizational fragmentation and conflict, the decentralization of authorities and resources is more rapid, less controlled and proceeds in an ad hoc manner. Under these conditions, alternative sources of political authority and organization built around patrimonial elites and based largely on local foci of power in subnational government attain more control over authorities and resources for economic policymaking. Brazil is frequently seen as the key archetype of this model. ₂₁  

Under both models, subnational governments attain a relatively high level of juridical, political and fiscal autonomy. Their capacities to act on these new authorities and resources, however, are mediated by other structural factors.

Fiscal constraints will limit the extent to which regional leaders and subnational development agencies can offer incentives to industrial investment. In new democracies where central states do not maintain a high degree of control over political decentralization, the result of subnational industrial promotion may be a fiscally unsustainable set of policies. Objective economic and fiscal limits, however, restrain regional leaders from pursuing unwise policies. ₂₂   In more developed democracies, central states will maintain more control over the fiscal autonomy of subnational governments. In either case, fiscal limits constrain subnational political ambitions.

Level of development is another relevant structural factor. “Backward regions” in shifted states do not maintain the same array of resources that core regions enjoy. The disadvantages of backwardness are inadequate infrastructure, poor location, the presence of maturing or public industries in decline, oppressive management-labor relations in one-company towns, the lack of industrial diversity and a general unattractiveness of these regions to foreign investors. ₂₃   The advantages of core economies are the presence of the most dynamic industries and professional services, proximity-induced productivity gains, well-developed infrastructure, and the general attractiveness of these regions to foreign capital. ₂₄   Nevertheless, many disadvantages of backwardness can be converted into propitious political and economic advantages. The advantages of backwardness include the tendency of economic crisis to catalyze political support for concerted action. ₂₅   Likewise, core regions become victims of their own success if agglomeration creates urban congestion, high labor and land costs, militant unionism and other diseconomies (disadvantages of core economies). Additionally, their own success and the policy attention of national states that this success accrues dissuade subnational governments in core regions from pursuing their own economic policies. ₂₆

Since subnational governments do not retain command over the key macroeconomic mechanisms of economic policy—interest rates, exchange rates and national commercial policy—they are ill-prepared for affecting the investment strategies of international firms following a global agenda. Firm interests are governed by the nature of their assets, labor needs and changes in market forces and macroeconomic indicators. ₂₇   In this context, the costs of carrying out a subnational industrial policy that will make a difference in a region’s economic development can be high. However, subnational governments, depending on their political organization and fiscal resources, provide infrastructure, political guarantees, labor regulations, information and finance that often prove crucial in determining an investment decision. ₂₈   Firm strategies must be taken into account in evaluating the evolution of any investment decision, yet national and subnational government have a role that has not been entirely excluded by globalization.

The political factors of my argument focus on the formation of political interests among subnational leaders for promoting industrial development. I assume that subnational government leaders are politicians who pursue rationally self-interested political goals, and their first-order preference is to maintain and expand their command over the state apparatus and the society over which it governs. ₂₉   The complexity of subnational development problems requires that politicians build a professionally-staffed array of public agencies capable of evaluating and acting upon these problems. Not all subnational politicians, however, relate their political interests with those of building such a subnational technocracy. Politicians with a weak or weakening political base often shore up their support by sacrificing institutional resources for political gain. ₃₀   Structural factors mediate this tendency. The more immediate and severe the economic crisis, the greater will be the political pressure on regional leaders to build technocratic institutions. In secondary regions these pressures are especially great; they are less salient in core regions. In cases where political society is weak, regional executives will face fewer, organized demands. The scarcity of fiscal resources and other structural obstacles to subnational industrial policy cause short-term political interests to take control.

Based on these and other reasons, the continuation of development policy agencies and institutions cannot depend solely upon political leaders. Politicians, no matter their level of support, eventually expire due to the end of their mandates or natural causes. Technocrats prefer to accomplish their professional tasks without relying on political conditions, yet they are often forced to follow political-technocrats—professionals who are technically competent and politically savvy. ₃₁   Political technocrats rely on their connections to government, other official organizations, business groups, firms, labor unions, and national government agencies in order to exchange information, preserve their resources and garner additional political support for an expansion of authorities and assets. In this study, these connections between political technocrats and other elements of political society are called horizontal embeddedness and they are crucial for the continuation of subnational industrial policy systems, especially when the political support of subnational leaders wanes. ₃₂   However, horizontal ties also open subnational development agencies to additional political influences. Still, the extent to which these agencies are already politically supported by regional leaders and have diverse, alternative sources of political support among their horizontal partners will reduce their tendency to rely too much on any one actor. A diverse political base gives the agencies some freedom from parochial influences by arming them with what Peter Evans describes as “a combination of internal coherence and external connectedness” (“embedded autonomy”). ₃₃  

Subnational development agencies must also foster close ties to their clients. Firms that become clients of the development agencies create vertically embedded associations with the subnational technocracy. The nature of these associations is based on exchange of information, resources and services between the official agency and the firm, the latter being usually, the recipient. The raison d’être of creating vertical ties is the cultivation of monitoring, information retrieval, and idea exchange between technocratic agencies and private firms. Vertical embeddedness allows policymakers to stay informed about the economic needs and opportunities industrial firms face. Such close ties improve the ability of technocrats to contour policy resources to address entrepreneurial concerns, while they foster the monitoring functions that make certain that private interests stick to the rules governing policy goals. These ties also attenuate the problems of asymmetric information that prevent private firms from realizing optimal investment opportunities and they encourage the exchange of ideas that leads to policy innovations. Finally, a vertically embedded policy system is better able to detect when market incentives are preferable to policy interventions. The greater the vertical embeddedness of the subnational policy system, the greater the sensitivity of policymakers to the economic costs and opportunities of public intervention.

The capacity of technocratic development agencies to create horizontal and vertical associations is not automatic. Both structural and more proximate political factors constrain the emergence of embeddedness. First, the relative organizational strength of political societal actors will determine to what extent these groups develop horizontal linkages with subnational technocracies. In national cases where the democratization experience weakened the organizational capacity of political society the range of horizontal associations will be narrow. The strongest horizontal linkages that will emerge, if any do, in these cases will be among official agencies on the subnational level. These horizontal associations can still prove strong if a diverse and flexible range of subnational agencies develops a consistent policy network. More isolated subnational agencies will more easily fall prey to political interests. In national cases where the democratization experience strengthened the organizational capacity of political society, subnational development agencies will enjoy a wider range of possible horizontal associations. If they depend on one political actor, however, they will likely fall prey to parochial interests. Diversity and flexibility of horizontal associations translate into institutional strength.

The indicators for horizontal embeddedness are the cohesion and flexibility of development agencies and institutions. The associations among policy agencies, government ministries, and members of political society that the concept of horizontal embeddedness entails are the key factors that preserve this mix of policy cohesion and flexibility. The indicators for horizontal embeddedness are most evident when they are tested by threats from parochial political interests, the removal of initial levels of political support and changes to the economic structure of opportunities that endanger policy performance and require new, technical directives. A horizontally embedded subnational policy system will face such threats and remain cohesive and flexible.

The indicators for vertical embeddedness are the frequency and familiarity of contacts between subnational development agencies and client firms. ₃₄   Evidence for these close and frequent ties are reactions from firm managers about the involvement of policymakers in the investment and productive projects of client firms and statements from policy managers on their level of access to private firms. Subnational policy systems that maintain close contact with client firms learn about the needs of business and the conditions of markets. Policies resulting from such exchanges are more likely to provide a useful set of solutions than directives that emerge from more arms-length associations between policymakers and private firms.

In this study the indicators of policy performance are relative improvements in the levels and quality of industrial investment and employment in the regional economy. By “relative” I mean in comparative perspective to past performance and similar policy systems in other regions. “Improvements in levels” refer to aggregate increases. “Improvements in quality” refer to productivity enhancement. Therefore my indicators are trends in positive industrial investment and employment in comparison to the results of subnational policy systems elsewhere and in other periods. These indicators for the dependent variable are inherently limited in that they do not detect long-term effects on competitiveness. Of course, the causes of firm competitiveness are legion, involving diverse factors such as technology, innovation, macroeconomic and regional policy, sectoral factors and market forces among other causes. Such is the number of potential independent variables that one is hard-pressed to flesh-out the particular influences of subnational industrial policy. Certainly the existence of a competitive industry does not necessarily mean that policy was the cause. ₃₅   Nevertheless, the quality of improvements to production and employment which are intricately linked to firm competitiveness is an important dimension of the indicators in this study. Investments that produce innovation and improve the quality of labor as a factor of production (rather than simply as a cost) are the most desirable. ₃₆   Such indicators are highlighted in this study whenever possible, particularly in the case of subnational policies that facilitate the exploitation of externalities and alleviate obstacles to investments designed to increase productivity. In these cases, productivity improvements cannot be made until market failures are alleviated.

My purpose in this study is analytical. I do not seek to explain the causes of economic growth, changes in labor markets or firm competitiveness. Instead, I seek to flesh-out why certain subnational policy systems perform better than others. The cases in this study illustrate how subnational public agencies promoted conditions that induced innovative and productive firm responses to market failures.

The empirical basis of my argument is a comparative study of contemporary Brazil and Spain that analyzes the industrial policy of three subnational regions in each country. These two countries were selected for their similarities and their differences. The similarities—the regional unevenness of state-led models, the exhaustion of those models and the turn to market-oriented economic policies during the 1980’s, and political decentralization linked with democratization—place both countries in the universe of cases of late industrializers with shifting states in uneven markets. Differences in democratization, relative fiscal constraints, and their position in global markets should make Spain a candidate as a “strong state,” a more stable democracy and a relatively more developed “success story” than the case of Brazil. Despite these differences, the subnational comparisons show that other factors explain common outcomes in regional development systems. These similarities suggest that structural factors affect the nature and performance of subnational industrial policy, but do not determine the outcome. Rather, the cases confirm that combinations of the structural, political and institutional factors of my argument explain the relative organization and performance of subnational industrial policy in Spain and Brazil. Analysis of the comparison will attempt to flesh-out the relative weight of each factor, although space limits prevent an exhaustive evaluation.

The next section will briefly describe how the crisis of state-led industrialization in both countries was accompanied by the political decentralization of the state. Subsequent sections will briefly examine the evolution of industrial policy in three Spanish regions (Asturias, Madrid and Andalucía) and three Brazilian states (Minas Gerais, São Paulo, and Rio de Janeiro).

