Joan Nelson

Political Capacities and Economic Reforms

Introduction

The politics of economic reforms have been much analyzed over the past two decades. The question of what political capacities and institutional arrangements are key to effective reforms has been one major focus of attention. During the l970s and l980s there was an on-going debate between those who asserted that only authoritarian governments could sustain sufficient macro-economic discipline to manage economies effectively, and those who challenged that view. By the late l980s, it was quite clear that broad generalizations about types of regimes -- democracies versus authoritarian systems -- were far too crude to offer useful generalizations and explanations. A much narrower version of the old debate persisted, however, in the effort to determine whether effective economic reforms required considerable concentration of executive authority and power (within the framework of more or less democratic as well as authoritarian systems). Some writers focused on specific mechanisms to shield decision-makers from political pressures and raise the costs of policy reversals, such as independent central banks. Other analysts focused on the effects of different numbers and kinds of veto points or " gates " in different systems or the effects of varied party systems. But our knowledge of the links between types of political institutions and the process of economic reforms remains at best sketchy.

One reason for our lack of knowledge is the fact that " economic reform " is not a sharply defined and homogeneous concept, but rather a broad and varied array of measures, operating in different arenas and on different time-tables and involving very different sets of actors. Moreover, the reform agenda evolves over time. Later measures become necessary to consolidate and win benefits from earlier ones. Different kinds of reforms pose different political challenges, and require different technical, administrative and political tactics and capacities. Section I sketches the contrasting character of different parts of the reform agenda, and explores some of the political implications.

Our ability to find clear generalizations about political institutions and capacities for reform is hampered by a second fact: the political context or climate for reform varies dramatically across countries and changes over time within countries. As a result, similar political institutions have different effects at different times or in different countries. To state the same point slightly differently: the, politics of economic reforms is profoundly influenced by widespread mind-sets within each country regarding reforms B mind-sets that are an outcome of each country ' s history and traditions and the nature of its experience with economic (and often political) decline and crisis. These aspects of the political context are not captured by examining countries ' varied political institutions and procedures, yet they powerfully shape the depth and duration of the reform effort. Section II discusses some of the ways that the political climate for reform is shaped by pre-reform circumstances, the shift from crisis to normal politics, and growing reform skepticism.

With these dimensions of variation in mind, Section III explores some of the ways that political institutions affect incentives and capacity for reform. The final section suggests some scenarios for late-stage reforms.

I. The Changing Reform Agenda and its Political Challenges

The broadening reform agenda.

The international debt crisis of the early l980s led multinational agencies, the governments of wealthy nations, and a growing number of poorer nations to adopt a reform agenda intended to restore economic stability, restart growth, reduce debt to manageable proportions, and restructure economies to reduce their vulnerability and improve prospects for sustained growth. This " international reform agenda " expanded dramatically in the course of the l980s and l990s.

At the beginning of the debt crisis, attention focused on macro-economic stabilization measures. That initial task was quickly expanded to include structural changes regarded as essential to restore growth and reduce debt. John Williamson ' s 1989 summary of the " Washington Consensus " listed, in addition to fiscal, monetary, and exchange rate measures, reforms to reduce government intervention and permit markets to function more effectively, including trade and financial liberalization, increased receptivity to foreign direct investment, deregulation, and privatization. These structural changes mostly entailed dismantling government regulations and restrictions on private economic transactions. The closest the Consensus came to more complex institutional reforms was the rather tentative inclusion, as the very last item, of " property rights protection. " Williamson noted that this was intended to signal recognition that institutional features were also important determinants of growth. By l989 the World Bank was beginning to use a broader concept, " creation of an enabling environment [for effective markets]. " Williamson remarked that concept might be preferable, but it remained largely undefined.

