Central banking, democratic governance and political authority: the case of Brazil in a comparative perspective

Banco central, governança democrática e autoridade política: o caso do Brasil em uma perspectiva comparativa

ABSTRACT

By examining the relationship between democratic governance and monetary authority this paper accomplishes two principal tasks. First, it justifies placing the study of monetary authority as a central item on the research agenda of political scientists. Beginning from the premise central banks constitute a special mode of political authority, we examine trade-offs between questions of transparency, democratic accountability, and public sector efficiency. Second, by conducting an empirical study of monetary authority for the Brazilian case, the paper inverts a commonly held assumption within the study of central banks. Rather than argue price stability follows from an autonomous central bank, the Brazilian case demonstrate nearly the opposite can take place.

KEYWORDS:
Central bank; central bank autonomy; monetary policy; political economy; accountability

RESUMO

Ao examinar a relação entre governança democrática e autoridade monetária, este artigo realiza duas tarefas principais. Primeiro, justifica colocar o estudo da autoridade monetária como um item central na agenda de pesquisa de cientistas políticos. Partindo da premissa de que os bancos centrais constituem um modo especial de autoridade política, examinamos as compensações entre questões de transparência, responsabilidade democrática e eficiência do setor público. Segundo, ao realizar um estudo empírico da autoridade monetária para o caso brasileiro, o artigo inverte uma suposição comum no estudo dos bancos centrais. Em vez de argumentar que a estabilidade de preços decorre de um banco central autônomo, o caso brasileiro demonstra que quase o oposto pode acontecer.

PALAVRAS-CHAVE:
Banco central; autonomia do banco central; política monetária; economia política; accountability

1. INTRODUCTION

With the recent transition to democracy in developing countries nearly complete, analysts have turned toward examining the conditions for effective democratic governance. Central to this research agenda has been a study of factors contributing toward democratic accountability, transparency, and public sector efficiency. Only by meeting those conditions will nascent democracies have a chance to fully consolidate their political institutions. This paper draws attention to a critical aspect of democratic governance which has so far been ignored in the literature-monetary authority. Despite the obvious and essential impact monetary authority has upon the political economy of developing countries, with notable exceptions the study of central banks has been largely left to economists.

We attempt to accomplish two tasks in this paper. First, we provide a justification for why the study of monetary authority and central banks should be an essential item on the research agenda of democratic governance for political scientists. In order to effectively discuss the relationship between democratic governance and monetary authority, however, one must expand the conceptual framework used to study central bank institutions. Specifically, analysis should shift away from an exclusive focus on how central banks attain autonomy from the political arena toward a study of central banks as a special mode of political authority. Only through doing so can we discern the specific trade-offs within monetary authority regarding questions of transparency, democratic accountability, and public sector efficiency.

Our second task is to suggest what an empirical study of monetary authority looks like given central banks are conceptualized as a mode of political authority. Drawing upon the Brazilian case, we invert a common assumption within the study of central banks; that price stability follows from an autonomous central bank. For the case of Brazil, we demonstrate nearly the opposite is taking place. Only with price stabilization and success of the recent Real Plan have the conditions for an autonomous central bank been created. The end of inflation weakened the very actors opposed to a centralization of monetary authority in the central bank, thus paving the way for its eventual autonomy from short term political pressures.

Price stability is of course a process still under way in Brazil, dependent upon the success of reforms geared to changing the pattern of financing State expenditures. Yet it is possible to show that price stability is leading to a reform of central bank institutions, not the other way around. In order to demonstrate this, however, our empirical study could not make a priori assumptions over who the relevant actors bargaining over monetary authority will be, nor that their relative bargaining positions will remain constant over time.

This paper consists of four sections. The first answers why the study of monetary authority is critical for democratic governance and proposes to expand the notion of monetary authority through treating it as a special mode of political authority. The second reviews the existing literature and the third conducts an empirical case study of Brazil. Contrary to conventional arguments, we argue the end of inflation has created the conditions for a centralization of monetary authority in the central bank. The fourth section mercifully concludes.

2. WHY THE INTEREST?

Why the interest in central bank autonomy for political scientists, and why is tackling this question necessary for an understanding of recent developments in Latin American neodemocracies? In this section we argue these questions should be placed on the comparative politics research agenda for two reasons: it is relevant to explain the diversity of current regional integration experiences and, in particular, because it impinges upon the prospects of democratic consolidation and the quality of democracy.

2.1. Establishment of monetary authority is critical for effective democratic governance

The current trend towards greater central banks statutory autonomy, and/or formal independence is a special chapter of State reform. This topic of political economy poses problems of democratic governance which are specific in transitional societies. It brings to the fore the complexity of the linkages between democratization, governance and State reform in neodemocracies. For this reason, we think it all important to approach central banking in terms conceptually broader than those adopted by economists. While the latter approach the topic in terms of the political economy of monetary authority, we argue that this constitutes a special mode of political authority. This shift of focus and the development of such perspective can yield insights to a specific family of Latin American neodemocracies - those who have been most exposed to protracted hyperinflationary pressures throughout their experiment in democratization, such as Argentina and Brazil.

The recent and convoluted Brazilian experience of central banking in particular provides a wealth of insights regarding the linkages between economic stabilization and democratic consolidation, for two major reasons. In the first place because throughout the 1980’s and early 1990’s Brazil represented an extreme case of the “untidy praxis” of central banking mentioned by Whitehead as typical of many countries in the region, and more broadly, of neodemocracies.¹ Secondly because of Brazil’s status, in a regional perspective, as a late-late comer in achieving a modicum of economic stability within an unstable and shifting political context of democratization.

One of the most striking developments of the 1990’s has been the general trend towards greater central banks statutory autonomy, or even formal independence, in both the dominant capitalist democracies and neodemocracies (South and East). This is taking place quite independently of the political and ideological tradition and image of the parties whose leaders were able to push in that direction, as recent events in Great Britain show. Who would expect the formal independence of the Bank of England would be one of the first initiatives of the new Labor government? On the other hand, in most countries of Latin American and in Southern Europe the quest for greater autonomy from the Federal Executive and/or from the political arena is part of a broader process of economic stabilization and restructuring taking place in the political context of democratization.

Differences between the two groups of countries should not be underestimated. ln OECD countries a great deal of political resources have been spent in debating over how much discretion should be delegated from individual central banks to a new institution responsible for monetary order at a regional level. The starting point of such a debate, however, is an established national monetary authority acknowledged not only by the associated members of the financial system, but also by the relevant populations. The question of how to obtain recognition of the specific powers central banks are expected to exercise has already been resolved within the territorial jurisdiction of each national state. As a recent brand of literature has shown, this is the long-run outcome of a number of different and convoluted institutional histories of European Central Banks (and of the Federal Reserve). Analogously, current research further indicates the recent wave of greater central bank autonomy also contains a diversity of institutional designs - far more than usually granted for by the orthodox approach to central banking.

This outcome has challenged an essential assumption of traditional approaches to Central Bank reform: that there is just one effective model of Central Bank independence to be adopted by neodemocracies.² Undeniably, though, strategic political actors and economic agents within dominant capitalist democracies have converged upon recognizing the authority of individual Central Banks within their national boundaries. However incremental the legitimization of Central Banks may have been in the past, and however diverse the political mechanisms and institutional devices by which they are accountable to other actors, critics of orthodoxy easily acknowledge Central Banks as an indispensable mode of authority. We start from a similar assumption, namely, that Central Banks are a necessary condition - and token - of an uncontested commitment to monetary order on the part of strategic sectors of the relevant societies.

The economic and political motivations for such a legitimization of monetary authority have been very diverse. One possible source may be a strong ideological commitment to price stability and/or concern with the uncontrolled passion of rulers, which in the case of England derived from the struggle of important sectors in civil Society against mercantilism. A different and more recent example is given by the German Bundesbank. Its successful performance for 40 years as an institution independent from the political arena owes much to the devastating effect hyperinflation had upon millions of Germans during the early 1920’s in addition to the political use of the Reichsbank as an instrument of the nazi regime³.

Whichever way legitimization was achieved (either through an ideological and socially embedded commitment to price stability and the rule of law, or by way of an economic and political trauma), a second condition for the exercise of monetary authority in a democratic framework is the acknowledgment of price stability as first priority public good. A number of other political requisites must be satisfied in order to reconcile the exercise of monetary authority with democratic governance. These are to a great (although variable) degree related to the premise of social legitimization discussed above.