 

The Decline of Industrial Coordination from Above—Spain and Brazil Compared

The early phases of heavy industrialization in the 1930’s and 1940’s in Spain and Brazil were dominated by public sector investments in steel, mining and petroleum. In Brazil, statism found ideological support in Getúlio Vargas’ Estado Novo (1937-1945) and its doctrine of using the state to exploit the wealth of the country. ₃₇   In Spain, General Francisco Franco created the Instituto Nacional de Industria (National Institute of Industry—INI) in 1941 as a holding company of state firms largely concentrated in petroleum, steel, mining and transportation. ₃₈   Inspired by the fascist Italian Istituto per la Recostruzione Industriale (IRI), the main goals of INI industrial policy were industrial autarky and national defense. ₃₉

The state-led model in both Spain and Brazil exacerbated existing geographic unevenness in the distribution of industry. In Spain, INI firms operating in manufacturing tended to locate in the developed industrial regions of Spain, particularly Madrid and Catalonia, while INI firms in raw materials or commodities such as coal and steel, located in the mature industrial regions of the North (the Cornisa Cantábrica). Juan Antonio Suanzes, the INI’s founder and president until 1963, argued for some purely “technical” criteria for placing INI firms that followed each region’s factor endowments. The resulting inequalities were predictable. By settling public firms in the most developed regions or in the regions with the greatest abundance of natural resources, the INI further concentrated industrialization in a handful of zones and specialized the industrial structures of some regions in mature sectors. ₄₀   Most of Spain’s other regions, including Andalucía, in the relatively poorer South, were neglected by INI. Unable to rely on a diversified, existing industrial infrastructure, these regions stagnated in their underdevelopment.

In Brazil, state-led industrialization did little to change the fact that São Paulo remained the home of more than 55% of the country’s industrial firms, yet localization policy was far more deliberate then INI’s policy in Spain. ₄₁   The post-1964 military governments attempted to weaken paulista opposition to authoritarian rule by reallocating official expenditures and public firms to the rural Northeast and North, a base of conservative support. The location of petrochemical poles in the Northeast and the creation of the National Cane Alcohol Program (Proálcool) sought to redistribute São Paulo’s industrial assets to other Brazilian states. Nevertheless, since the industries of other states depended on the consumer market in São Paulo, federal policies failed to make a dent in the state’s industrial dominance. Public expenditures on São Paulo even increased. As the chief domestic market for petroleum and steel, economic constraints compelled federal authorities to place additional refineries and steel mills in São Paulo. Arguably, Proálcool did more to revitalize São Paulo’s agroindustrial sector than to create a lasting industrial structure in the poor Northeast. ₄₂

In both countries, however, the larger model of state-led growth could not be sustained over time. Each pattern of state-led industrialization entered a severe period of crisis in the 1970’s and 1980’s. Both had international catalyzers. In the case of Brazil it was an escalating external debt that developed into a fiscal crisis of the state. ₄₃   Public sector-led growth policies under the military’s Second Development Plan generated the largest external debt in Latin America. Growing deficits and financial speculation led to mega-inflation during the 1980’s. In Spain the cause of economic crisis was stagflation in Western Europe resulting from the first oil price shocks of 1973. Although the Spanish economy began its liberalization in 1959 with Franco’s macroeconomic reform, Spain’s public sector still represented a growing burden on official accounts that would need to be reduced to avoid soaring deficits and promote the country’s entry into the European Community (EC) in 1986.

In Brazil, the fiscal crisis of the state led policymakers to reduce the autonomy and entrepreneurial role of public firms. ₄₄   Public production as a percentage of gross investment fell from 21% in 1982 to 6% in 1990. ₄₅   The sharp reduction in the entrepreneurial capacity of the Brazilian public sector, however, did not produce a more stable economy in the medium-term. The continued postponement of extensive fiscal and administrative reform during the first civilian governments of the transition to democracy (José Sarney 1985-1990; Fernando Collor 1990-1992; and Itamar Franco 1992-1994) led from one failed stabilization plan to another. Monthly mega-inflationary rates exceeding 60% were common for much of the late 1980’s and early 1990’s. Industrial policy became a secondary consideration. Between 1985 and 1991, no fewer than 10 different proposals for an “industrial policy” were circulated by Brazilian ministries and hyped by the presidents and cabinet officers they served. Each of these proposals, however, did not get past the position paper stage. ₄₆   Macroeconomic reform remained the political priority of national politicians.

Spain’s monetary and fiscal problems were not as severe as Brazil’s. The country’s industrial crisis occurred in the 1970’s when the first oil price shocks of 1973 threatened stagflation in Western Europe. The INI industries stagnated as foreign and domestic demand declined precipitously during the decade, generating inflation, large trade deficits and high levels of unemployment. Continued INI losses and chronic trade deficits created the conditions for a fiscal crisis in Spain in the early 1980’s. The budget deficit grew from 1.7% to 5.5% of GDP in the short period from 1979 to 1982. ₄₇   Between 1980 and 1993, the budget deficit averaged 4.6% of GDP and the public debt averaged 37% of GDP during 1980-1992 but it became progressively worse, topping out at an estimated 60% by 1994. ₄₈

The priorities of planning were soon outweighed by the fiscal and macroeconomic exigencies of reducing inflation and the public budget deficit. ₄₉   Despite the election of the Socialists (PSOE) led by Felipe González in 1982, tight monetary and salary policies were continued over from the previous Centrist government to keep inflation low. Centrist and Socialist policymakers alike were convinced that they had to prepare domestic industry for competition from more advanced European economies and to redress the fiscal imbalances and monetary instability that could prohibit accession to the EC. ₅₀   The dominance of EC policy priorities limited the scope of industrial reform by disassembling the entrepreneurial projects of the INI, but EC fiscal transfers also made a more extensive restructuring of the public sector possible. ₅₁   The Socialists pursued a wide-ranging industrial reconversion policy that premiered in a White Book report (Libro Blanco) in July 1983, delivered by then Minister of Industry, Carlos Solchaga. Both the White Book and its legislative successor, the Law of Reconversion and Reindustrialization (1984), targeted sectors in industrial decline for subsidized credit and structural aid. The key sectors of the reconversion—steel, metallurgy, appliances and textiles—all received large amounts of state aid. Zones of Urgent Reindustrialization (ZURs) surrounding the most hard-pressed sectors were provided with preferential credit and subsidies and were the beneficiaries of funds for labor relocation. ₅₂   The result was a neoliberal policy that succeeded more at macroeconomic stabilization than at reducing unemployment or promoting industrial growth. The INI spent large amounts of public money to remodel and eliminate entire lines of production which reduced jobs and salaries. Total employment in the INI fell from 216,700 workers in 1983 to 154,500 in 1989, almost a 30% reduction. ₅₃   Like the French before them, the Spanish Socialists abandoned expansionism and embraced neoliberal reform. ₅₄

Despite the apparent regional and historical differences between Brazil and Spain, two countries literally an ocean apart, the evolution of macroeconomic and industrial policy took very similar directions during the 1980’s and the 1990’s. The national public sector in both countries underwent a period of severe decline, jeopardizing older patterns of state-led development. As the national state retracted from its history of extensive intervention in the market, subnational governments became more involved. To a great extent, this was the result of a process of political decentralization that emerged from democratic transitions in both countries.

 

Shifting the State: The Politics of Decentralization

In both Brazil and Spain, the 1980’s witnessed a rapid decentralization of both fiscal resources and policy autonomy as a result of political bargaining between local and national interests. In each case, transitions to democratic rule provided dynamic bases for these processes.

The Brazilian military, which had originally embraced a politically centralized formula of authoritarianism, eventually decentralized taxes and intergovernmental transfers through a 1967 tax reform. ₅₅   The act was an attempt to counterbalance the power of the Southern and Southeastern states where much of the opposition to authoritarianism emerged. ₅₆   Yet it also set a precedent for expanding subnational autonomy before constitutional democracy was established. As a result, fiscal decentralization could not be controlled as the democratization process proceeded. Local politicians, in the scramble to position themselves for power in the gubernatorial elections of 1982 and the new civilian government in 1985, bitterly attacked the military’s emphasis on centralization. Added to the cacophony of dozens of newly autonomous social and political actors, from unions to social movements, who joined the attack on centralism and the emergence of the Congress as an increasingly important venue for lobbying and deal-making, the stage was set for a dramatic decentralization of fiscal authority. ₅₇   All of these conditions expanded the “politics of the governors”—the empowerment of state and municipal governments and their local bases of patrimonial politics. ₅₈   Regional lobbies exerted a tremendous influence on the Constituent Assembly charged with drafting the new constitution in 1987. ₅₉   As a result, the 1988 Constitution would make Brazil one of the most fiscally decentralized federalisms in the world.

After figuring for revenue sharing and transfers, in 1974, the national state disposed of 50.2% of all revenue, the states 36.2% and the municipalities 13.6%. By 1988, the national state disposed of 33.4% of revenue while the states boosted their share to 50.7% and the municipalities to 15.9%. ₆₀   Most important, the 1988 reform gave the states reserved authorities to engage in economic policies not within the jurisdiction of the national state. The open-endedness of this provision granted the states carte blanche for directing spending at an array of “development” needs. More than 90% of intergovernmental fiscal transfers were not earmarked for particular purposes by Brasília and were free to be spent at the discretion of state and municipal governments. ₆₁

Political decentralization was also extensive in Spain, but it did not occur as rapidly or uncontrolled as the Brazilian process. The emphasis placed during the Spanish transition on building political society allowed political decentralization to be more institutionalized than it was in Brazil. ₆₂   Nevertheless, the overall result was similar with differences in degree. The institutions governing political decentralization in Spain were imbued with a degree of open-endedness that allowed subnational government to expand their authorities and resources during the 1980’s significantly. As in Brazil, the nature of the democratic transition directly shaped the degree of this open-endedness. From the beginning, the question of regional autonomy dominated the emergence of Spanish democracy. Separatist violence in Basque and calls by the Catalonians for autonomy in 1977 snowballed into a wider debate concerning the structure of the state. Soon all the Spanish regions moved to claim their own rights of autonomy precisely during a time democratic institutions were being conceived. ₆₃   The 1978 Constitution guaranteed the regions “autonomy for the administration of their respective interests” (Article 137) and the controversial Title VIII authorized a negotiated route to autonomy based on the development of statutes of autonomy for each region, dividing responsibilities for taxation and shares of intergovernmental transfers.