More than a decade later, at the beginning of the new century, the reform agenda has ballooned to include a broad array of institutional reforms, and to emphasize poverty reduction as well as growth and stability. Responsible macro-economic management and reduced state intervention in the economy remain crucial, but they are now viewed as far from sufficient for growth and poverty reduction. Reform of the state itself, including the civil service, the police, the system of justice, and reduced corruption are part of the essential " enabling environment. " Social sector reforms in pensions, health and education, as well as far-reaching changes in labor markets and industrial relations are also squarely on the expanded international agenda. These further reforms are much more demanding than the initial agenda: they require not merely the dismantling of regulations, tariffs, and subsidies but fundamental changes in the design and operations of core public functions and institutions.

Individual countries tended to recapitulate, in their own reform histories, the broadening of the international reform agenda. Except in those few countries (like Colombia) that avoided severe problems of destabilization and debt, reforms initially focused on macro-economic stabilization. Where those efforts were successful, reform-oriented governments usually moved quickly to varying degrees of trade and financial liberalization, opening of capital accounts, privatization, and deregulation. Only later did attention turn to complex institutional changes in social sectors, labor markets, and reforms of state institutions. Action on these " second generation " or " late stage " issues was often limited and indecisive. The sequence of reforms, of course, was not clear-cut. Macro-economic management demanded constant vigilance and periodic tightening. Liberalization and privatization programs continued to unfold over many years, even while governments turned to later-stage reforms. But reform efforts in most countries broadly followed the three stages: initial stabilization; some degree of liberalization; and later and slower institutional reforms in social sectors, labor markets, and reform of the state itself.

The varied characteristics of different kinds of reforms

Different kinds of policy and structural reforms pose different political challenges. It is useful to think about a spectrum of types of reforms, arrayed according to several characteristics:

i. The degree to which reform decisions are guided by a clear model or vision supported by a strong technical consensus, or at least by clear parameters for debate.

ii. The number and variety of actors who must collaborate in order to design and implement major changes.

iii. The length of time needed to implement reforms, due solely to technical requirements, even if there were no political delays.

iv. The kinds and amounts of detailed information required to fine-tune and implement each successive stage of the reform, and the extent to which mechanisms already exist for collecting that information.

v. The evident costs of postponing action (or, more broadly, the incidence, timing, and certainty of expected costs and benefits).

At one end of the spectrum are certain macro-economic reforms, especially those centered on price changes such as devaluation or interest rate adjustments. In inflationary contexts, these measures are guided by a fairly clear technical consensus, despite sharp debate about precise timing and design details. The key changes can be put into effect by a handful of high economic officials (with the backing of top political authority). They require little organizational or institutional change, and they take effect extremely rapidly. They require certain basic economic indicators, which are available in all but the least sophisticated countries. Postponing action has obvious and high costs.

Liberalization measures such as reduced average level and dispersion of tariffs, or reduced controls on financial sector operations entail more legal changes and some organizational restructuring. While the broad direction and objectives of change are

fairly well-established, there is considerable room for debate and adjustment of precise design to specific country circumstances. A wider range of officials and some private sector actors must collaborate on the reforms, and they are likely to take several or many months to put into effect. Deeper financial sector reforms require considerable detailed information; mechanisms to collect and analyze that data often need to be created or upgraded. The costs and benefits of rapid versus slower reforms are less clear-cut than for macro-economic measures.

At the far end of the spectrum are systemic reforms in the major social services, primarily education and health care finance and delivery. Multiple models are available, influenced by very different national and regional traditions and histories. More important, there is only limited consensus among technical specialists regarding basic principles of reform. Experts argue bitterly over the merits of, say, single-payer health care systems or charter schools. They agree only very partially on the principles that should guide the degree and design of privatization or decentralization. Therefore, public debate regarding the design and priorities of reform tends to be diffuse and inconclusive.

Even after initial agreement is reached regarding social service reforms, implementing them is extremely complex. Executive agencies and legislatures at national, state, and local levels are usually involved. Reforms intended to increase efficiency and save money in the long-run may nonetheless have high up-front costs. Not only the Ministry of Finance but often sub-national financial authorities must concur. Many social sector reforms require years to implement. A great deal of detailed information is required to fine-tune design of successive steps. Much of that information is not available without new arrangements to gather it. All of these complications are reinforced by the fact that, even where there is widespread dissatisfaction with the status quo, postponing action does not carry obvious and prompt risks.