A convenient starting point to proceed discussing the impact Central Bank autonomy may have upon democratic governance is to recall the special functions and powers attributed to central banks. A first question is why, and on what grounds, have economists properly labeled central banks, or institutions performing similar functions, monetary authorities? The answer rests with their specific functions in shaping monetary and exchange rate policies, securing financial stability, and ultimately acting as guardians of the currency. The effective performance of those tasks is closely related to the exercise of specific powers: the regulation of the financial system, monetary supervision, and maintenance of external creditworthiness. In this capacity they are rule-givers as much as rule-enforcers through their capacity to enforce penalties and generate acquiescence on the part of all members of the financial system to existing regulation.⁴

This function should be distinguished analytically, as Weber reminded us, from another more indirect mode of power also enjoyed by those economic institutions. This latter consists of providing incentives for actors in the financial markets to respond to guidelines of credit and monetary policies in a cooperative way according to a different criteria: their own variable individual calculations of what is in their best interest. This latter mode of power derives from the fact that central banks enjoy a monopoly or quasi-monopoly over credit, whereas their capacity as rule-givers stems from their expected performance and legitimate intervention in the economy as guardians of an established monetary order. We can therefore infer monetary authority is a special mode of political authority and of power, thus providing two reasons to approach Central Banks as an important chapter within the study of political economy.

Notwithstanding the key role the concept of authority has for political theory - democratic or not - central banking still remains a topic explored primarily by economists. This does not mean economists have neglected to explore political dimensions to central banking. They, however, tend to leave aside a major problem extremely relevant for political scientists which encompasses normative issues and questions of institutional engineering - particularly relevant for neodemocracies. Namely, how to reconcile the exercise of monetary authority with a modicum of democracy? That question is of no special concern for economic analysis because at best democracy has been treated and remains a residual category.

This may help explain why much theorizing about the social and political role of those institutions has focused predominantly on two dimensions of politics. Namely: the impact Central Bank autonomy has upon growth, employment, and other developmental issues; the requisite political conditions for a Central Bank to perform its expected economic functions - a dimension both technical and normative.

The conventional approach takes for granted central bank independence improves the quality of democracy, given the insulation of central bankers from “undue” particularistic pressures and from the politicization of technically complex monetary issues. If, however, we want to understand the political logic and dynamics of the process by which the exercise of monetary authority is achieved in a democratic context the mere recognition of the highly technical nature of monetary operations and the potential danger from particularistic pressures is not a satisfactory starting point. Instead, the analyst should acknowledge that a key requisite of such process is an act of political delegation of authority - from politicians to other actors - which may indeed have beneficial consequences. On condition that analysts and policy makers alike insert the issue of accountability and transparency within this act of delegation.

To the extent that, in the name of monetary order, insulation from the political arena is indispensable for protecting against self-servicing particularistic interests, the analyst is confronted with one of the most fundamental (and oldest) challenges for democratic theory: who guards the guardians⁵? And, by implication, which set of institutions are best equipped to counteract the risks of delegating the necessary autonomy to institutions regulating monetary authority⁶? This leads us to explore well-known dilemmas pertaining to democratic governance best explored by Dahl, related to the permanent tension between autonomy and control in a pluralist democracy. Two questions follow: to whom should central bankers be accountable, and what are the limits to transparency? The latter is complicated by the fact that secrecy, or at least non-immediate disclosure, is often an important tool for effective monetary management not only on the grounds of economic efficacy, but also as a means to protect against speculative attacks.

Economic liberalism takes for granted three assumptions. First, price stability is a public good. The second, is that because inflation is more detrimental to social sectors less able to protect their incomes from the “inflationary tax”, economic stabilization is instrumental to achieve a greater degree of equity in the long- run. Third, central bank independence is a principal condition for actors and the institutions in charge of price stability to act in the interests of society. ln other words, the conventional approach presumes a conception of what constitutes good government and a good society. These are of course major criteria in the legitimization of Central Banks in dominant democracies which the concept of monetary authority draws heavily upon.

2.2. Democratic governance and inflationary regimes: towards a typology

To what extent do same criteria for good government delineated above hold for neodemocracies? In our view, neodemocracies as an analytic category is too broad for our purposes. In order to explore the problems associated with the establishment of a legitimate and accountable monetary authority one must first make the distinction between two families of neodemocracies. On one hand, there are countries who experienced runaway inflation and that followed an explosive path as a consequence of both adjustments imposed by the 1982 external shock and past policy-choices⁷ They fall under the rubric of hyperinflationary experiences. In such cases, economic stabilization became a priority, inseparable from economic restructuring and, in particular, the reform of the State. This is so because hyperinflation is in effect synonymous with the end of any monetary regime. This is a term which, translated into political science parlance, is nothing less than a disruption of the existing monetary order. Hyperinflation can therefore be interpreted as a process of accelerated loss of monetary - and political - authority. It may further be interpreted as the combined outcome of a fiscal and legitimization crisis of the state affecting both itself and its regulatory capacities, and therefore, the long-term relationship it established with the economy and society.

Most Latin American countries fit this category, although one can draw additional distinctions amongst the most conspicuous hyperinflationary cases -Argentina, Bolivia and Brazil. Whereas in the first two countries hyperinflationary pressures developed into an effective hyperinflationary crisis, in Brazil such pressures were kept under permanent, yet precarious, control thanks to generalized indexation lasting until 1993. This mode of control, which as far as decision-making was concerned, verged on a praxis of systematic brinkmanship.⁸

On the other hand, there are neodemocracies for which the task of reconciling economic restructuring with democratization could be achieved in a context of comparatively manageable rate of inflation. This was due in great part to a more amenable fiscal crisis, related to a comparatively reduced exposure to the 1982 debt crisis and to an externally supportive environment created by the European Union, which provided incentives for a smooth change in the prevailing pattern of State financing. Portugal and Spain fit this category.

In what follows we shall address problems typical of the first group of countries, resorting to evidence supplied primarily by Brazil, but eventually by Argentina and Bolivia. This family of neodemocracies have distinctive features. They represent extreme cases where economic restructuring and democratization are taking place in a context in which monetary order has been and still is a goal, not a reality. We are therefore assuming that although hyperinflationary pressures have been eliminated and inflation brought dramatically to unprecedented low levels, the achievement of self-sustained economic stabilization is a long-run process dependent upon the implementation of economic and institutional reforms, and the continuance of the external bonanza.

This context imposes a number of additional conditions upon an effective State reform compatible with democratic governance, which principally concerns the reestablishment of monetary authority under the context of a broader, and conflict-ridden, process of recreating democratic political authority. Two conditions are essential to this process. The first is the recognition of self-sustained economic stability as a public good, with the expectation this goal is inseparable from greater equity. The second condition is an agreed process of institutional innovations geared to rendering monetary authority accountable to society without becoming prey of the political arena.

These conditions may be referred to as a quest for embedded authority, typical of a situation in which the State ‘s capacity for regulation and supervision will be dependent inter alia upon a changing pattern of financing and composition of its expenditures. This implies, of course, a different state-society relationship, the implications of which have been dealt elsewhere.

We are therefore positing the question of Central Bank autonomy⁹ should be considered in a yet broader perspective than the one proposed by critics of the conventional approach. We can begin with Laurence Whitehead’s critique to a presumed convergence toward a single model of central banking in neodemocracies; an assumption resting at the heart of neoliberal prescriptions. A distinction must be drawn, however, between two separate issues. The first is the observed and indispensable trend toward greater central bank autonomy from the contradictory pressures of competitive politics and its use as an instrument of party politics. This does not imply, however, an uncritical acceptance of the single-model version that bearers of the conventional approach tend to prescribe to neodemocracies.¹⁰

In countries like Brazil and Argentina a second issue to be explored, concerns both the pace and the guidelines for this mode of State reform to be effective. The establishment of a new monetary order poses specific problems to democratic governance - such as the legitimization of monetary authority and the institutional innovations needed to make it more accountable to politicians and society. In our view monetary authority has yet to be fully established in these neodemocracies. Furthermore, other dilemmas should be considered if democratic governance is to be counted as a legitimate competing goal (alongside with other ones). Some implications follow from this proposal, which we will briefly review.

First, the political dynamic of economic stabilization concentrates power, at least at its initial stage. A principal feature of neodemocracies consists in re-concentrating power within the institution responsible for monetary authority - a process which implies multiple acts of delegation. These acts of delegation range from: politicians to experts and eventually to whole bureaucracies, from politicians situated in the Executive and/or in Congress, from politicians holding Executive positions at the State level in a federation, and from the Judiciary as the authorized interpreter of the legal order. The other side of this multifaceted game is, of course, the multiplication of veto players¹¹.