The system conceived by the 1978 Constitution was meant to be flexible, to build the “state of the autonomies” gradually, over time and through a process of negotiation between the central state and the regions. ₆₄   Nevertheless, different needs and political interests among the regions thwarted central state ambitions to generalize a model of autonomy to be applied to all cases. ₆₅   The negotiation of statutes of autonomy with each region had the effect of further compartmentalizing the autonomy process. Different regions obtained distinct and sometimes “exclusive” rights. ₆₆   Although there were differences in degree, all the regional governments successfully gained increases in fiscal resources and authority during the long evolution of the statutes of autonomy. The open-endedness of the stipulated political economic rights of the regions allowed for a great deal of leeway in the interpretation of the capacity of the regions to promote industry. The negotiated process allowed the regions to claim authority through their statutes of autonomy that were not specifically contained in the 1978 Constitution. ₆₇   Thus many regions developed their own public sectors during the early to mid 1980’s, launching an array of agencies designed to promote industrial investment. By the end of 1988, the regions were spending a total of 35% of the national budget, but 75% of the monies allocated to promoting industrial investment. ₆₈

Subnational governments in both countries viewed these institutional transformations as fortuitous. Faced with a declining national industrial policy, regional governments concerned about promoting industrial development could expect little from their central states. Political decentralization in Spain and Brazil provided authorities and resources to mount local responses. Yet not all subnational regions were prepared or capable.

 

The Spanish Subnational Cases

Asturias During the 1980’s, Asturias seemed an unlikely candidate for autonomy. Unlike Catalonia and Basque, Asturias lacked a distinctive national and linguistic identity. The 1978 Constitution did not place the region on the “fast track” to autonomy. Yet Asturias faced arguably the worst economic crisis in Spain. As a secondary region in Spain, far to the northwest, away from the dynamic core economies of Madrid and Catalonia, Asturias underwent a process of industrial decline and restructuring. The public mining and steel firms that had once given Asturians an industrial identity rapidly cut their production and downsized their workforces during the 1980’s as part of the Socialists’ industrial ‘reconversion.’ Between 1980 and 1992, 24% of all industrial jobs were phased-out, with 55% of those losses occurring in the public sector. ₆₉

Asturian unions, the most politically influential actors in the region, mobilized throughout the 1980’s and into the early 1990’s to slowdown the ‘reconversion.’ The mining unions were led by José Angel Fernández Villa, a fiery leader who exerted great influence within the national and regional PSOE. Manuel Fernández Lito, the head of the steel unions, also retained tremendous influence on national politicians. The Asturian regional government, which remained under the control of the Socialists from 1983 to 1995, was commanded by Pedro de Silva Cienfuegos-Jovellanos for most of this period (1983-90). Both the Asturian unions and the regional government shared interests in limiting the extent of the industrial ‘reconversion’ and garnering national and EC monies to allay the social costs. The regional government, however, also sought the political goal of building Asturian autonomy.

This second goal created strong political interests on the part of Silva to construct a regional technocracy that would promote industrial investment and employment. The Socialist president had been elected on the promise of reorienting national industrial policy to take into account regional concerns. Yet his initial contacts with the national PSOE leadership had convinced him that the industrial ‘reconversion’ could not be significantly amended. ₇₀   Silva also viewed the Socialists’ regional programs, the ZURs, as inefficient and limited. Faced with severe economic exigencies and lacking both a strong juridical and ethnic basis for autonomy, Silva increasingly spoke of the regional government’s role in implementing industrial policy as a key legitimating pillar of regional autonomy. ₇₁   Encouraged by the arguments of political technocrats in the Asturian Society of Economic and Industrial Studies (SADEI), the region’s official research agency, and the regional Chancellery of Economy and Industry, Silva took advantage of the open-endedness of the Asturian Statute of Autonomy’s economic policymaking provisions to create an array of public development agencies.

The new Asturian technocracy, led by the Institute of Regional Promotion (IFR), created in 1983, and the Service of Assistance and Entrepreneurial Promotion (SAYPE), initiated in 1988, among other agencies, created a number of horizontal and vertical associations that helped to expand industrial investment and employment in Asturias. Backed by IFR’s market and infrastructure studies, Silva and his staff of political technocrats established relations in 1989 with Germán Lastra, a native Asturian who headed the DuPont Nemours corporation’s fledgling Spanish subsidiary. Despite significant legal, union and environmental challenges to DuPont in the U.S. and Europe, the firm decided to place a multi year, billion dollar investment in Asturias during the 1990’s. The decision seemed improbable considering that Asturias maintained less than 1% of total foreign direct investment in Spain, housed the country’s most politically active unions and a sensitive environmental movement. The unlikely choice of Asturias for DuPont’s largest European investment was the result of concessions engineered by Silva with the unions and local environmental groups, national and regional subsidies totaling $267 million, and the technical support supplied by IFR. ₇₂   During the 1990’s, DuPont’s presence in Asturias would account for just more than 7 percent of the region’s GDP and about 6,000 jobs, a significant contribution to a secondary Spanish region in industrial decline. Here, horizontal ties among members of political society and the development agencies assured that DuPont would face a united front in Asturias while vertical ties between regional technocrats and firm managers eased negotiations that consolidated the multinational’s investment plan in the region.

The building of horizontal and vertical linkages within the Asturian industrial policy system also proved crucial in a completely different area of policy: the promotion of small- and medium-sized firms. These firms are the chief employers in Asturias and they are perennially cut off from sources of finance due to their size and lack of collateral. ₇₃   Aiding small firms became a political priority of the Asturian political technocracy once the national government reduced subsidies for small Asturian firms during the early 1990’s. ₇₄   In response, IFR and SAYPE organized a Support Plan for Small- and Medium-Sized Firms, and then negotiated with Asturian unions and regional politicians an array of funding policies in 1993 (e.g., subsidized interest loans, collateral support, and target loans for productivity-enhancing technologies).

Based on a small-n study of 10 small- and medium-sized firms conducted by the author in March-May, 1996, firm managers reported moderate to significant improvements in the efficiency of production and employment due to agency activities. ₇₅   Most of these respondents argued that such changes would not have been possible but for agency intervention, particularly the manner in which financial support was tailored to improve efficiency in the short-term (a critical factor for small firms that typically suffer short life spans). ₇₆   Firm managers noted that close and frequent visits by IFR officials figured prominently as a factor that greatly improved the significance and effectiveness of agency programs. By exchanging crucial information about the priority needs of these firms, specialized finance could be focused on areas of production and employment that would generate the most improvements in efficiency. Of course, IFR and SAYPE officials were willing to pay for monitoring costs by maintaining close relations with client firms, thus reducing the monitoring costs of banks and easing the flow of financial resources normally impeded by information asymmetries. Simultaneously, firm managers made certain that official promises were met once their enterprises satisfied agreed-upon requirements for receipt of program finance.

An evaluation of Asturian industrial policy between 1985 and 1991 will reveal that industrial employment promoted by the regional agencies amounted to 40% of the number of industrial jobs lost between 1980 and 1992 (35,717). ₇₇   SAYPE, which works exclusively with small- and medium-sized firms, helped replace 71%, 46%, and 78% of all industrial jobs lost in 1991, 1992, and 1993, respectively. ₇₈   These figures do not completely reverse Asturias’ total unemployment (71,000, 20%), but the figures do signify an important cut into the impact of deindustrialization. Agency activities have had a significant effect on industrial investment as well. In 1993, IFR and SAYPE programs made possible 28% of Asturias’ total industrial investment that year. The figures are not indicative of a total recovery, but they are significant results for a subnational industrial policy system in a secondary region with few resources, militant unions and still-nascent juridical autonomy. More important, the Asturian case illustrates how political support can create the basis for building a subnational industrial policy system and how horizontal and vertical ties mediate the continuity and performance of development policy.

Madrid In contrast to Asturias, the Community of Madrid is a core region in Spain.

Madrid accounts for 16% of Spain’s GDP and 13% of all industrial employment. About half of all foreign investors, 250 of the largest 500 firms in Spain and 31 of the 32 foreign banks authorized to do business in Spain, are in Madrid. ₇₉   Not surprisingly, all roads lead to Madrid in the literal if not in the figurative sense. La Plaza del Sol, the center of the city of Madrid, is known as “kilometer zero”—the geographic heart of Spain’s interconnected roadways and highways. Such centrality and industrial development gave the Community of Madrid a self-sufficiency that Spain’s secondary regions do not retain. Madrid’s diversified array of industries, administrative and financial services created an “introverted” agglomeration of economies that insulated the region from the vagaries of international trade. ₈₀   Unlike Asturias, Madrid was also spared the throes of the industrial ‘reconversion.’ Job-loss rates during the ‘reconversion’ were less than the Spanish average and job-creation rates were greater than the Spanish average during the 1980’s. ₈₁   Madrid’s industrial structure, with its array of advanced technological firms and administrative services, was also more diverse and more productive than was the case in Asturias. As a result, the region’s political leaders had weak interests in promoting industrial investment.

Joaquín Leguina, the Socialist president of the region from 1983 to 1995, did not seek to develop Madrid’s “slow track” autonomy by linking it to development concerns. In part this was true because the national state supplied enough resources to address the region’s more important urban planning concerns. The distribution of industry within the region to relatively poorer segments was the focus of national and regional efforts. In this area, a regional technocracy was created, led by the Institute of Development of Madrid (IMADE). Yet this agency was poorly linked to other regional and national agencies, including the regional mutual collateral society, Avalmadrid, a regional agency dedicated to reducing the transaction costs of small firms seeking finance. Avalmadrid maintained daily contact with the banks, but the agency tended to foster arms-length relations with IMADE which it viewed as an agency pushing risky small- and medium-sized firm projects. ₈₂   Avalmadrid tended to be conservative in its lending practices, frustrating IMADE staff who could not consolidate financial support for firm projects. ₈₃   Without useful horizontal associations, the lack of political support for IMADE served to undercut the agency’s functions and further isolate it. José Luis Fernández Noriega, Leguina’s Chancellor of the Economy, made IMADE’s directorate a revolving door of political appointments. The agency experienced three changes of director in 1992 and 1993. In October 1993, Fernández Noriega stripped the agency of most of its affiliated public firms. Nine separate societies were closed in an attempt to “rationalize” IMADE’s structure. ₈₄   Cases of fraud among IMADE’s administrators weakened the agency itself in 1994. ₈₅

Agency links to firms, in this institutional climate, could not be maintained. A survey of Madrid firms by the Confederación Empresarial de Madrid (Entrepreneurial Confederation of Madrid—CEIM), the region’s chief business association, in 1987 found that only 5% of the region’s firms believed that the regional government was doing important work and only 1.4% of responding firms were aware of the extent of the government’s activities. ₈₆  

Given the regional government’s political and institutional weaknesses, Madrid was unprepared for addressing the sharp fall in industrial employment that occurred after 1988 due to corporate downsizing and the economic recession. More than 45,000 jobs were phased out between 1988 and 1992 alone. Repeated attempts during this period to forge “pacts” with regional business and labor leaders failed as IMADE and the rest of Madrid’s development apparatus remained isolated and incoherent.