The varied character of different reforms -- availability or absence of a consensus model or clear parameters for debate, timetable, number and variety of actors, information requirements, apparent costs of delay -- shape the political challenges. If many actors must co-operate to put a reform into effect, any one of them can weaken or stop the reform. In other words, there are many potential veto actors. Decisions taken by the executive run high risks of being blocked in the legislature or sabotaged in the course of implementation. Moreover, the large number of actors increases transaction and enforcement costs. If implementation takes many years, there are many potential veto opportunities. The length of time required to get most complex institutional reforms up and running also means that the benefits of the reforms may not become apparent for some time. Therefore it may be hard to mobilize pro-reform coalitions to counter opposition from vested interests, which are likely to resist from the outset. Information requirements also affect the course of reform. Lack of information may stall action; new information may alter perceptions and reopen debates. Complex institutional reforms are the result of an extended process, not an event. The process is subject to stops and starts; issues regarded as closed may be re-opened and steps already taken may need to be repeated. The process is not linear, but iterative.

In short, different kinds of reforms pose quite different political challenges, for reasons intrinsic to the character of the reforms themselves . The fact that late-stage reform agendas concentrate on complex institutional reforms helps to explain why the pace of reform in most countries almost always slows substantially after initial stages.

The varied characteristics of different kinds of reforms also suggests why reforms in some sectors have made much more progress than others, in cross-national perspective. For example, far-reaching pension reforms have been adopted in many more countries, in and beyond Latin America, than have introduced similarly basic changes in education or health care systems.

There are many reasons for the contrast; only two are noted here. First, spreading awareness of the Chilean model by the early l990s, and the World Bank ' s endorsement of mixed public-private approaches provided a clear alternative to traditional Pay As You Go (PAYG) arrangements. The two alternative paradigms became the focus of reform debates. As already noted, no similar " dominant reform model " exists for education or health care and finance. Reformers are divided among many possible approaches, and debate tends to be fragmented and inconclusive. Second, pension reform is facilitated by the comparatively simple and centralized administrative structure required to launch a partly privatized system. In sharp contrast to the lengthy implementation of major social service reforms, once pension reform is approved by the legislature, it can be put into operation quickly -- in some cases within six months. Thus the politics of implementation are minimized. Much more could be said about the politics of pension reforms, but the point here is to illustrate how the characteristics of different sectors shape the politics, and therefore the probability and pace of reform.

The phases of specific reform efforts.

Detailed histories of specific reform efforts suggest an additional layer of variation in politics. Most reforms must move through several phases. At a minimum, there is an initial stage (often lasting many years) during which the individuals and groups concerned about a problem try to get it on the agenda of the executive branch. Initiative may come from within or outside of the government. Once some of the appropriate officials believe the issue demands action, the main task of the second phase is reaching (sufficient) agreement within the executive on a concrete design. Unless the reform can be put in place entirely through executive decree, there is a third stage of winning legislative approval and adequate public acceptance. The fourth phase, launching and sustaining implementation, may be fairly simple and rapid (for instance, putting into effect a new set of tariffs) or may take years and involve multiple agencies and levels of government, like the sweeping health sector reforms adopted in Colombia in l993. The key players and arenas for action, the major political challenges, and the relevant tactics differ in each phase. Table 1 provides a schematic sketch of the political tasks and tactics associated with each phase. Since the tasks change (despite considerable overlap), the mix of capacities required to push reform forward changes from one phase to the next.

Conclusions: Political effects of the changing agenda.

To move beyond the broadest generalizations regarding the politics of economic reform and the capacities required to promote them, the concept of " reform " itself must be taken apart. Different kinds of reforms pose quite different political challenges. Even the different phases of any specific reform entail different political tasks and demand different tactics and capacities.