Second, the technical complexities of monetary management regarding its financial reorganization and regulation is further enhanced by another factor. When hyperinflation sets in, together with its inseparable companion, the fiscal crisis of the state, a tidy praxis of central banking requires a modicum of coordination between monetary and fiscal policies. This condition can be satisfied in many ways, one of which is to provide the President of the Central Bank veto power over (expansionist) fiscal policies - thus preventing the risks derived from indirect challenges to the established monetary guidelines, and that is to say, its own authority.

Exploring the links between fiscal and monetary policy for stabilization in a hyperinflationary context is out of the scope of this paper. We, however, wish to make a point pertaining to this relationship. A major constraint facing the delegation of power to monetary authority is the usually low level of understanding of its technical issues systemic policy consequences. Stated in a more positive light, technical expertise among politicians within a democratic framework is a precondition for establishing a new monetary and regulatory financial system.¹² This is an ideal long-run condition, which is far from being satisfied even in dominant democracies yet still more difficult to achieve in countries where uneven development - social, economic and political - is the rule. The problem, however, is more critical in neodemocracies under hyperinflationary stress. The question for us is: which modes of political and institutional intermediation are required in order to bridge the gap between technical knowledge and expertise concentrated in the hands of a given economic team on one hand and on the other the actual level of understanding of politicians with a mandate, and whose fiat is indispensable?

Incentives to bridge the gap more or less quickly may come from various sources. We posit that in cases in which institutions are not flexible and effective enough to secure the legitimacy derived from the representative system, the incentive to bridge that gap and to help check hyperinflationary disruption may come directly from a definite change in the preferences of the population. Such was the case in Brazil and most Latin American countries. This is why beyond all differences related to diversity of the relevant party systems, and in particular beyond the technical differences in their stabilization programs, Bolivia (1986), Argentina (1991-2), and the latest comer Brazil (1993-4) show a political denominator in common.

In all three cases, the building up of an agreed economic strategy was dependent upon three conditions. First, a shift in the ordinal preferences of society toward making stabilization a first order priority goal. This process may be translated as the emergence of a new criteria of legitimacy owing much to the role of hyperinflation in the wake of a number of failed stabilization programs within new democracies. Under such circumstances there is a shift in what should be expected from democracy. Above all, there is a gradual de-linking between immediate economic welfare and treating democracy as a goal and value on in its own. The second condition is related to the quality of the leadership. An executive must have the political capacity to propose a technically adequate stabilization program in addition to create a political coalition necessary to respond to a newly emerging anti-inflationary coalition¹³. From this follows a third condition, still related to the quality of the leadership. In order to bridge the gap between technicians and politicians within the framework of a disrupted monetary order one must delegate a high degree of autonomy to the economic team, and therefore to the monetary authority, while at the same time secure the support of Congress, and eventually of local executives. This was a condition satisfied in the case of Brazil under Cardoso, Menem in Argentina, and in particular, of Bolivia under Victor Paz Estensoro.¹⁴ In the case of Brazil, we can speak of a paradox by which, in the absence of an institutionalized fully autonomous power of the monetary authority, the Central Bank was granted the right to proceed as if it were “independent” in the wake of the Mexican crisis of 1994-5. The de-politicization of the Central Bank was dependent upon a political decision - an indicator of the precarious statutory autonomy of the Brazilian monetary authority.

Finally, decentralization poses as crucial problem for countries (like Brazil and Argentina) who face the need to establish credible monetary authority while improving the conditions for democratic governance. Economic and political decentralization is often claimed essential for viable democratic governance through its beneficial effects upon accountability, transparency and public sector efficiency. The literature to date on decentralization, however, has not explicitly incorporated the question of monetary authority within its analysis.

3. ALLEGATIONS IN THE LITERATURE

The prevailing wisdom concerning monetary authority is basically normative. It stipulates the independence of central banks from the political arena as a precondition for achieving durable price stability. In order to uphold the economy’s long-term interest in price stability, monetary authority should be delegated to a central bank independent from the political arena and in particular from short term electoral interests.¹⁵ The well-known study by Alesina -which finds a positive correlation between independent central banks and price stability¹⁶ - is interpreted as evidence of a causation. Economic growth and price stability therefore rests on ‘getting the institutions right’. While not dismissing the potentially large role political institutions have on policy outcomes, the Brazilian case qualifies the generalizability of arguments focusing on ‘getting the institutions right’. This is because, as we demonstrate, conditions for creating an increasingly autonomous central bank have only recently been generated because of price stability, not the other way round.

There are positive lines of research which attempt to explain the conditions under which politicians delegate monetary authority to an independent Central Bank from a respective international and national level of analysis. Sylvia Maxfield argues politicians delegate monetary authority to an independent Central Bank as a mechanism to signal international creditors their commitment to a policy of price stability in order to attract investment when international credit is tight.¹⁷ International credibility is achieved through a policy of “tying one’s hands” - by delegating monetary authority to central bankers. However, there still exists a considerable amount of “slack” at the domestic level which Maxfield’s approach cannot account for as her own country case study of Brazil suggests.

Brazil is a case which warrants the analyst to develop domestic level hypothesis¹⁸ Dispersal of monetary authority - either dependent upon the federal executive,¹⁹ or effectively divided between levels of government - has been the rule. It remained so throughout sustained periods of tight international credit and economic stagnation such as the 1980s and 1990s.

Domestic approaches to Central Bank independence have primarily been developed in the OECD context and can be divided in two categories: institutional and social preference explanations²⁰. It is sufficient here to recall that institutional explanations generally presume delegating authority to a Central Bank is costly²¹. This would be easier to achieve in countries where corporatist structures and practices prevail. Social preference explanations, treat institutions as intervening variables and focus on the preferences of dominant groups who influence economic policymaking. Central Bank autonomy is therefore most likely in countries where the financial sector is relatively stronger.²²

The argument we develop for the Brazilian case study adopts a domestic level of analysis which contains elements of both a social preference and institutional explanation. Our explanation, however, differentiates from the above orientations in a few significant ways. On the institutional side we agree that policy change in politics with a large number of veto gates is difficult²³ but the Brazilian case indicates how the relative bargaining position of actors occupying those veto gates can change over time. A strictly institutional orientation would not have the tools necessary to examine when such change is possible and would be unable to explain the recent centralization of monetary authority in the Brazilian Central Bank.

In order to account for the dynamic element inherent in the politics of monetary authority, the analyst must turn to the relevant actors involved. Unlike the social-preference explanations previously cited, however, we do not assume the relevant actors impinging upon monetary policy will be the same across countries, nor, that their relative bargaining positions will remain constant. In Brazil the relevant actors disputing control over monetary authority include the federal executive, state governments, and the private financial sector. After the 1988 Constitution was promulgated the Senate and the Judiciary also are part of the bargaining game and must be included as additional veto players. The achievement of greater monetary autonomy requires, therefore, multiple balancing acts on the part of the Executive. Other countries may have very different actors bargaining over it. Furthermore, in the case of Brazil the stalemate over monetary authority and its eventual independence is beginning to dissolve because the end of inflation reduced the relative bargaining strength of the actors opposed to a centralization of monetary authority in the Central Bank.

The literature on decentralization has neglected the distinctive features of monetary management, the conditions for its effectiveness, and the special dilemmas it poses for democratic control. Furthermore, it fails to make a satisfactory analytical distinction between the political logic of decentralization and that of federalism in a democratic framework. In Brazil (and in a lesser degree in Argentina) states and municipalities played an active role in the passage from high inflation to hyperinflation within the context of democratization - a process inseparable from their claims for greater autonomy.

In what follows we are principally concerned with examining monetary authority from the standpoint of the relationship between the Brazilian Central Bank and state banks. This focus is justified for three reasons. Democratic governance is first predicated upon coherent control by the incumbent government over the proper State apparatus.²⁴ Second, the creation of quasi-money by state banks has an obvious direct impact on the expansion of the monetary base. Finally, some of the private financial institutions which incurred in dubious practices and in systematic non-compliance with existing rules as the Banco Econômico in Bahia or Bamerindus in Paraná, enjoyed a quasi-monopoly in the local financial system. This explains their informal status as regional banks and the enormous bargaining power vis-à-vis the Central Bank²⁵.