The Community of Madrid had few of the factors that explained Asturias’ more successful experience with industrial policy. First, regional leaders lacked strong incentives to pursue industrial policy since, as a core economy, Madrid did not feel the dislocating effects of the industrial ‘reconversion’ and it relied upon a substantial base of domestic and foreign industrial investment. More important, the institutional development of the region’s technocratic agencies suffered from a lack of horizontal linkages. IMADE’s organizational and financial weaknesses stemmed from arms-length ties to the region’s other development agencies. The failure of policy and the system’s eventual politicization during the early 1990’s made the agency unable to innovate new arrangements with business, labor and Madrid’s other technocratic agencies. Poorly supported and insulated by regional government, IMADE and the rest of Madrid’s development policy system remained passive observers during a period of significant economic change.

Andalucía As one of Spain’s most underindustrialized regions, Andalucía had few of the conditions that favored the development of Madrid’s core economy. Like Asturias, Andalucía is a secondary region in Spain, a peripheral territory with unproductive firms and an industrial structure concentrated in economic (and geographic) enclaves in basic chemicals (Huelva), shipbuilding (Cádiz), and auto assembly (Linares). This inflexible structure was a poor employer. Andalucía maintained unemployment rates more than 30% during the 1980’s, higher than the already high Spanish average of 20%. The Socialists who came to power in Andalucía in 1982 had few interests in changing this structure radically. Rafael Escudero, the region’s first autonomous president, employed Andalucía’s development concerns as part of a wider populist strategy to buttress the influence of the PSOE. Following the examples of Basque and Catalonia, Escudero argued that Andalusians were “oppressed,” but due to the economic exploitation of the region’s raw materials by the rest of Spain, not on ethnic grounds.

As a result, the regional technocracy that emerged to deal with Andalusian underdevelopment, led by the Andalusian Institute of Industrial Promotion (IPIA, later the Andalusian Institute of Promotion, IFA), was manipulated by the populist politics of Socialist leaders. ₈₇   Unlinked to the influential Andalusian business associations that saw IPIA/IFA as a rival for representing the region’s businesses, the institute depended upon the political support of PSOE leaders who routinely used the agency’s directorate to make political appointments. ₈₈  

The politicization of the IFA forced the agency’s technocrats to commit themselves to costly interventions in unviable large firms in politically sensitive areas. In 1994, for example, the IFA bought out a failing Suzuki land rover plant in Linares called “Santana” under pressure from the enclave’s unions. ₈₉   The firm lost millions under public ownership. Between May 1993 and August 1994, the IFA invested more than $269 million to save 13 unviable firms, including Santana. ₉₀   This amount exceeded the total for the IFA’s other industrial promotion programs in agriculture, industry and services in 1993; a most inefficient outcome.

In comparison to Asturias, Andalucía lacked the political and institutional factors that explain the successful formation and performance of subnational industrial policy. First, the IFA system was politically manipulated by successive Socialist governments and politicized in certain cases by local labor unions (e.g., Linares). Second, the IFA was poorly connected to alternative sources of political support. Neither regional business associations nor other sister agencies in the policy system gave the IFA a base of political or logistical support. In this institutional context, the vertical ties between the IFA and Andalusian firms were more top-down rather than exchange-oriented. That made it easy for public agencies such as IFA to impose their own politicized doctrines on how their clients (and in the case of firms such as Santana, their going concerns) would operate in the region. These relations were unable to balance productive and fiscal concerns, leading to costly public investments with poor short- and medium-term results for industrial development. In Asturias, by contrast, where political interests failed to intervene as directly due to the strength of horizontal ties, vertical embeddedness produced more responsive and sustainable policy results.

 

The Brazilian Subnational Cases

Minas Gerais Like Asturias and Andalucía, Minas Gerais is a secondary region. Historically, Minas’ economy depended upon national public investments in steel and mining and the vitality of São Paulo’s industries just to the south. During Brazil’s developmentalist period the state government’s political elites, followers of Juscelino Kubitschek (a former mineiro governor and Brazilian president from 1955 to 1960), built an array of development agencies designed to diversify the inflexible industrial structure of the state. The state utility company (CEMIG), the state development bank (BDMG) and the Institute of Industrial Development (INDI) emerged during the 1950’s and 1960’s as technocratically-managed and politically supported agencies that were “pockets of efficiency” within state government. ₉₁   Unlike the Spanish subnational cases, however, the mineiro agencies were politically supported by a dense network of inter-bureaucratic ties. Mineiro entrepreneurs were politically weak and the state’s oligarchical political class deferred economic policy to political technocrats. ₉₂   Mineiro labor was also weak due to a long history of clientelist labor relations in one-company towns dominated by public firms and government-supported efforts to control labor mobilization on national issues such as privatization. ₉₃

During the 1970’s, the developmentalist policies of the mineiro agencies generated significant results. Through an array of infrastructural, fiscal, and financial mechanisms, the agencies successfully attracted and made possible more than $7 billion of new investment and more than 200,000 industrial jobs during the decade. ₉₄   Foreign investors flocked to Minas Gerais. Almost one-fourth of all foreign investment in Brazil between 1970 and 1977 went to Minas. ₉₅   The industrial structure of the state was transformed. By 1980, Minas was producing 20% of Brazil’s capital goods, whereas in 1970 it produced less than 8%, and this transition occurred when national expenditures on Minas were in decline compared with other states. ₉₆

The 1980’s would challenge the mineiro system in many new ways. First, democratization placed self-interested, traditional elites in the newly empowered seat of governor. Tancredo Neves, who was elected Minas’ governor in 1982 and would later run for president, largely ignored the state’s planning structure since he associated it with the previous authoritarian regime. ₉₇   Newton Cardoso, a self-styled “man of public works” with a political base among oligarchical, rural elites in northern Minas, became governor in 1986. Cardoso proceeded to attack the mineiro technocracy with political appointments and summary dismissals. Second, the economic crisis of the 1980’s eroded the agencies’ client base. Foreign and domestic investment slowed significantly during the decade. ₉₈   The larger fiscal crisis of Brazil’s developmentalist model was reflected in the decline of developmentalist policy in Minas Gerais.

The mineiro agencies survived these intense political and economic threats by exploiting their horizontal linkages. Interlinked by mutual ownership and the circulation of political technocrats, CEMIG, INDI, BDMG and other mineiro agencies maintained a cohesive core of elites who could not be easily rooted out of their positions during the Newton Cardoso administration. More important, these actors coalesced behind an economic diagnostic produced in 1989 by BDMG which became a blueprint for restructuring the mineiro planning system. In 1991, the election of Hélio Garcia, a former mayor of Belo Horizonte and governor of the state, provided an open political door to the mineiro political technocrats. Stunned by actual and threatened interventions by the Brazilian Central Bank in Minas’ state banks, and facing a deteriorating fiscal and productive structure, Garcia gave a carte blanche to a small group of political technocrats to rework the mineiro agencies. Garcia’s successor in 1994, Eduardo Azeredo, a political technocrat himself, further strengthened the role of political technocrats by placing prominent members of the class in the key economic positions of state government.

The new mineiro industrial policy system that emerged during the 1990’s emphasized the pursuit of economic externalities in Minas. Perhaps the most prominent example of success in this regard was the diversification of the auto manufacturing and supplier sector in Minas Gerais. Armed with a new array of fiscal incentives and infrastructure (funded in part by international financial agencies), the BDMG, INDI and CEMIG promoted the expansion of the state’s sole automobile producer, Fiat. The agencies financed the relocation and expansion of dozens of the automaker’s suppliers to locations proximate to the firm’s Betim plant. As a result, Fiat was able to launch a series of just-in-time links with its suppliers that greatly increased the firm’s domestic market share from 13.4% in 1980 and 24.5% in 1990 to 33.4% in 1994, making Fiat Brazil’s number one automaker in that year. The agencies also subtly opposed Fiat’s interests by promoting the expansion of a massive auto supplier production network in southern Minas that served São Paulo’s major automakers (GM, Ford, Volkswagen) and then by employing fiscal and infrastructural mechanisms, implemented through secret agency-firm negotiations, to lure a new Mercedes investment to the area in 1995. ₉₉  

Dozens more investments were made possible in underindustrialized areas of Minas by similar commitments of public support. In the chemical poles of the western part of the state and in industrial districts surrounding Belo Horizonte, agency technocrats attempted to create new industrial activities by prompting firms to move from São Paulo and create external linkages with sectors in the state and foreign capital looking for new locations in Brazil. Plant visits, meetings at the offices of BDMG and INDI and the use of financial and fiscal incentives managed by BDMG and the state government, informational resources organized by INDI, and infrastructural facilities maintained by CEMIG all became crucial mechanisms in the attraction of investment and the promotion of external economies. ₁₀₀   In these ways, horizontal ties preserved the state’s development agencies during the 1980’s and helped to innovate policy responses to the decline of developmentalism. Vertical ties between the mineiro agencies and private firms adjusted policy responses to fit economic opportunities and the development goals of the subnational policy system. These associations soon proved successful as Minas Gerais became, during the 1990’s, one of the fastest growing industrial regions in Brazil despite its underdevelopment and the country’s macroeconomic and fiscal problems.