Discussions of the politics of reform often fail to recognize these variations. Many economists used to B and some still do B talk about " political will " on the part of top-level leaders as the necessary and sufficient requirement for effective reform. Some of the metaphorical language used in discussions of reform convey a similar message: " bite the bullet " , " just do it. " That implicit image of the reform process may roughly describe a single-shot devaluation decision. But it is clearly very misleading for more complex measures. Recognizing the varied character and political challenges of different reforms, and the tendency for complex institutional changes to be late and slow are first steps toward understanding why some kinds of reforms move faster than others, and why the pace of reforms tend to slow down almost everywhere.

II. Elite and Public Attitudes Toward Reforms

Shifts in public attitudes toward reform are a second set of factors effecting the depth and persistence of the reform effort. Especially in democracies, widespread public attitudes and perceptions toward reform create a favorable or unfavorable political climate for change. Measures and strategies that are feasible in a favorable climate may be much more difficult and costly in unfavorable climates of opinion. Attitudes toward reform vary widely across countries. They also change in partly predictable ways within countries, in the course of sustained reform efforts. Evolving public attitudes are important factors in explaining why some countries have pursued rapid and far-reaching structural reforms, while others have followed a much more gradual path, and still others have repeatedly started but then abandoned reform efforts.

The political effects of pre-reform conditions.

There is no one-to-one connection between objective measures of economic problems and reform responses. Some governments adopt corrective measures in reaction to rather mild adverse trends or signals; others fail to act decisively for years or even decades. In many countries, only persistent and increasingly severe economic (and sometimes political) pressures and the failure of earlier modest corrective measures have created conditions for strong reforms. Pre-reform conditions help to explain not only whether a government launches reforms, but also the vigor and durability of the reform effort.

Recent evidence on this point comes from a massive research program conducted by the Economic Commission for Latin America and the Caribbean (ECLAC), comparing reform efforts and outcomes throughout the region. The program focuses on five categories of " first generation " structural reforms: trade liberalization, liberalization of domestic financial markets, opening the capital account of the balance of payments, privatization, and tax system neutrality. The study is particularly interested in the effects of reforms in these five areas on growth, equity, and employment. It gives special attention to nine countries with long reform histories. Four of the nine B Argentina, Bolivia, Chile and Peru B are described as " aggressive reformers, " undertaking a wide range of reforms quite rapidly. The other five B Brazil, Colombia, Costa Rica, Jamaica, and Mexico B are classified as " cautious reformers. " What most clearly distinguishes the aggressive from the cautious reformers are indicators of the severity and duration of economic and political difficulties before market-oriented reforms were launched.

Table 2 is reproduced here from the advance summary of the forthcoming overview volume.

All of the aggressive reformers had suffered hyperinflation just before launching what proved to be the initial steps in a sustained and vigorous course of reforms. All had experienced negative average growth of GDP for the five years preceding reforms. All, according to the estimates of the authors, were characterized by " low governability " (defined as " ability to make and carry out policy decisions " ) prior to reforms. Put slightly differently, Chile had suffered something close to a civil war; governments in Bolivia in the early l980s and Argentina in the late l980s were virtually paralyzed by strikes, protests, and a disintegration of governing institutions; Peru was increasingly convulsed by Shining Path terrorism.

The cautious reformers, in contrast, had faced less acute economic and political difficulties. In the five years prior to launching reforms, all five averaged slow or moderate income growth; Brazil and Colombia had averaged roughly 4.5% a year. Over the same period, average inflation had been quite moderate in Colombia, Costa Rica, and especially in Jamaica. Brazil ' s average inflation was considerably higher, though still below most of the aggressive reformers; its impact was also substantially cushioned for elites and middle classes by extensive indexing. All of the cautious reformers had experienced some protests and other political strains, but none approached the point of political paralysis.