An approach which does not make a priori assumptions over who will be the relevant actors, nor what their relative bargaining strengths will be over time can further shed light on accounting for the wide empirical variation in central banking models adopted.²⁶ The question of autonomy from who “leads thus to a preliminary task: identifying the shifts in social interests underlying the new trend towards Central Bank independence. This is vital if criteria such as transparency and accountability are to be included as necessary constraints to the exercise of “effective monetary authority”.

4. BRAZIl’S QUEST FOR EMBEDDED AUTHORITY

4.1. The political game prior to the Real

Brazil has achieved a dramatic reversal of its hyperinflationary pressures in 1994-5 under the aegis of the Plano Real initiated by then Minister of Finance Fernando Henrique Cardoso. In a regional perspective Brazil can be considered a late-late comer regarding price stabilization because of its numerous and unsuccessful heterodox shocks in 1986, 1987, 1989, and 1990. By 1992 the government resorted to orthodox monetary and fiscal short-run policies with only partial success.

As in other parent countries, the recent Brazilian economic stabilization brought about a dramatic change in perspectives. Price stability came to be reckoned as a first order public good, reflecting a clear change in social preferences and criteria for political legitimization - something unprecedented in contemporary Latin American history. The experience of hyper-inflation and/or hyperinflationary crisis instigated a lowering of expectations concerning economic welfare and democracy. All over the region such changes inflicted a dramatic blow to economic populism. It became clear the populist policies and style of problem-solving would not yield electoral dividends as it happened in the past. This change in the political and economic climate holds true for economic structures, political regimes and systems of representations across countries as diverse as Peru, Bolivia, Argentina, Mexico and Brazil.

Why has Brazil been one of the slowest reformers? Although answering this question thoroughly is out of the scope of this paper, one must highlight its importance for two reasons. First, economic stabilization is a process still under way. As a goal it is far from consolidated, for its continued success depends upon fiscal, social security, and administrative constitutional reforms that require Congressional approval. Second, some of the political constraints to managing hyperinflation are still at work and impinge heavily upon the difficulty to establish an embedded monetary and fiscal authority as a specific mode of political authority.

In what follows we will be concerned with two kinds of constraints. The first is long-run and refers to how monetary authority was handled under a national and federal perspective prior to the democratization of the 1980s. This discussion will provide the background necessary to address the main question of this section: how has exercise of monetary authority changed subsequent to economic stabilization?

In regional perspective, one of the distinctive traditions of Brazilian policymaking, in addition to an explicit vocation toward continuous growth at all costs, has been a difficulty to establish a single monetary authority. From this standpoint Brazil is also a late-late comer. The establishment of a single monetary authority was politically viable only, under the authoritarian regime. Moreover, except for a brief interlude of 1964-7, the Central Bank since its creation in 1964 has experienced very little autonomy.

One could tel1 the history of the eight failed attempts at stabilization throughout the populist democracy (1945-64) in terms of the poiitical constraint posed by the protracted dispute between two rival centers of monetary authority - each subjected to different political and economic pressures. Politicization of monetary management in that period therefore prevailed under the guise of the contradictory policies made possible by the existence of a dual monetary authority; Banco do Brasil and Sumoc (Superintendência da Moeda e de Crédito).

Subsequently, the convoluted history of Brazilian Central Banking since 1964 provides a wealth of insights for political analysts concerned with explaining the contrast between economic programs and ideas on one hand and the effective results of economic policymaking on the other. For, although the intentions and the rules laid down by the architects of Central Bank reform in 1964 were clearly stated in terms establishing an autonomous Central Bank from the political arena, such a proposal was easily defeated three years later under the subsequent military government committed to expansionist policies.²⁷ It was only in 1986 and, subsequently in 1988 under the first civilian government, that two reforms were implemented to establish basic mechanisms for effective monetary management and a minimum degree of transparency in national accounting.²⁸

The history of monetary authority in post-war Brazil has therefore less to do with the nature of the political regime than with a commitment to economic performance cast in terms of accelerated growth at all costs²⁹. International liquidity, of course, played an important role from 1974-8, as Sylvia Maxfield argues. The decision taken by General Geisel in 1974, however, to prolong the period of accelerated growth and deepen import substitution despite the first oil shock (which required adjustments) was a domestic political option³⁰.

Brazil has become one of the slowest reformers for a second reason directly related to the kind of dual monetary authority, which emerged in tandem with democratization. This dual monetary authority is intimately linked to the federal question and decentralization, for the governors acted as centrifugal forces who pressured for fiscal decentralization and used their state banks to emit money, thus fiercely challenging the constitutional authority of the Central Bank. An important explanation for the ability of governors to use their state banks in such a manner, and thus undermine all stabilization plans undertaken by the new civilian regime (1986, 1987, 1989, 1990 and 1991), relates to the sequence by which democracy was restored. The contrast between the Brazilian and Spanish strategies may help clarify this point.

By the time Brazilian civilian rule was re-established in 1985-6, the strategy of political devolution adopted by the military had been under way for almost 10 years. For our purposes it is sufficient to stress two major aspects. First, devolution at the state level occurred prior to that at federal level. Open gubernatorial elections were re-introduced first in 1982, with the opposition party (PMDB) electing a significant number of governors committed to democratization in the most economically and politically powerful states (São Paulo, Minas Gerais, Rio de Janeiro e Rio Grande do Sul). Second, the economic strategy adopted by the military government of Ernesto Geisel (1974-8) ensued a process of accelerated economic decentralization at an early stage of political liberalization. Geisel’s radically changed the previous pattern of industrialization which focused on the already industrialized south-east.

This sequence is critical to understand the manner in which the political drive for autonomy at the state level translated into a centrifugal pull. Economic, and eventually democratic devolution at state level prior to liberalization at the federal level³¹ helps explain the difficulties later faced by the federal Executive to exercise a coordinated authority and implement an agreed strategy and support amongst governors. A political economy of federalism should take into account such a political and economic sequencing.³²

An explanation for the difficulty of any subsequent civilian government to build a governing coalition, should take into account this kind of political constraint. For this sequence set the stage for a second type of political constraint to economic policymaking and stabilization - the new Constitution promulgated in 1988. The new Constitution transferred a fixed percentage of federal taxes to states and municipalities without a concurrent transfer of spending responsibilities and represents the culmination to process of fiscal decentralization beginning in 1974. Such a constitutionally determined fiscal allocation imposes an important constraint upon economic management and the crafting of social policies.

The ability of local banks to create quasi-money through their administration of state government fiscal deficits - either through carrying state bonds or providing direct loans - has had two important consequences. First, state banks became a favorite mechanism of state governors to evaded meeting elementary requirements of transparency and accountability established by monetary and fiscal authorities at the federal level. Second, this implied a deliberate challenge to the constitutionally established monopoly over money creation at the federal level.

In other words, the disruption of the monetary order in democratic Brazil since the early 1980 ‘s was closely related to the operation of centrifugal inter-governmental forces which effectively created rival centers of monetary power between levels of government. This was part and parcel of the dramatic loss of both monetary and fiscal authority, which is inseparable from hyperinflation - a process leading to the absence of any monetary and fiscal regime. In effect, Brazil ‘s source of monetary disorder can be characterized as a monetary tragedy of the commons. While state governments had the individual interest to use their state banks as effective creators of money in order to finance their budget deficits, one could argue the collective good of price stability was jeopardized - which is a federal, not state responsibility.

That is why when examining the more extreme case of Brazil it is possible to speak of a monetary and fiscal rebellion as a process inseparable from hyperinflation.³³ The complex bargaining act required for self-sustained economic stabilization in such a neodemocracy places additional strains on political leadership, for it constitutes a permanent trade-off between two goals. On one hand the need to check and to bring under permanent control centrifugal forces which deny central monetary authority - a need which must be first turned into a political goal by the rulers. On the other hand, the Executive is faced with the need to build an agreed economic and political! strategy to reshape the political order consonant with federalist principles.

The study of decentralization, at least for the Brazilian case, must incorporate the question of a coherent monetary authority as an essential requisite for the viability of a new federalist pact - in addition to reconciling decentralization with the objectives of democratic governance: accountability, transparency, predictability in the rule of law, and public sector efficiency.

4.2. The political game after the Real

This section argues monetary authority has been increasingly centralized in the Central Bank because the federal executive has increased its bargaining power over governors since the implementation of the 1994 Real stabilization plan.