São Paulo As the core of Brazilian industry, São Paulo’s state government developed few political interests to promote industrial investment. Rather, state efforts were more focused on relieving the urban congestion and high labor and land costs produced by the intense concentration of industries in the greater metropolitan area of the city of São Paulo. ₁₀₁ The militancy of increasingly autonomous labor unions during the late 1970’s and throughout the 1980’s and 1990’s produced additional problems for managing industrial investment in the state. These problems, however, did not cause the paulista leadership to develop their own technocratic institutions. São Paulo’s governors and major politicians during and after the transition to democracy became more concerned with their own short-term political interests in state and national politics. The most prominent among them, including Paulo Maluf (governor, 1979-82), Orestes Quércia (governor, 1987-90), and Luiz Antonio Fleury Filho (governor, 1991-94), among others, engaged in internecine political conflicts within and between political parties. These politicians distributed political rents and engaged in extensive fraud to buttress their individual bases of political support. One of the major effects of the state political class’ extensive patrimonialism was the generation of a severe fiscal crisis during the 1990’s which led to a massive $20 billion debt in the state bank, BANESPA. Paulista business and labor were unable to form an alternative support base for industrial policy, although there were attempts during the 1990’s. ₁₀₂   Due to these pressures and constraints, São Paulo was institutionally ill-prepared for dealing with its evolving problems in the distribution of industries and industrial relations.

In comparison to Minas Gerais, several factors account for this case of failed industrial policy: 1) the lack of strong political interests in favor of promoting investment; 2) the underdevelopment of institutions to implement industrial policy; and 3) the lack of fiscal resources to build an industrial policy. The fact that São Paulo’s major industrial development problems deal with a failure of urban planning and not underinvestment partly explains the lack of political interest in industrial promotion. Moreover, the short-sighted personalist interests of major paulista politicians explain why these failures of urban planning continue. The historical dominance of federal industrial policy, its subsequent deterioration during the 1980’s and 1990’s, the emerging fiscal crisis of the state government and the complexity of industrial restructuring also raised the costs of producing a technocratic and flexible set of institutions capable of dealing with the industrial development concerns of paulista business and labor.

Rio de Janeiro Although Rio de Janeiro state ₁₀₃   once retained more than 28% of Brazil’s industries, it gradually lost this share to São Paulo and other states, dropping to 15.6% in 1970. ₁₀₄   The state’s industrial activity declined precipitously during the economic crises of the 1980’s; the lowest growth rate for all of Brazil’s states during the early part of the decade. ₁₀₅   Like Andalucía, Rio’s industries were largely located in unproductive enclaves in steel, shipbuilding, and textiles. Rio’s political class, however, developed few strong political interests for promoting a more diversified industrial structure. Like the paulista political elite, Rio’s political leaders were obsessed with building and maintaining a clientelist network of power. During the height of the military government, Rio’s most prominent opposition politician, Antönio de Pádua Chagas Freitas, built an impressive political machine within the official opposition party, MDB, in Rio between 1965 and 1982. ₁₀₆   As in Minas Gerais, the transition to democracy empowered other traditional elites who sought to replicate Chagas Freitas’ system. Leonel Brizola, an old denizen of the Brazilian left, became governor in 1982 and again in 1990 based on his own network of political clientelism. Like his counterparts in São Paulo, Brizola thirsted for national power and thus ran for the Brazilian presidency perennially, following a mixture of nationalist and populist ideas. ₁₀₇   Under Brizola’s personalist rule, Rio could develop few lasting technocratic institutions for restructuring the state’s flagging industries.

The few that did emerge were dominated by Brizola’s brand of populism. The state government maintained an array of “industrial districts” managed by a Company of Industrial Development (CODIN). Brizola, and his successor between 1986-90, Moreira Franco, used the districts to subsidize land speculation and both manipulated CODIN through their political appointments. ₁₀₈   As a result, the CODIN could develop few workable linkages with national or state-level agencies, business or labor groups. Both business and labor were, as in Minas, politically weak in comparison to the state’s patrimonial network.

The travails of the CODIN industrial districts created between 1986 and 1990 illustrate the weaknesses of Rio’s industrial policy system. Moreira Franco’s major projects were the creation of industrial districts for fashion design, the production of optical equipment, footwear and computers. All were smashing failures. The fashion firms needed capital, something well beyond the domain of the CODIN. The lense- and frame-producing firms of the optical district were largely small, artesanal operations and, like the fashion designers, lacked capital. The footwear firms were poor, artesanal and operated primarily in the informal sector. Not only did they not have the money to transport their machinery to the site of the industrial district, but they feared that the state would prosecute them for working without a license if they were to setup shop in the government’s industrial districts. The computer growth pole was the greatest flop. The policy depended upon the continuity of the national informatics law that protected domestic producers. When President Fernando Collor eliminated the policy in 1990, just two years after the inauguration of the industrial district, the fluminense project collapsed. CODIN officials claimed, inexplicably, that they had compiled a history of success by selling these industrial districts to firms. However, a second look would show that although the land had been sold, (again, at subsidized prices), no new industrial activity had been initiated in these areas. As one observer claimed: “The government was subsidizing land speculation, not industrial development." ₁₀₉   In all of these cases, policy was formulated with the primary goal of extending the governor’s populist image. Few or no countervailing economic views could make such a policy more responsive to the industrial needs of Rio given the austere top-down nature of policy implementation. As a result, the economic rationale for CODIN’s industrial districts did not exist.

Brizola’s exit in 1994 and the incoming Marcello Alencar governorship led to the appointment of ex-National Development Bank political technocrats to CODIN’s directorate. Alencar backed the new CODIN with resources and political support in an effort to improve Rio’s decaying public image and economic structure. Despite this commitment, the current CODIN leadership is well aware of the state’s patrimonial system and the threats it poses to the emerging technocratic management of industrial policy in Rio de Janeiro. ₁₁₀   Their concerns verify the importance of building a horizontally embedded network capable of withstanding the infusion of parochial interests into the management of Rio’s new industrial policy.

Rio’s subnational industrial policy system lacked many of the factors that explained the success of industrial policy in Minas Gerais. Unlike the INDI-BDMG-CEMIG system in Minas, CODIN in Rio was institutionally isolated. It lacked the horizontal ties that allowed the mineiro system to survive several attempts at politicization. As a result, CODIN was easily manipulated by populist voices in Rio’s state government. Like IMADE in Madrid and the IFA in Andalucía, CODIN’s ties with firms were exclusively top-down. CODIN did not listen to the economic logic of their proposed industrial policies since they followed a more powerful political rationale. The fact that the present leadership of CODIN views the building of new relations with other technocratic organizations as a priority to guarantee its evolution beyond the term of the present governor is significant testimony for the importance of forging horizontal and vertical ties.

 

Conclusion

The structural, political, and institutional factors of my argument explain the organization and performance of the subnational cases discussed above. The distinct legacies of democratization in Spain and Brazil clarify how a gradual process of institutional change in the former could create a basis for the decentralization of resources and authorities to a developing political society. In Brazil, the process of decentralization was more ad hoc and tended to revitalize the power of traditional, patrimonial elites not a wider political society. In São Paulo and Rio de Janeiro, this factor helped to explain why institutions representing the interests of the subnational economy did not emerge and why the Asturian industrial policy system could rely on a more developed political society to maintain itself. The latter proved crucial since Asturian labor and its links with the regional Socialist party created a base for decentralizing national and EU resources to this secondary region. The Asturian agencies were also able to develop horizontal ties with labor once the political support of the post-Silva governments waned. Yet this structural factor does not explain how the mineiro system could create horizontal ties despite the weakness of political society.

Fiscal constraints and distinct levels of development also help to explain the range of opportunities and interests in the cases. Core regions such as Madrid and São Paulo shared few interests in promoting industrial investment and employment. In these cases, the major problems facing regional elites dealt with urban planning and investment distribution. Madrid could do more than São Paulo due to the Spanish capital’s relative flexibility of fiscal resources and less patrimonial group of political elites. However, both subnational systems failed to develop a lasting set of technocratic institutions for lack political backing and the ineffectiveness of efforts to forge horizontal associations. The study’s secondary regions—Andalucía, Asturias, Minas Gerais and Rio de Janeiro—all showed more political interest in developing industrial policies, but only Asturias and Minas Gerais, the most institutionally developed cases, performed relatively well. Also, despite fiscal constraints in Minas, the mineiro political leadership was able to reorganize the system and intervene in the market selectively. The failure of the subnational industrial policy systems in both Andalucía and Rio emphasized the importance of the interests of regional politicians to support development agencies and for these agencies to develop lasting associations with other official actors, political society and client firms. None of these factors were as salient in Andalucía and Rio as they were in Asturias and Minas. Fiscal constraints and level of development are relevant structural constraints on subnational politics, but consideration of the latter is needed to offer complete explanations for observed differences among the cases.

Firm interests and globalization also intervened in all of the subnational cases. The demonstrated ability of Asturias and Minas Gerais to listen to multinational interests and offer an effective set of incentives suggests that firm interests guide investment strategies but do not entirely determine outcomes. If even subnational governments can substantially affect multinational firm interests, then the power of globalization to decide development outcomes is greatly exaggerated. Subnational governments within shifted states matter.

The strength of the political support offered industrial policy systems in Asturias and Minas Gerais and the relative weaknesses of such support in São Paulo, Rio de Janeiro, Andalucía and Madrid, were crucial for the formation of these systems. Silva’s support for the development agencies in Asturias as part of his interests in building a base for Asturian autonomy and Garcia’s and Azeredo’s interests in re-empowering the mineiro political technocracy point to the importance of political support. This crucial factor was missing in the four other cases. Partial systems emerged or reemerged in Madrid, Andalucía and Rio due to moderate or weakening political support at different times. Yet in São Paulo, where political interests never supported subnational development policy, absolutely no significant institutions emerged.

The cases of Asturias and Minas also show that political support is not enough for maintaining a subnational industrial policy. Particular aspects of the development institutions in these systems explain their perseverance. Horizontal associations between Asturian labor unions, client firms, business elites and the development agencies determined how the Asturian industrial policy system could survive the political exit of Silva and the decline of national funding. Likewise, the developmentalist mineiro agencies survived both economic and political threats due to horizontal ties among themselves. These links became the basis for revitalizing the system through the 1989 BDMG diagnostic. The isolated and politicized IFA in Andalucía, IMADE in Madrid and CODIN in Rio, illustrate further how horizontal associations help to explain the relative flexibility of subnational industrial policy systems and their subsequent maintenance. Table 1.1 illustrates how the mix of political support and embeddedness differentiated the ability of each of the six subnational policy systems to respond to economic challenges.

Finally, the cases of Asturias and Minas Gerais show that vertical ties between agencies and client firms affect the economic performance of official projects. In these two cases, literally an ocean apart and facing distinct structural challenges, cases of successful intervention were characterized by close exchanges of information and a high level of monitoring between subnational development agencies and firms. The cases of DuPont and the small- and medium-sized firms of the study in Asturias and Fiat, Mercedes and their suppliers in Minas Gerais provide indicative cases.