Prolonged and increasingly acute economic and political disintegration in the four aggressive reformers convinced a growing part of elites and the general public that the countries ' problems were due not to the incompetence or wrong-headedness of specific leaders nor the vagaries of weather and international prices, but to deep-seated structural problems that had to be corrected. Like much of the citizenry of Eastern Europe and the Soviet Union in l989, people became convinced that old formulas and ideologies no longer worked. Radical changes were, indeed, not only necessary but urgent: hyperinflation and political chaos were terrifying, and people desperately wanted governments that could take charge and reverse the disintegration.

In the more cautious reformers, in contrast, specific crises (like the unprecedented and frightening loss of value of the Costa Rican peso in l981) led to public demand for specific action. But there was no widespread sense that the economic system itself was fundamentally flawed. On the contrary, there was considerable pride in the nation ' s democratic political system and extensive social benefits. Somewhat similarly, in Brazil, public and elite opinion clung throughout the l980s to the belief that the Brazilian economy was so large and strong that while short-run macro-economic adjustments might be needed, more basic changes in policy and structure were not. Only around the time of Collor ' s election in l990 was the belief emerging that the country faced a " deep crisis of the inward-looking, state-driven growth model developed in the l930s. " In general, in the cautious reformers "[T]he central actors believed the countries were basically sound and had much worth preserving. "

In short, a country ' s recent economic and political experience affects the interpretation of current economic difficulties. If most people believe that today ' s economic problems are temporary, or traceable to some specific and correctable flaw, proposals for far-reaching reforms will be resisted. If, in contrast, the view is widespread that there are basic structural weaknesses that must be addressed, aggressive reform is far more likely to be tolerated or even welcomed.

From crisis politics to normal politics.

Even in countries where much of the elite and public are convinced that reforms must be broad and deep, tolerance for reform is likely to dwindle over time. Where much of the public feels that the country is in crisis, political leaders (and especially new leaders, whether or not elected) have considerable political space to maneuver. In such circumstances, groups opposed to specific reforms may find themselves isolated, without allies. Legislators, unions and other interest groups, and the general public may acquiesce not only in measures, but in decision processes they would normally protest. In Argentina, Menem used the decree powers of the Presidency almost 60 times in l991, by-passing the legislature. In Poland a year earlier, then-Minister of Finance Leszek Balcerowicz sent twenty-one complex and far-reaching bills to the legislature, demanding action on the entire package within days.

Balcerowicz, architect of Poland ' s initial reform period and currently Prime Minister, describes this period as a " window of opportunity " , and emphasizes that it is brief. If the government ' s initial efforts fail, its credibility is damaged and it is likely to face growing difficulties. Success in containing the crisis enhances the government ' s political capital, but ironically may also encourage the feeling that more normal interests can safely be asserted.

Beyond reform fatigue: reform skepticism.

Analyses of the politics of economic reforms often refer to reform fatigue: the decreased tolerance of citizens, legislators, and politicians for disruptive change. Beyond reform fatigue, a different and more fundamental shift in mood is now emerging. Opinion polls indicate that the broad thrust of market-oriented reforms has been quite widely accepted in countries like Argentina, Bolivia, Chile, Mexico and Peru. But in several of these countries there is also a growing malaise, a deep concern over some of the outcomes of the reforms: increased inequality, high unemployment and increased insecurity as the social and labor relations protections that used to benefit large middle and formal sector working classes have eroded.

A recent (l999) large cross-national survey of fourteen Latin American countries found that in all but two, a majority of respondents felt they lived less well than their parents had. Only in Brazil did as many as a quarter think they were doing better than their parents. In Peru, survey data permit comparing actual changes in income over roughly the past decade, to the respondents ' perceptions of gain or loss. Surprisingly, sizable majorities of those whose incomes had risen substantially by objective measures felt that they had lost ground. More generally, workers with formal sector jobs feel vulnerable to layoffs; and growing numbers are on temporary or part-time contracts.