The federal Executive has been able to centralize monetary authority in the central bank for both political and economic reasons. On the political end both the interests and bargaining power of the federal Executive have changed. Not only does the current administration have an unprecedented interest to sustain the present stabilization plan, but the Executive has gained leverage over the legislature, and therefore governors, due to a wedded presidential, legislative, and state-wide elections in 1994. On the economic end the federal government increased its bargaining power over governors because the end of inflation substantially weakened state government finances, thus making governors dependent upon a federal bailout. Fernando Henrique Cardoso took advantage of both of these political and economic factors to condition a federal bailout upon a centralization of monetary authority in the central bank.³⁴

The federal Executive is usually the branch of government held electorally accountable for maintaining macroeconomic stability, therefore any federal Executive would have an interest to maintain a stabilization program. The government of Fernando Henrique Cardoso, however, differs from previous administrations by the extent of its interest in upholding a stabilization program. Fernando Henrique launched and won his bid for the presidency almost entirely based upon his role as crafting the Real stabilization program as Minister of Finance under the outgoing Itamar Franco administration (1993-4). Whereas previous presidents could craft a stabilization program after being elected, have it failed, and still have time for a second “go”, Fernando Henrique does not have that option because he won the presidency on a specific stabilization program - the Real. His political career, and chances for reelection in 1998, are dependent upon the continued success of the Real.³⁵

In addition to an unparalleled interest in sustaining the current stabilization program, the Executive branch gained added leverage over the legislature and governors through the wedded presidential, legislative, and state elections of November, 1994.³⁶ For the first time in Brazil’s recent democracy, legislative and gubernatorial candidates had the potential of riding presidential coattails. The successful presidential bid of Fernando Henrique (PSDB) indeed appears to have influenced both the legislative and gubernatorial elections.³⁷

The unparalleled interest in sustaining a stabilization plan combined with a wedded election in 1994, however, do not constitute sufficient causes for a successful centralization of monetary authority. Essential to this process has been the economic impact the end of inflation had upon state government finances. State governments have been running fiscal deficits for quite some time before 1994. The end of inflation, however, made those deficits unsustainable. The Real stabilization plan squeezed State finances through both a monetary and fiscal mechanism. On the monetary end state governments could no longer use the large “floating” revenue of state banks, and on the fiscal end states could no longer use inflation to corrode real spending on items like public wages and were obligated to disburse significantly higher debt payments because of the higher interest rates.

High inflation generates winners and losers. A clear winner under high inflation in Brazil was the financial sector. Each year between 1990 and 1993, the banking sector (both public and private) generated inflationary revenue to the sum of four percent of the country’s GDP - with public banks appropriating roughly two thirds of the total³⁸. In fact, floating revenue became a principal reason for preventing many state banks from entering complete bankruptcy during much of the 1980s and early 1990s. While most of their credit operations (assets) were directed to the public sector in the form of long-term financing, their liabilities derived primarily from short term deposits, or bank bonds (CDBs etc). Without the floating revenue, state banks had increasing difficulties to meet their cash requirements. When a bank does not balance its account by the end of the business day, it has two options: either borrow the money over the “inter-bank market” (mercado inter-bancário) at a heafty interest rate, or resort to the Central Bank’s discount line (linha de redesconto). By the end of 1994 the private market was no longer accepting state bank CDBs or RDBs (bank bonds), and thus prominent State banks like Banespa and Banerj (who make up 60% of the financial sector) repeatedly turned to the Central Bank’s-providing the impetus for their subsequent intervention. In sum, without inflation revenue many state banks became insolvent, thus inducing a Central Bank intervention. This placed a pinch upon State finances because state banks were essentially the managers of State government debt, and through that role subsidized State finances.

The end of inflation further deteriorated state government finances through two fiscal mechanisms. First, states could no longer keep budgetary costs down through using inflation as a means to corrode away real spending on items like public payroll.³⁹ As a result, public payroll consumed an increasing percentage share of state budgets.

Second, the end of inflation had the effect of dramatically increasing debt obligations. Much like the stabilization program in Argentina, the Real plan was based upon a stabilization of the currency through the exchange rate. Such a stabilization plan forces the government kept a high interest rate in order to attract foreign capital, which consequently had the effect of higher debt payments on behalf of state and municipal governments. The chart below demonstrates how state level debt mushroomed at an astonishing rate after the stabilization program was implemented. Note, however, the dramatic increase derives from the two categories of debt whose interest rates are not fixed; bonds and debt owed to state banks.

TRENDS IN STOCK OF STATE DEBT BY CATEGORY (R$ BILLION)

Before proceeding with the argument, however, two qualifications are in order. First, we do not argue the end of inflation caused the fiscal crisis many state governments are currently experiencing. The first major state borrowing boom began in the 1970s under the military government’s developmental project, and during the 1980s state governments were able to roll over that debt, and contract new debt through the use of their state banks. Rather than create the fiscal crisis, the end of inflation eliminated the mechanisms which sustained chronic state government fiscal imbalances.

Second, it is important to note the weakening bargaining position of state governments began prior to the imposition of the Real. State and municipal governments began a process of debt rescheduling during the early 1990s, and both the Senate and the Central Bank imposed more stringent limits on the ability of subnational governments to contract new debt. With each new round of state debt crisis during the 1980s and 1990s the federal government, and Central Bank, gained incremental leverage over State government finances. Following the state bank crisis of 1987, for example, the Central Bank, through Decree Law 2321, gained the ability to assume temporary control over insolvent state banks. The same decree law further established more stringent rules regarding the judicial accountability of state administrators for improper banking practices while the state banks placed under federal intervention in 1987 were eventually returned to their respective state governments with no significant judicial action taken against its bank managers, the central bank gained a new tool to discipline state banks⁴⁰.

The next round of state government financial crisis during the early 1990s further placed new limits on state banks and state government finances. In 1990 the National Monetary Council imposed more stringent limits on the ability of state banks to loan to the public sector,⁴¹ and in 1992 the judicial accountability public administrators for their actions in state enterprises was further tightened⁴² In addition to imposing more restrictions on the state banks, the Senate, as the branch of government responsible for setting legal debt limits for states and municipalities, further limited the ability of subnational governments to contract new debt⁴³.

Thus, the end of inflation in 1994 did not initiate the process through which the federal government is exerting greater control over State finances - either through Central Bank discipline over state banks or through legislative restrictions over the permissible level of debt states are able to contract. The ability of the central bank to exert discipline over state banks must be viewed as an incremental process beginning in the early 1980s which were imposed with each successive state banking crisis. These periodic crises were induced both by the political use of such banks at the time of elections and by the various, and inevitably failed, stabilization programs which temporarily ended inflation. Each state banking crisis has coincided the electoral calendar (1982, 1986, and 1990) and stabilization plans (1986, 1989-90).

We argue, however, the recent crisis of state banks induced by the end of inflation in 1994 differs from the previous state banking crisis on two and potentially three fronts. First, as we demonstrated, the Executive branch has an unparalleled interest in disciplining state banks in order to sustain the stabilization program, and second the Real plan has so far proven a much more sustainable stabilization plan than its predecessors, thus inducing more drastic pinch on state bank, and state government finances. Third, the central bank had more tools at its disposal to discipline state banks in 1994 in part because of the periodic state banking crisis describe above. The federal government therefore took advantage of both political and economic factors which weakened the bargaining position of state governors in order to, in addition to other. measures, condition a federal bailout on a centralization of monetary authority in the Central Bank.⁴⁴

The evaluation by the federal government that disciplining state-level financial institutions was a necessary component to making the Real Plan work was made evident by declarations of members of Fernando Henrique’s economic team soon after his electoral victory in November of 1994.⁴⁵ The government didn’t lose any time. On December 31, 1994, the last day of the presiding state government administrations, the Central Bank intervened in the two largest state banks belonging respectively to the states of São Paulo and Rio de Janeiro: Banespa and Banerj. In addition to infusing “national” elements to a predominant local and regional congressional election, Fernando Henrique Cardoso further took advantage of a concurrent turnover of state and federal governments⁴⁶. By intervening on the exact date of turnover the outgoing governor doesn’t have a chance to mobilize his political allies against the intervention, and the incoming governor doesn’t have as large of an incentive. One the one hand the incoming governor isn’t directly accountable for the intervention, and more importantly, the governor may have a financially healthy bank returned to the state halfway through the term.⁴⁷. While legislators from São Paulo and Rio fought against the stated aims of the Central Bank to eventually privatize both banks, initial opposition to the intervention might have been greater if it had taken place during the middle of a gubernatorial term.