Comparison of Minas and Asturias with the four unsuccessful subnational policy systems shows that vertical associations are unlikely to form or will be manipulated by political interests without a strong contribution by the other dimension of embeddedness, horizontal associations.

Table 1.1 Summary of the Six Subnational Policy Systems

Political Support
Embeddedness	High	Low
High     (On-balance consistency in horizontal and vertical ties)	Minas Gerais (1991-present)/Asturias (1982-1990) Structural incentives: high Political support: high Political Priority: institutions Horizontal ties: strong Vertical ties: strong	Minas Gerais (1984-1990) Structural incentives: high Political support: weak Political priority: patronage Horizontal ties: strong Vertical ties: strong     Asturias (1991-present) Structural incentives: high Political support: moderate/weak Political priority: short-term electoral interests Horizontal ties: strong Vertical ties: strong
Low       (On-balance inconsistency in horizontal and vertical ties)	Rio de Janeiro/Andalucía Structural incentives: high Political support: high Political priority: patronage Horizontal ties: weak Vertical ties: strong in some cases; development policy system undermined overall	São Paulo/Madrid Structural incentives: low Political support: low Political priority: patronage/short-term political interests Horizontal ties: weak Vertical ties: weak

In Andalucía, for example, strong vertical ties did develop between the IFA and domestic and foreign firms. These relations, however, were more top-down than they were in Asturias and Minas Gerais. Andalusian government officials dominated the investment strategies of client firms through a public demiurge approach. Such relations worsened the politicization of the Andalusian subnational industrial policy system by excluding from view the productive and investment concerns of market actors. The same was true of Rio’s CODIN which followed a similar top-down approach, driven by the populist politics of several governors. Rather than the emergence of systems of information exchange and monitoring which would have detected abuses, the old CODIN system created an exploitative and unsustainable sequence of investments.

Given the complexity of political decentralization and its differentiated effects on subnational industrial policy in Spain and Brazil, it is impossible to claim that there is a single model of “efficient” federalism or a single model of “economic governance” that explains the organization and performance of these subnational cases. Rather, only by studying the way that the politics and institutions of subnational government change can we appreciate the richness of shifted states. We must first, however, do away with our disciplinary tendencies to associate analytical and political “strength” with models that favor political centralization.

 

---

Endnotes

Note 1: Alfred Marshall, The Principles of Economics (1920); Albert O. Hirschman, The Strategy of Economic Development, (New Haven: Yale University Press, 1958); Gunnar Myrdal, Economic Theory and Under-Developed Regions, (London: Gerald Duckworth & Co.).  Back.

Note 2: Richard M. Locke, “The Composite Economy: Local Politics and Industrial Change in Contemporary Italy,” Economy and Society, 25:4, (November, 1996), p. 484.  Back.

Note 3: Paul Krugman, Geography and Trade (Cambridge: MIT Press, 1991), p. 99. Krugman speaks for many economic geographers when he notes that the modern neglect of the spatial dimensions of investment in contemporary economics is all the more surprising given that Alfred Marshall introduced the essential economic concept of “externalities” in the context of developing a theory on the localization of industries. Economic and human geographers, of course, did not ignore the spatial unevenness of development. See Allen J. Scott and Michael Storper, “Industrial Change and Territorial Organization: A Summing Up,” in A. J. Scott and M. Storper, eds., Production, Work, Territory: The Geographical Anatomy of Industrial Capitalism (Boston: Allen & Unwin, 1986); and Michael Storper and Richard Walker, The Capitalist Imperative: Territory, Technology and Industrial Growth (New York: Basil Blackwell, 1989).  Back.

Note 4: Comparisons of the “successful” East Asian cases of export-oriented industrialization versus the “unsuccessful” Latin American cases of import-substitution industrialization underscored these arguments. See Gary Gereffi and Donald L. Wyman, eds. Manufacturing Miracles: Paths of Industrialization in Latin America and East Asia (Princeton: Princeton University Press, 1990).  Back.

Note 5: Robert Wade, Governing the Market: Economic Theory and the Role of Government in East Asian Industrialization (Princeton: Princeton University Press, 1990); Chalmers Johnson, MITI and the Japanese Miracle: The Growth of Industrial Policy, 1925-1975 (Stanford: Stanford University Press, 1982); Alice Amsden, Asia’s Next Giant: South Korea and Late Industrialization (New York: Oxford University Press, 1989); Frederic C. Deyo, ed., The Political Economy of the New Asian Industrialism (Ithaca: Cornell University Press, 1987).  Back.

Note 6: Peter B. Evans, Embedded Autonomy: States and Industrial Transformation (Princeton: Princeton University Press, 1995). Evans’ approach attempted to correct the tendency of strong state arguments to champion the “insulation” of the technocratic elite by claiming that “a state that was only autonomous would lack both sources of intelligence and the ability to rely on decentralized private implementation” (p. 12). Yet, once again, the main protagonist of Evans’ “embedded autonomous” state was the bureaucratically developed central state linked narrowly with business elites; the East Asian model redux.  Back.

Note 7: Stephan Haggard and Robert R. Kaufman, The Political Economy of Democratic Transitions (Princeton: Princeton University Press, 1995), p. 370. These authors were clearly influenced by the arguments of Samuel Huntington, Political Order in Changing Societies (New Haven: Yale University Press, 1968).  Back.

Note 8: Some of the key works in fiscal federalist studies are Richard A. Musgrave and Peggy B. Musgrave, Public Finance in Theory and Practice, 5th Ed. (New York: McGraw-Hill, 1989); Wallace E. Oates, Fiscal Federalism (New York: Harcourt Brace Jovanovich, 1972). Recent Inter-American Development Bank reports have applied “fiscal federalist” arguments to developing countries. See Inter-American Development Bank, Economic and Social Progress in Latin America, 1994 Report. Special Report: Fiscal Decentralization (Washington, D.C.: IDB/Johns Hopkins University Press, 1994). On the “subsidiarity principle” in the EU see Jacques Pelkmans, “Governing European Union: From Pre-Federal to Federal Economic Integration?,” in Karen Knop, Sylvia Ostry, Richard Simeon, and Katherine Swinton, eds., Rethinking Federalism: Citizens, Markets and Governments in a Changing World. Vancouver: University of British Columbia Press, 1995).  Back.

Note 9: Richard M. Locke, Remaking the Italian Economy (Ithaca: Cornell University Press, 1995); Sebastiano Brusco, “The Emilian Model: Productive Decentralisation and Social Integration,” Cambridge Journal of Economics, 6 (1982); Sebastiano Brusco and Ezio Righi, “Local Government, Industrial Policy and Social Consensus: The Case of Modena (Italy),” Economy and Society, 18:4, (November, 1989); and Frank Pyke, Giacomo Becattini and Werner Sengenberger, Industrial Districts and Inter-Firm Cooperation in Italy (Geneva: International Institute for Labour Studies, 1990).  Back.

Note 10: Gary Herrigel, Reconceptualizing the Sources of German Industrial Power (New York: Cambridge University Press, 1995).  Back.

Note 11: AnnaLee Saxenian, Regional Advantage: Culture and Competition in Silicon Valley and Route 128 (Cambridge: Harvard University Press, 1994); and Manuel Castells, The Informational City: Information Technology, Economic Restructuring, and the Urban-Regional Process (Cambridge: Blackwell, 1989).  Back.

Note 12: Studies of this kind are predictably scarce. One exception is Harald Fuhr, “Mobilizing Local Resources in Latin America: Decentralization, Institutional Reforms and Small-Scale Enterprises,” in Brigitte Spath, ed., Small Firms and Development in Latin America: The Role of the Institutional Environment, Human Resources and Industrial Relations (Geneva: International Institute for Labour Studies, 1993). Fuhr’s preliminary work confirms the existence of formidable political and economic obstacles to the broad application of the Third Italy model to developing countries.  Back.

Note 13: Although her approach to subnational policymaking is a bit more nuanced than the approach used here, Judith Tendler also emphasizes the importance of mutual monitoring that close and consistent relations between policymakers and private clients facilitates. See Judith Tendler, Good Government in the Tropics, (Baltimore: Johns Hopkins University Press, 1997), pp. 139-140.  Back.

Note 14: Stephen S. Cohen and John Zysman, Manufacturing Matters: The Myth of the Post-Industrial Economy (New York: Basic Books, 1987); Fernando Fajnzylber, Unavoidable Industrial Restructuring in Latin America (Durham: Duke University Press, 1990).  Back.

Note 15: Evans (1995), p. 18.  Back.

Note 16: This approach is similar to Locke’s “micropolitical” approach to Italy. Locke also focuses on “the politics of strategic choice to explain variation within states...", p. 181. similarly, institutions governing economic policy and political decentralization evolve historically and are shaped by continuing political processes and events. this treatment of institutions differs from the “rational choice,” organization theory of institutions which depicts institutions as static entities constraining the actions of political actors. see kathleen thelen and sven steinmo, “historical institutionalism in comparative politics,” in sven steinmo, kathleen thelen, and frank longstreth, eds., structuring politics: historical institutionalism in comparative perspective (new york: cambridge university press, 1992).  Back.

Note 17: I define “structural” as national-institutional and socio-economic constraints.  Back.

Note 18: My argument here is based on Leonardo Parri’s “territorial political exchange” analytical framework. See Leonardo Parri, “Territorial Politics and Political Exchange: American Federalism and French Unitarianism Reconsidered,” in Bernd Marin, ed., Governance and Generalized Exchange: Self-Organizing Policy Networks in Action (Boulder: Westview, 1990). A more systematic version of this argument is found in Susan Shirk, The Political Logic of Economic Reform in China (Berkeley: University of California Press, 1993).  Back.

Note 19: Stepan defines “political society” as “that arena in which the polity specifically arranges itself for political contestation to gain control over public power and the state apparatus". See Alfred Stepan, Rethinking Military Politics: Brazil and the Southern Cone (Princeton: Princeton University Press, 1988), p. 4. Although Stepan operationalizes the term to intra-governmental lobbies and legislative politics, I will expand the meaning of the term in the political economic realm to include business, labor, and consumer organizations.  Back.