Once the Central Bank had conducted the intervention of Banespa and Banerj the government used whatever tools at its disposal to increase the effective threat of privatization - and this was accomplished through slowly increasing the financial market to foreign competition⁴⁸. With new foreign competitors looking to establish themselves in the Brazilian market, the privatization threat becomes much more credible due to potential buyers.

In sum, the political game between the executive, legislators, and governors has substantially altered during the 1990s.

4.3 Evidence to increased Central Bank discipline over the monetary system

The federal government has been able to condition a bailout of state finances upon a centralization of monetary authority in the central bank through two mechanisms: the rescheduling of state debt, and Central Bank bailouts of state banks. While the former is also a prerogative of the Senate, the latter is exclusively of the Central Bank.⁴⁹ Unlike previous federal bailout of state finances, the recent round of Central Bank interventions and State debt rescheduling have demanded greater countermeasures on behalf of state governments. In monetary authority states have been induced to strictly adhere to existing central bank resolutions, and more importantly, to privatize or transform their state banks into development agencies. Central Bank discipline over the financial sector, however, has not been limited to state governments - the recent rescue of private banks has been conditioned on similar measures.

As the previous sections demonstrated, Central Bank intervention of state banks is not a novelty of the 1990s. State bank crisis, and subsequent central bank rescue, have coincided with Brazil’s electoral calendar and its stabilization programs⁵⁰. The Central Bank attempted to impose restrictions on the ability of such a crisis from reoccurring, but such measures either lacked in scope or were simply ignored. At the extreme, state non-compliance of Central Bank resolutions took the form of a blatant public defiance. In 1993, for example, a state bank from Brazil’s northeast region opened six new agencies without the requisite prior approval of the Central Bank. The state bank’s president was called to Brasília, and verbally sanctioned for his actions. Following the meeting the bank opened another six agencies, upon which the governor stated to a Central Bank director, “The bank is mine, and I do what I want”⁵¹

As the example given above indicates, much of the history of Central Bank-state bank relations has been one of successive non-compliance of central bank regulations. The state banking sector experienced periodic crisis coinciding with Brazil’s electoral calendar and economic stabilization plans. In each instance the central bank provided financial relief to state banks in return for administrative and banking practice reforms which were subsequently not met. During the early 1980’s the central bank provided conditioned financial relief to state banks through two programs: the PAC (Programa de Apoio Creditício) in 1983 and PROREF (Programa de Recuperação Econômica e Financeira), in 1984.⁵² Both programs conditioned financial assistance upon a series of measures to reduce operating expenses, close deficit agencies, and oblige states to recapitalize their banks. Most conditions went unheeded not only then but in subsequent attempts to establish monetary discipline in response to the banking crisis of 1986-7, and early 1990’s.⁵³

One must be careful, however, to ascribe the periodic state banking crisis as a mere non-compliance of sound banking regulations. Journalists and financial analysts alike often cite the high allocation of state banking credit to state governments as a direct challenge to Brazil ‘s 1964 banking legislation which limited credit operations of all banking institutions with their respective controllers to 10%⁵⁴. Despite the fact such limits were increased to 30% in the 1980’s state banks often would direct up to 60-80% of their credit operations to their respective state governments during the late 1980’s. State banks, however, were able to surpass such limits given each additional credit operation was approved by the central bank and National Monetary Council under the rubric of “extra-limit debt”(a legally permissible category)⁵⁵.

The current interventions of state banks differ from those of the 1980s and early 1990s on two important counts. The first pertains to the adherence of Central Bank regulations. While the central bank began to tighten the ability of state banks to loan to their controllers during the early 1990s, the Plano Real further tightened the controls over state banking practices which were adhered to. The National Monetary Council reestablished limits on the concessions of new loans toward the public sector, including the carrying of state bonds (Resolution 1990 of 6/30/94), prohibited the rescue by the Central Bank or treasury in order to solely recapitalize state banks (Resolution 1995 of 6/30/94) and further increased the judicial accountability of state bank administrators.

Second, and probably more importantly, full federal rescue has been conditioned on either an eventual privatization or transformation of the bank into a development agency. According to Provisional Measure 1514 edited by the executive, states are provided two options over federal rescue of their state banks. ln order for the federal government to refinance the entire portion of state debt owed to the respective state bank, the state government must agree to either privatize, liquidate, or transform the institution into a development agency. State governments, however, can retain control over their banks if they are willing to accept a federal restructuring of only half their debt to the state banks. Given the acute stage of the crisis, however, the later alternative often isn’t an option.⁵⁶

In sum, the bargaining game between the federal executive, governors and the legislature has shifted. While the federal government began the slow process of controlling state finances and state banks during the early 1990s, the Plano Real added great impetus.

5. CONCLUSION

An important dimension of State reform and of economic restructuring in neodemocracies and western dominant democracies pertains to the function and changing jurisdiction of central banks under the context of globalization and of regional integration. International factors such as the liquidity crisis of the 1980’s and the acute need to attract investments in peripheral economies in Latin American, Africa, East Europe and more recently South Asia, have provided important incentives to improve international creditworthiness. This indeed is the international background against which the world-wide trend towards greater central banks statutory autonomy observed in the l 990’s becomes intelligible.

The problems we addressed in this paper, however, led us to propose an analytical shift of focus, which we reckon is also policy relevant. We started by exploring three related questions. The first and more general is why, despite the critical impact the exercise of monetary authority has upon democratic governance, has this topic been nearly absent from the relevant political literature? A second set of questions had to do with an uncritical acceptance over how monetary authority is currently used by economists - for whom democracy is a residual category (at best). If, as we believe, the exercise of monetary authority is relevant for political scientist because it impinges directly on the question of democratic governance, the distinctive attributes of this specific mode of authority should be incorporated within our analytical framework. Finally, a third question was raised in connection with the comparatively “untidy praxis” of central banking in most Latin American countries notwithstanding the legal changes geared at granting them ever greater autonomy in the 1990s. We follow the criticisms to the conventional approach raised by Laurence Whitehead, who challenged the notion of a single effective central banking model in which to measure countries by. At the same time we tried to advance a step further, by introducing another question: to what extent categories such as peripheral economic systems and neodemocracies are too broad to explore the dilemmas posed by the need to reconcile the tasks (ideally) performed by central banks as rule-givers on one hand and democratic governance on the other?

A first conclusion is that monetary authority is a mode of political authority which like the judiciary poses an important question for democratic framework: who guards the guardians of both the currency and law? The process of legitimization of such authority depends in great measure to the value attached to economic stability as a first order public good.

With support from empirical evidence supplied by a sub-family of Latin American countries like Argentina and Brazil, we conclude the process of legitimization by which economic stability is reckoned a first order public good is specially convoluted in Latin American neodemocracies. In order to demonstrate this we thought it crucial to draw an important distinction between two kinds of neodemocracies. Those where monetary disorder reached the stage of hyperinflation (accompanied or not by hyperinflationary crisis), from those where political stresses inseparable from economic restructuring were less threatening to the maintenance of a modicum of monetary order. Argentina and Brazil belong to the former. Thus, we chose examples drawn from such extreme cases in which, together with Bolivia in 1986, the issue at stake was the construction of a new monetary order. Hyperinflation is inseparable from the absence of any monetary and fiscal regime - the State is powerless to impose the rule of law (if there is one). Brazil finds itself in this predicament for reasons linked to the federal structure, and in particular to the influence governors have exerted in the Congress due to Brazil’s logic of democratization. Brazilian governors illustrate how challenges to fiscal and monetary authority tend to take the form of fiscal and monetary rebellion, in the name of democracy. We tried to show the mode of economic and political federalism adopted during political liberalization and subsequent democratization impinged directly upon monetary order because it created a multiplication of rival centers of power prepared to create money - which in principle is an exclusive prerogative of the centralized monetary authority.

We conclude for the case of Brazil that at least one of the conditions for legitimization of Central Bank authority has been achieved to the extent that economic stabilization became a first order public good for the populace at large and the existing government. The plan’s redistributive impact in favor of lower economic segments, however, should not be dismissed as a factor to the endorsement and popularity enjoyed by the Plano Real, and subsequently its major architect, Fernando Henrique Cardoso.

The recent centralization of monetary authority in the Central Bank yields a final and valuable analytical lesson. The Central Bank has in Brazil has been able to gain increasing discipline over the monetary system in part because of the economic stabilization plan - not the other way around, as predicted by conventional wisdom. Economic stabilization has less to do with “getting the institutions right” and is more a consequence of a dynamic bargaining game between the federal executive, legislators, and subnational governments. When studying monetary authority and central bank institutions, the analyst should identify the relevant actors, their interests, and how economic and political conjunctures can shift the relevant bargaining position of those very actors. Only then can the analyst study monetary authority as political authority.

REFERÊNCIAS BIBLIOGRÁFICAS

• ALESINA, Alberto (1988). “Macroeconomics and politics.” In FISCHER, Stanley (ed.), NBER Macroeconomics Annual. Cambridge, Cambridge University Press.
• BANAIAN, K., LANEY, L. & WILLETT, T. (1986). “Central bank independence: an international comparison.” In TOMA, E. F. & TOMA, M. (eds.). Central bankers, bureaucratic incentives, and monetary policy. Boston, Academic Publishers, 199-219.
• CAMPOS, Roberto (1995). Lanterna na Popa. Rio de Janeiro.
• FANELLI, J.M., FRENKEL, R. & ROZENWURCEL, G. (1994). “Growth and structural reform in Latin America. Where do we stand.” In SMITH, W., ACUNÃ, C. & GAMARRA, E., Latin American political economy in the age of neo-liberal reform, New Brunswick, Transaction Books.
• GOODMAN, John B. (1992). Monetary sovereignty: the politics of central banking in western Europe. Ithaca: Cornell University Press.
• GRILLI, V, MASCIANDARO, D. & TABELLINI, G. “Political monetary institutions and public financial policies in the industrial countries.” Economic Policy: A European Economic Forum, 13, October, 342-92.
• HALL, Peter A. (1994). “Central Bank independence and coordinated wage bargaining: their interaction in Germany and Europe.” German Politics and Society, 31, Spring, 1-23.
• LINZ, J. & STEPAN, A. (1996). Problems of transition and democratic consolidation. Princeton: Princeton University Press.
• LOHMANN, Susanne (1994). “Federalism and Central Bank autonomy: the politics of German monetary policy, 1957-92.” Department of Political Science, UCLA, October.
• MALLOY, James M. & CONAGHAN, Catherine. (1996). Unsettling statecraft, democracy and liberalism in the central Andes. Pittsburgh: Pittsburgh University Press.
• MAXFIELD, Sylvia (1997). Gatekeepers of growth: the international political economy of central banking in developing countries. Princeton: Princeton University Press.
• SOLA, Lourdes & KUGELMAS, Eduardo. (1997). “Statecraft, economic liberalization and democracy. The case of Brazil in comparative perspective.” Paper presented at the Congress of the Latin American Studies Association, Guadalajara, April 17.
• SOLA, Lourdes (1994). “State, structural reform and democratization.” In SMITH, W., ACUNA, C. & GAMARRA, E. Democracy, markets and structural reforms. New Brunswick: Transaction Books.
• SOLA, Lourdes. (1996). “Reforma do Estado, ordem legal e democratização.” In Projeto de Reforma do Estado e Políticas Públicas, nº 38, Comissão Econômica para a América Latina, Santiago do Chile.
• SOLA, Lourdes (1998). Ideias econômicas, decisões políticas. Desenvolvimento, estabilidade Populismo. Editora da Universidade de São Paulo (EDUSP), forthcoming.
• WHITEHEAD, Laurence (1997). “Models of central banking: how much convergence in neodemocracies?” Nuffield College, draft.
• WOOLLEY, John T. (1985). Monetary politics. Cambridge, Cambridge University Press.