Note 20: For an articulation of this argument and a comparison to the contrasting case of Brazil, see Guillermo O’Donnell, “Transitions, Continuities, and Paradoxes,” in Scott Mainwaring, Guillermo O’Donnell, and J. Samuel Valenzuela, eds., Issues in Democratic Consolidation: The New South American Democracies in Comparative Perspective (Notre Dame: University of Notre Dame Press, 1992).  Back.

Note 21: Ibid.  Back.

Note 22: This argument is posited by Allan Tupper, “Federalism and the Politics of Industrial Policy,” in André Blais, ed., Industrial Policy (Toronto: University of Toronto Press, 1986).  Back.

Note 23: R. J. Johnston, “The State, The Region, and the Division of Labor,” in A. J. Scott and M. Storper, eds. Production, Work Territory: The Geographical Anatomy of Industrial Capitalism (Boston: Allen & Unwin, 1986).  Back.

Note 24: Storper and Walker, p. 84.  Back.

Note 25: Ray Hudson and David Sadler, “Contesting Works Closures in Western Europe’s Old Industrial Regions: Defending Place or Betraying Class?” in A. J. Scott and M. Storper, eds., Production, Work, Territory: The Geographical Anatomy of Industrial Capitalism (Boston: Allen & Unwin, 1986).  Back.

Note 26: Hirschman, p. 185.  Back.

Note 27: Storper & Walker refer to these firm needs as “locational specifications”—labor, natural resources and consumer markets. The authors also add the division of labor within and between industries, linkages with suppliers and other firms in the input-output chain. The authors argue that “locational capabilities” are distinct from “locational specifications” in the sense that the former refers to the capacity of a firm to secure its needs at a given location. In this way, firm needs are linked to level of development (my third structural factor). See Storper and Walker, p. 73.  Back.

Note 28: Regional policies which affect factor prices will be particularly effective in attracting industries facing increased sectoral competition. See Anne Markusen, Profit Cycles, Oligopoly, and Regional Development (Cambridge: MIT Press, 1985).  Back.

Note 29: This assumption is based on Barbara Geddes, Politician’s Dilemma: Building State Capacity in Latin America (Berkeley: University of California Press, 1994), pp. 7-8.  Back.

Note 30: Ibid.  Back.

Note 31: This conception of technocrats and political-technocrats follows the distinction between técnicos and political técnicos in Ben Ross Schneider, Politics Within the State: Elite Bureaucrats and Industrial Policy in Authoritarian Brazil (Pittsburgh: University of Pittsburgh Press, 1991).  Back.

Note 32: The term “embeddedness” is used here much as Evans employs the term to describe the official linkages between central state agencies and business organizations. See Evans (1995).  Back.

Note 33: Peter B. Evans, “The State as Problem and Solution: Predation, Embedded Autonomy and Structural Change,” in Stephan Haggard and Robert R. Kaufman, eds. The Politics of Economic Adjustment: International Constraints, Distributive Conflicts, and the State (Princeton: Princeton University Press, 1992), p. 176.  Back.

Note 34: Both frequency and familiarity seem recursively related, and in most of the empirical cases in this study, they are. Conceptually, however, they are different to the extent that frequency of contact might not necessarily breed familiarity if the knowledge transmitted between agency and firm during visits and meetings is not substantial. Familiarity, on the contrary, assumes a substantial amount of information exchange.  Back.

Note 35: Paul Krugman, “Targeted Industrial Policies: Theory and Evidence,” in Industrial Change and Public Policy: A Symposium Sponsored by the Federal Reserve Bank of Kansas City, (Jackson Hole, Wyoming, August 24-26, 1983), pp. 140-141.  Back.

Note 36: Implicit in this formulation is the difference between a “high road” and a “low road” to competitiveness. In the “high road” scenario, improvements to production attempt to maximize the quality of products and innovation at the same time that labor quality is integrated as a source of productivity. By contrast, the “low road” scenario depicts a competitive strategy based on cost factors. For more on this literature, see Babson (1995).  Back.

Note 37: Peter B. Evans, Dependent Development: The Alliance of Multinational, State, and Local Capital in Brazil (Princeton: Princeton University Press, 1979), pp. 89-90.  Back.

Note 38: It should be noted that all Spanish public firms were not governed by INI. After 1970, two more holdings—the Dirección General del Patrimonio del Estado (the General Direction of the Patrimony of the State—DGPE), a collection of film, communications, food and tobacco companies—and the Instituto Nacional de Hidrocarburos (National Institute of Hydrocarbons—INH) a collection of energy firms, were created. INI, however, still remained the most important public sector holding company. This paper will focus mainly on INI firms since they are representative of the crisis of the Spanish public sector as a whole.  Back.

Note 39: Pablo Martín Aceña and Francisco Comín, INI: 50 años de industrialización en España (Madrid: Espasa Calpe, 1991), pp. 78-82.  Back.

Note 40: Pablo Martín Aceña and Francisco Comín, “La acción regional del Instituto Nacional de Industria, 1941-1976,” in Pautas regionales de la industrialización española (Siglos XIX y XX) (Barcelona: Editorial Ariel, 1990).  Back.

Note 41: Wilson Cano, Raízes da Concentração Industrial em São Paulo (Rio de Janeiro: Difel, 1977).  Back.

Note 42: Wilson Cano, “Concentración, desconcentración y descentralización en Brasil,” in José Luis Curbelo, Francisco Albuquerque, Carlos A. de Mattos and Juan Ramón Cuadrado, eds., Territorios en transformación: análisis y propuestas (Madrid: Fondo Europeo de Desarrollo Regional, 1994), pp. 232-233.  Back.

Note 43: Luiz Carlos Bresser Pereira, “Economic Reforms and Economic Growth: Efficiency and Politics in Latin America,” in Luiz Carlos Bresser Pereira, José María Maravall and Adam Przeworski, eds., Economic Reforms in New Democracies: A Social-Democratic Approach (New York: Cambridge University Press, 1993).  Back.

Note 44: Antönio Barros de Castro, “Renegade Development: Rise and Demise of State-Led Development in Brazil,” in William C. Smith, Carlos H. Acuña & Eduardo A. Gamarra, eds. Democracy, Markets, and Structural Reform in Latin America (New Brunswick: Transaction Publishers, 1994), p. 201; and Rogério Werneck, Empresas Estatais e Política Macroeconömica (Rio de Janeiro: Editora Campus, 1991).  Back.

Note 45: DESEP, Indicadores DESEP (São Paulo: DESEP/CUT, 1994), p. 43.  Back.

Note 46: Wilson Suzigan, “Política Comercial e Perspectivas da Indústria Brasileira,” UNICAMP Discussion Paper N° 13 (December), 1992.  Back.

Note 47: Luís Angel Rojo, “La crisis de la economía española, 1973-1984,” in Jordi Nadal and Jaume Torras, eds., La economía espanola en el siglo XX: una perspectiva historica (Barcelona: Editorial Ariel, 1987), p. 196.  Back.

Note 48: Anuario de Economía: 1995 Negocios (Madrid: Iberdrola, 1995), p. 258, Table  Back.

Note 49: These goals were institutionalized in the Pacts of Moncloa in 1977. Among some of the anti-inflationary policies contained in the Moncloa Pacts were a tight monetary policy and reductions in public expenditures on pensions, social security and the civil service. Enrique Fuentes Quintana, “De los pactos de la Moncloa a la Constitución (julio 1977-diciembre 1978),” in José Luis García Delgado, ed., Economía española de la transición y la democracia, (Madrid: Centro de Investigaciones Sociologicas, 1990); and Joan Trullen i Thomas, Fundamentos economicos de la transición política española: la política economica de los acuerdos de la Moncloa (Madrid: Ministerio de Trabajo y Seguridad Social, 1994).  Back.

Note 50: Author interview with Ignácio Bayón, Centrist Minister of Industry (1980-1982), July 7, 1994, Madrid; and Otto Holman, “In Search of Hegemony: Socialist Government and the Internationalization of Domestic Politics in Spain,” International Journal of Political Economy 19, 3 (Fall) 1989, p. 95.  Back.

Note 51: Author interview with Álvaro Espina, former Secretary of State for Industry, Ministry of Industry, June 29, 1994, Madrid; and author interview with Fernando Maravall, former General Technical Secretary, Ministry of Industry, and adviser to the Socialist Minister of Industry, Carlos Solchaga (1983-1986), June 27, 1994.  Back.

Note 52: Mikel Navarro Arancegui, Política de reconversión: balance crítico (Madrid: EUDEMA, 1990).  Back.

Note 53: Martín Aceña and Comín, p. 542.  Back.

Note 54: Donald Share, Dilemmas of Social Democracy: The Spanish Socialist Workers Party in the 1980s (New York: Greenwood Press, 1989), p. 73.  Back.

Note 55: At the same time, however, the military’s 1967 tax reform reduced the autonomy of the states to tax. This increased the dependence on debt to finance local development. By one estimate, in 1981 over one-third of Brazil’s total external indebtedness was held by the state governments. See Lawrence S. Graham, The State and Policy Outcomes in Latin America, (New York: Praeger, 1990), p. 87.  Back.

Note 56: Wayne A. Selcher, “O Futuro do Federalismo na Nova República,” Revista de Administração Pública 24:1 (1990).  Back.

Note 57: Ibid, p. 179.  Back.

Note 58: Fernando Luiz Abrúcio, Os Barões da Federação: O Poder dos Governadores no Brasil Pós-Autoritário, Unpublished master’s thesis, Universidade de São Paulo (1994); Frances Hagopian, Traditional Politics and Regime Change in Brazil, (Cambridge: Cambridge University Press, 1996).  Back.

Note 59: H. L. Campos, Os Interesses Regionais na Assembléia Nacional Constituente, Unpublished master’s thesis, Universidade de Campinas (1991).  Back.

Note 60: Anwar Shah, The New Fiscal Federalism in Brazil, World Bank Discussion Papers, N° 124 (1991), p. 17.  Back.

Note 61: José Roberto Rodrigues Afonso, “Descentralização Fiscal: Revendo Idéias,” Ensaios FEE, Porto Alegre 15:2 (1994), p. 356.  Back.

Note 62: The degree of control over the transition by the military in Brazil compared to the relative flexibility of the Spanish transition, the historical emergence of a strong political society in Spain compared with the reign of personalism in Brazil, especially among the most conservative segments of the opposition, are some of the prominent factors that explain these differences. See O’Donnell, pp. 29-31.  Back.

Note 63: Victor Pérez-Díaz, “Gobernabilidad y mesogobiernos: autonomías regionales y neocorporatismo en España,” Papeles de Economía Española N° 21 (1984), p. 49.  Back.

Note 64: Joan Subirats Humet, “Evaluación del processo autonómico: algunos indicadores,” in A.F. Laraudogoitia & E.M. Artaraz, eds., Poder Político y Comunidades Autónomas, (Vitoria-Gasteiz: Eusko Legebiltzarra Parlamento Vasco, 1991), p. 107.  Back.

Note 65: A. Ballesteros, “La función pública de las comunidades autónomas,” in S. Martin-Retorrillo, ed., Pasado, presente y futuro de las comunidades autónomas (Madrid: Instituto de Estudios Económicos, 1989), p. 176.  Back.

Note 66: The governments of Basque and Navarra, for example, asserted their historic “foral” rights in taxing and financing themselves. Catalonia, Valencia, Andalucía and Galicia gained additional control over education and other social expenditures under the “rapid route” (vía rápida) to autonomy. The other Spanish regions were placed in the “slow track” (vía lenta) which provided a less extensive array of capacities, yet a significant number.  Back.

Note 67: José María Serrano Sanz, “La intervención económica en el Estado de las Autonomías, in José Luis García Delgado, ed., España: Economía (Madrid: Espasa-Calpe, 1988), pp. 511-512.  Back.

Note 68: This is at best a cursory estimate. See El País August 8, 1988.  Back.

Note 69: CREP (Comisión de Coordinación de los Representantes del Principado en la Empresa Pública), Informe sobre la empresa pública industrial del INI en el Principado de Asturias: antecedentes y perspectivas de futuro, Unpublished document, (December 1992), p. 21.  Back.

Note 70: Author interview with Pedro de Silva Cienfuegos-Jovellanos, June 26, 1995, Gijón, Asturias.  Back.

Note 71: Several government documents highlight the link the regional government made between the building of regional autonomy and the expansion of its authority over industrial policy. See especially Consejería de Hacienda y Economía, Programa de Desarrollo Regional 1985-88 (Oviedo: Principado de Asturias, 1986), p. 135.  Back.

Note 72: Based on author interviews with Asturian union, government and business leaders, May-June 1995 and March 1996. On DuPont’s decision-making concerning the Asturian deal, see El País, (January 28, 1990); Química Hoy (January, 1992); Chemical and Engineering News (June 11, 1990).  Back.

Note 73: Numerous studies by Spanish economists have established that information asymmetries create prohibitive costs for banks seeking to finance small firms. As a result, banks require high collateral and net worth, conditions that small firms rarely satisfy. Small firm projects are often riskier, thus elevating the costs of study for the banks and the costs of providing requisite preliminary collateral for the firms. These costs create significant obstacles to small- and medium-sized firms seeking finance in Spain. See Manuel Martín Rodríguez, “El sistema financiero y la financiación de las pyme,” Papeles de Economía Española N° 65, (1995); and Ramón Caminal, “El papel de las restricciones de crédito y las políticas públicas en la financiación de la pequeña y mediana empresa,” Papeles de Economía Española N° 65, (1995).  Back.

Note 74: Marisol Esteban and Roberto Velasco, “La política industrial en la España de las autonomías,” Papeles de economía española N° 67 (1996), p. 295.  Back.

Note 75: A more detailed description of the study is not possible here. See Alfred P. Montero, Shifting States in Uneven Markets: Political Decentralization and Subnational Industrial Policy in Contemporary Spain and Brazil, Ph. D. Dissertation, Department of Political Science, Columbia University (1997), Chapter 4.  Back.

Note 76: The functions performed by the agencies were widely viewed by Asturian firms as extremely valuable. This is confirmed by a recent study of 198 Asturian firms. See Manuel Castells and Juan Antonio Vázquez, Estrategias para la reindustrialización de Asturias (Madrid: Editorial Civitas, 1994), pp. 652-653.  Back.

Note 77: Based on numbers in several studies. See Paz Benito del Pozo, “La industria en Asturias: un declive prolongado,” in Joaquín Bosque and Ricardo Méndez, eds., Cambio industrial y desarrollo regional en España (Barcelona: Oikos-Tau, 1995), p. 131; and CREP.  Back.

Note 78: SAYPE, Memoría de Actividades (1991-1993), multiple issues.  Back.

Note 79: Luis J. Sanchez Ortiz, “Madrid, capital del capital,” Economía y Sociedad N° 4 (December, 1990); Ricardo Méndez, “La nueva industria en la Comunidad de Madrid,” in Joaquín Bosque and Ricardo Méndez, eds., Cambio industrial y desarrollo regional en España (Barcelona: Oikos-Tau, 1995), p. 415.  Back.

Note 80: Madrid exported only 6% of its industrial production in 1989. That rate is half the national average for regional economies. See José Luis García Delgado and Cándido Muñoz Cidad, “La industria de Madrid de la segunda mitad del decenío de 1980,” Economía industrial N° 279/280 (May/August, 1991), p. 213.  Back.

Note 81: José Luis García Delgado and Juan I. Palacio Morena, “La estructura industrial de Madrid: bases formativas, evolución y transformaciones recientes,” Economía industrial N° 263/264 (September/December 1988), p. 229.  Back.

Note 82: Author interview with Gloria Ruíz Camacho, Economist for Avalmadrid, May 14, 1996, Madrid.  Back.

Note 83: Author interview with José Javier Peña Linares, Director of Information Services, IMADE, May 13, 1996, Madrid.  Back.

Note 84: For more on this episode, see El País, (October 30), 1993.  Back.

Note 85: Regional politicians accused high-level administrators of subsidizing firms in which they had a private interest. The accusations developed into a full-blown investigation that uncovered a web of corruption far larger than what had been originally suspected. See El Mundo, (October and November), 1994 on the case of Ibercoop.  Back.

Note 86: Reports of this poll were published in El País, (June 24), 1987. My own preliminary discussions with businesses in Madrid reflected these opinions. Managers of Espindesa, for example, a chemical firm that had received a small subsidy through IMADE in 1991, told me that IMADE’s activities were poorly organized and had no effect on production or employment in their case. Agency officials, I was told, showed little interest with Espindesa’s concerns.  Back.

Note 87: Michael Barzelay and José María O’Kean, Strategic Public Management in Andalusia: Concepts, Experience and Analysis (Madrid: Instituto de Estudios Fiscales, 1989).  Back.

Note 88: Author interviews with IFA officials, April 1996, Seville, Andalucía; several issues of El País, Andalucía section (1986-1996).  Back.

Note 89: Several issues of El País, Andalucía section, (1994); and El Mundo (1994).  Back.

Note 90: El País, August 19, 1994.  Back.

Note 91: See Luis Aureliano Gama de Andrade, Technocracy and Development: The Case of Minas Gerais, Ph. D. dissertation, University of Michigan (1980); and Clélio Campolina Diniz, Estado e Capital Estrangeiro na Industrialização Mineira (Belo Horizonte: PROED, 1981).  Back.

Note 92: Hagopian, pp. 87 & 124; and Schneider, Chapter 6.  Back.

Note 93: Montero, Chapter Seven.  Back.

Note 94: Clélio Campolina Diniz, “Economía e Planejamento em Minas Gerais,” Revista de Estudos Políticos N° 58 (January 1984), p. 267.  Back.

Note 95: Paulo Brant, “O Setor Industrial em Minas Gerais: Características, Desempenho Recente e Perspectivas,” in II Seminário sobre a Economia Mineira: Historia Econömica de Minas Gerais; A Economia Mineira dos Anos 80 (Belo Horizonte: Diamantina/CEDEPLAR, 1983), p. 322.  Back.

Note 96: Diniz (1981), p. 214; and BDMG, Economia Mineira 1989: Diagnóstico e Perspectivas, Volume I (Belo Horizonte: Government of Minas Gerais, 1989), p. 33.  Back.

Note 97: Author interview with Marilena Chaves, Assistant Secretary of Planning, SEPLAN, June 14, 1996, Belo Horizonte, Minas Gerais.  Back.

Note 98: BDMG, p. 28.  Back.

Note 99: Based on author interviews with INDI and mineiro government officials; and various issues of Diário do Comércio (1995-1996).  Back.

Note 100: See Montero, Chapter Seven.  Back.

Note 101: Veja (November 9) 1994.  Back.

Note 102: The most prominent was the “sectoral chambers” experiment in tripartite bargaining. See Eli Diniz, “Reformas Econömicas e Democracia no Brasil dos Anos 90: As Câmaras Setoriais como Fórum de Negociação,” Dados 37:2 (1994); and Scott Martin, “Beyond Corporatism: New Patterns of Representation in the Brazilian Auto Industry,” in Douglas A. Chalmers, Carlos M. Vilas, Katherine Hite, Scott B. Martin, Kerianne Piester, and Monique Segarra, eds. The New Politics of Inequality in Latin America: Rethinking Participation and Representation, (New York: Oxford University Press, 1997).  Back.

Note 103: Guanabara (the contemporary city of Rio de Janeiro) was the federal district and the administrative capital of Brazil until the early 1960’s when the capital was moved to Brasília. During the mid 1970’s, Guanabara was integrated into the surrounding area, thus creating the contemporary state of Rio de Janeiro.  Back.

Note 104: Hildete Pereira de Melo and Cláudio Monteiro Considera, “Industrialização Fluminense 1930/1980,” Revista do Rio de Janeiro 1:3 (May/August 1986), p. 113.  Back.

Note 105: Clélio Campolina Diniz and Fabiana Borges Teixeira dos Santos, “Região Sudeste: Desempenho Econömico, Heterogeneidade Estrutural e Perspectivas,” in Evolução e Perspectivas das Desigualdades Regionais no Desenvolvimento Brasileiro (São Paulo: FUNDAP, 1994), Table 2.  Back.

Note 106: Eli Diniz, Voto e Máquina Política: Patronagem e Clientelismo no Rio de Janeiro (Rio de Janeiro: Paz e Terra, 1982).  Back.

Note 107: Thomas E. Skidmore, The Politics of Military Rule in Brazil, 1964-85 (New York: Oxford University Press, 1988), pp. 218-220 & 233-235.  Back.

Note 108: Author interviews with CODIN officials, May-June, 1996.  Back.

Note 109: Author interviews at CODIN, May 1996, Rio de Janeiro.  Back.

Note 110: Author interview with Marco Antonio de Araújo Lima, CODIN president, June 5, 1996, Rio de Janeiro.  Back.