• 1
  Laurence Whitehead, “Models of Central Banking: How Much Convergence in Neodemocracies?”, Nuffield College, draft, 1997.
• 2
  Idem.
• 3
  This was not the only motivation. The Allied powers actively supported restructuring of the monetary order and the finances of the German State in order to prevent a new war effort. The independence of the Central Bank vis-à-vis the domestic political arena was crucial for this purpose.
• 4
  See Laurence Whitehead on this specific point, op.cit.
• 5
  This question, as it is well known, also applies to the judiciary, as guardian of the law.
• 6
  Moreover, to the extent that effective autonomy in practice has been accompanied by the exercise of a veto power by Central Banks over fiscal policy, there is no question about the specific problems raised for democratic governance and for the impact on the balance of power between winners and losers.
• 7
  For the notion of an explosive path as a consequence of the economic adjustments imposed by the external shock of 1982 combined with past policy choices see José Maria Fanelli, Roberto Frenkel & Guillermo Rozenwurcell (1994). One of the authors has developed further this approach to the case of Brazil. See Lourdes Sola (1994).
• 8
  See Lourdes Sola & Eduardo Kugelmas (1997). For our purposes, generalized indexation can be interpreted as the device by which the distributive conflicts underlying accelerating inflation were accommodated and prevented from turning into uncontrolled hyperinflation for long time.
• 9
  We note, however, many make the distinction between autonomy and independence. While the latter refers to a severence from the political arena, the former indicates the ability to resist short term political pressures. Independence never really exists in practice.
• 10
  As Whitehead shows, the latter brush aside inter alia, three important questions: 1) the convoluted history of Central Banks in dominant democracies and the variety of paths followed by them; 2) the present diversity of institutional designs which vary according to governmental rules geared to strike a balance between monetary stability and other, possibly conflicting, goals; 3) the need to identify the shifts in social interests underlying the general trend towards Central Bank independence.
• 11
  This is more so in cases like that of Brazil where a very new and minutely detailed constitution opens multiple fronts for divergent interpretations, not to speak of the infra-constitutional legislation which is still dependent on Congressional decision. For the legal constraints posed by the Constitution see Lourdes Sola (1996).
• 12
  We are assuming that deregulation of the financial system does not preclude regulation at another level on behalf not only of monetary order but also of transparency, accountability and predictability (the rule of law) pending on infra-constitutional legislation.
• 13
  We are following Albert Hirschman in his assumption that - at a certain point inflation generates a new social coalition in favor of stabilization. Our point, however, is that political intermediation and the quality of leadership are additional conditions which may be met or not but are relevant for the choice of effective policies also in the sense that they are adequate to the social profile of that coalition. This implies, of course, the choice of the economic team.
• 14
  For the Bolivian experiment, see specially James M. Malloy & Catherine Conaghan (1996). For the Brazilian case, see Sola and Kugelmas, op.cit.
• 15
  Central Bank governors should have fixed mandates which don’t coincide with an electoral calendar, and more generally their appointment should follow technical criteria predominantly.
• 16
  Alberto Alesina (1988); see also Vittorio Grilli, Donato Masciandaro & Guido Tabellini (A European Economic Forum 13) for use of such arguments in Brazil see Roberto Campos, Lanterna na Popa (1995).
• 17
  Sylvia Maxfield (1997). Her argument is specifically relevant in a context of a growing internationalization of securities markets in the 1990s.
• 18
  Maxfield’s country case studies don’t focus exclusively on the need to obtain credit. In most of her cases she blends international with national level variable, yet we contend her general theoretical orientation is ill suited for the explaining the Brazilian case.
• 19
  Not only dependent upon the federal executive, but also divided between two federal organs: the SUMOC and the Bank of Brazil prior to 1964, and the Central Bank and the Bank of Brazil after to the military coup. The exception was the short interim period during Brazil’s first military government of Castello Branco (1964-7).
• 20
  Peter A. Hall (1994).
• 21
  Once established central banks are more likely to endure in polities with high political competition and a large number of veto gates. See, Susanne Lohmann (1994); King Banaian, Leroy O. Laney & Thomas D. Willett (1986). Although note the relationship posited by these two authors between federalism and central bank autonomy would be the inverse in a context where the central bank isn’t autonomous. A large number of veto gates hinders institutional change, thus federal countries without an independent central bank would be the slowest to implement those institutions.
• 22
  John B. Goodman (1992); John t. Woolley (1985).
• 23
  Stepan and Linz makes this type of argument for federal countries, see Juan Linz & Alfred Stepan (1996).
• 24
  Malloy & Connaghan (1996) succinctly make this point.
• 25
  This would partially explain the Central Bank belated action to discipline such banks after the Real.
• 26
  Laurence Whitehead (1997).
• 27
  See Roberto Campos (1995).
• 28
  The separation between monetary and fiscal budgets introduced by the economic team under the first civilian government, 1986, made it viable to eliminate one of the black boxes created by Antonio Delfim Netto. Namely, the non-discrimination between those two budgets was a mechanism by which the Executive evaded accountability.
• 29
  For stabilization plans under the populist democracy, see Lourdes Sola (forthcoming).
• 30
  Two reasons explain this. First, continuous growth may have appeared necessary to the military in order to cement a political coalition they wished to create in support of a strategy of gradual and controlled political liberalization. Second, the choice in favor of growth at the expense of inflation and an exchange rate crisis was always taken by previous Brazilian governmental elites.
• 31
  Here we are referring to the federal executive. Congressional elections were held during most of the military period.
• 32
  Stepan makes this argument in connection to the electoral sequencing. From the standpoint of federalist devolution Brazil and Spain are polar cases. Great part of the success of the devolution strategy adopted in Spain owes much to the opposite sequence - that is to say, general elections and conditions of legitimacy were established at national level before devolution to autonomias took place. This is more striking considering Spain is a case of multinational federalism, comparatively less manageable.
• 33
  See Sola (1994 e 1996). While out of the scope of this paper, monetary and fiscal rebellion occurred in the private and public sector alike. Public and private actors often financed their investments through evading tax payments and delaying social security payments collected. Under high inflation, such practices became a source of long-term finance that the market was not prepared to supply.
• 34
  In addition to centralizing monetary authority in the Central Bank, the federal government has made state bailouts conditioned upon a series of other measures beyond monetary authority (future revenue, administrative reform, and privatization of state state-owned enterprises).
• 35
  The Brazilian Congress approved a constitutional amendment to permit executives of all three levels of government to run for re-election. Again, this institutional change further increased the power of FHC, as noted by an increased flocking of congressional deputies to the president’s party after the amendment was approved.
• 36
  The Brazilian transition to democracy was marked not only by the re-introduction of direct elections for governor prior to the presidential elections, but also by concurrent gubernatorial and national legislative elections in 1982, 1986, and 1990. Direct presidential elections were only re-introduced in 1989 - an “off’” election year. Brazilian legislative candidates during the 1980s and early 1990s therefore benefited from riding “gubernatorial coattails” rather than presidential. For the effect this electoral calendar had upon subnational influences over the national legislature, see, Linz & Stepan (1996); Fernando Abrucio, 1994. “Os barões da federação: o poder dos governadores no Brasil pós-autoritário”. Master’s Thesis, University of São Paulo; Christopher Garman, Stephan Haggard & Elisa Willis. 1996. Decentralization in Latin America. Paper presented at APSA; for the timing of elections at the national level, see Matthew Shugart & Carey. Presidents and Assemblies, 1992.
• 37
  The PSDB increased its representation in the Chamber of deputies from 38 to 62 seats (out of 513), in the Senate from one to nine (out of 54), and maybe more importantly, its gubernatorial posts from one to six (out of twenty-seven), including the most prominent states of São Paulo, Rio de Janeiro and Minas Gerais. See Jairo Marconi Nicolau, Multipartidarismo e democracia. Rio de Janeiro, Fundação Getúlio Vargas, 1996.
• 38
  IBGE. Sistema financeiro: uma análise a partir das Contas Nacionais 1990-5. Andima, 1996.
• 39
  Spending on payroll often followed an electoral calendar. Governors and mayors would swell the public sector with political appointments prior to elections in order to help elect their successors, and once in office the newly elected governments would reduce spending on payroll through allowing inflation to corrode the real the wage bill. This was done either through poorly indexing wages, or through tardy payroll disbursements.
• 40
  One could argue the federal interventions of 1987 indicated the strength of state governments in relation to the federal executive. Most governors in fact desired a federal intervention of their insolvent state banks upon taking office in 1987 (the only opposition came from Rio de Janeiro’s Brizola), for they were under the correct assumption their banks would be returned to them financially healthy at the end of their administration intime for the next elections. How the central bank decides to use their ability to intervene state banks, however, is analytically distinct from the incremental process we are trying to depict in which the central bank gains more tools to discipline state financial agents. For the willing participation of governors in the 1987 interventions see: Folha de São Paulo 1 /31/88, Istoé 3/4/87, Correio Brasiliense 8/10/87.
• 41
  CMN Resolution 1718 of 5/29/90, CMN Resolution 1775 of 12/26.
• 42
  Lei da Improbidade - Código Penal - Law 8429 of 6/2/92.
• 43
  Senate resolution n. 94, of 12/15/89; Senate Resolution n. 58 of 12/13/90; and Senate Resolution n. 36 of 6/32/92. In 1993 a Constitutional Amendment (n.3, 3/17/93), was further passed which prohibited the emission of any new state bonds except for the payment of principal. See Eduardo Refinetti Guardia, 1995, “A crise fiscal dos estados” Caderno PUC Economia, 3, pp.65-87; and Gilton Carneiro dos Santos, “Contingenciamento do crédito e finanças estaduais”, Programa de Formação de Técnicos do Banco Central, Curso de Formação Superior, 1993.
• 44
  Since this paper is focusing on monetary authority, we limit our discussion primarily to how federal bailout of state finances has affected Central Bank credibility in the financial sector. The current rescheduling of state finances, however, involves many elements which go beyond the monetary sector. In order to receive federal financing, state governments are further being obligated to yield guarantees in future revenue (from own taxes and constitutionally allocated shares of federal taxes) and assets of their state-owned enterprises.
• 45
  In the Senate confirmation hearing for the President of the Central Bank, Persio Arida, for example, declared he was in favor of privatizing state banks in order to sustain price stability. Jornal do Brasil, 12/14/94.
• 46
  Since prior to 1997 governors and presidents were not permitted to run for re-election, an election, by definition, constituted a turnover in government.
• 47
  The Central Bank intervention of both banks was also helped by the fact both incoming governors Mario Covas of São Paulo, and Marcelo Allencar of Rio de Janeiro belong to the same political party as the president (PSDB). The fact both belong to the same party, however, could influence the outcome in either way. One could equally expect the governor to reduce a state “rebellion” against the federal government, as one would expect the federal government to give the governor favorable treatment.
• 48
  The purchase of Bamerindus by HSB is a case in point.
• 49
  As it will become clear, the two cannot in practice be separated. Because state banks have been the administrators of state government debts, negotiation over state banks by the central bank often must be made in tandem with a restructuring of the state government debt - a Senate prerogative.
• 50
  Brazil ‘s gubernatorial elections were first opened in 1982, with subsequent elections for the governor’s office held every four years; 1986, 1990, and 1994. The first Central Bank bailout of state banks occurred in 1983, followed by a series of interventions in 1987 and 1991. These crises, however, also coincide with the various stabilization programs adopted over the last 10 years: the Plano Cruzado of 1986, and the Plano Collor of 1990. For explanations which focus on the electoral connection see Gustavo Loyola, “Os Bancos Públicos Estaduais”, Banco Central, 1992; and for an analysis which focuses on the impact of stabilization, see Wadico Buchi, “Banespa: Uma Visão Realista”, mimeo, published in Folha de S.Paulo: 9/17/95.
• 51
  O Estado de São Paulo, 1/17/93.
• 52
  CMN - Conselho Monetário Nacional-vote 233 of7/20/83 for PAC and CMN Vote 446 of 4/4/ 84 for PROREF.
• 53
  For an analysis of how state banks repeatedly ignored central bank stipulations established in those packages see Eduardo de Carvalho Andrade. “Os bancos comerciais estaduais no Brasil: do final dos anos 60 à crise dos anos 80” Master’s Thesis, PUC-RJ. 1992. See also Gustavo Loyola, “Os Bancos Públicos Estaduais”, op. cit, Paes, op. cit. See also Francisco Lopreato, “Crise e Financiamento dos Governos Estaduais 1980-88, Campinas, Unicamp, IE; Ogasavara, R. & J. Vasconcelos, “Análise econômico-financeira dos bancos estaduais”. Documento de Política, nº 7, IPEA, Rio de Janeiro, 1992. See also, World Bank, “Brazil. The dilemma of Brazil’s state banking system: an analysis and suggestions for reform, February 27, 1990.
• 54
  Law 4595 of 12/31/1964.
• 55
  The category of extra-limit debt was created under the 1975 Central Bank Resolution 446, which interestingly coincides with the military government’s decision to slowly liberalize the political sphere. For a direct analysis over how the military government’s decision to induce state and municipal debt through the use of state banks served their political interests, see Christopher Garman, forthcoming electoral dissertation, University of California, San Diego.
• 56
  Provisional Measure nº 1514, 8/7/1996.

• 57
  JEL Classification: E58.

Publication Dates

• Publication in this collection
  27 June 2022
• Date of issue
  Apr-Jun 1998