In a recent paper, Dani Rodrik suggests three basic sources of this widespread insecurity: the traumatic depressions of the l980s; heightened vulnerability to external shocks as a result of open economy reforms and other factors; and declining confidence in social and political institutions capable of responding to demands for improved economic security. Despite the consolidation of formal democratic processes in most of Latin America in the l990s, the sclerosis or weakening of traditional political parties in many countries, the decline of unions, and the erosion of corporatist arrangements make much of the middle and working classes feel they no longer have effective channels for voice or participation in decision-making. Surveys indicate much smaller proportions of voters claim a clear party affiliation. Many factors have contributed to the decline of traditional parties. Global shifts in intellectual currents and the collapse of communism have played obvious roles, as has the decline in union membership and power. Neo-liberal reforms in many countries have dried up some traditional sources of patronage. At the same time, neo-liberal reforms have contributed to an increased concentration of wealth in many countries. It is widely assumed that there is a corresponding increase in the political influence of the very wealthy. In most of the region, formal and informal power structures and processes have not proved flexible and innovative enough to cope effectively with rapidly evolving pressures of globalization and the social demands and conflicts growing out of changing economies.

III. How do political institutions and procedures affect economic reforms?

Even first-generation reforms require some minimum level of political stability and administrative capacity. It is hard to imagine an effective program of structural adjustment in Somalia or Haiti. Yet effective reform programs have been launched and sustained in countries with rather precarious political stability and limited administrative resources. An observer of Bolivia in l985 might understandably have been skeptical about the capacities of the system to carry out a sustained reform effort. Bolivia had seen seven military and two weak civilian governments between l978 and l982. From l982 to l985, the government of Hernan Siles Zuazo was whipsawed between the influential employers ' association and militant union protestors who repeatedly occupied the capital. Yet from l985 on, with some hesitations and interruptions, Bolivia has carried out fairly sustained and aggressive reform programs. We are left with an untidy conclusion: above some fairly low threshold level, political stability and administrative competence facilitate but are not essential for first-generation reforms.

What can we say about the structure of political institutions, as distinct from their stability? We have discarded the old unhelpful debate as to whether democracies are capable of economic reform. But it is still widely asserted that strong executive authority, insulated from excessive political pressures, is crucial for at least the early stages of reform. To what extent is this assertion true?

Virtually all significant structural reforms require executive branch initiative and guidance. This is true even though ideas for reform often originate outside of the government, and despite the fact that the main impetus for reform sometimes comes from business or civil society. Government initiative and guidance is almost always crucial even for reforms that rely substantially on non-official groups or agencies for effective implementation.

In ECLAC ' s four aggressive reformers, individual leaders strongly committed to reform played key roles. It is hard to imagine the sweeping policy re-orientations in Argentina in the l990s without Minister of Finance Domingo Cavallo backed by President Carlos Menem ; in Bolivia in the late l980s without Minister of Planning Gonzalo Sanchez de Lozada backed by President Victor Paz Estenssoro (and in a later administration Lozada as President); in Chile in the l970s and l980s without Sergio de Castro, Hernan Buchi and others backed by Pinochet; and in Peru in the l990s without President Fujimori. At first glance these cases (and others outside of the region) support Arnold Harberger ' s formula that a " handful of heros " accounted for the " secrets of [reform] success. "

But committed leadership is far from a sufficient condition for effective reform. Leaders like President Carlos Andres Perez of Venezuela or, more recently, Minister of Finance and later President Henrique Cardoso of Brazil have not lacked commitment to far-reaching policy and structural changes. Their success was (or is) limited not only by the climate of public opinion discussed above, but also by the political institutions and context within which they work.

The simplest question regarding institutional constraints is to what degree the chief executive is constitutionally and legally empowered to act alone, by-passing other branches of government. Political leaders in two of ECLAC ' s aggressive reformers could act with comparatively few constraints. Pinochet of course dismantled the constraints of competitive elections and subordinated Chile ' s legislature and judiciary. Fujimori ' s powers were less sweeping, but by l993 he had essentially destroyed the autonomy of Peru ' s legislature and judiciary. In contrast, executives in Argentina and Bolivia had to operate within the constraints of democratic institutions. However, Menem made extensive use of his decree powers, enacting fifty-nine decrees of " Necessity and Urgency " in l991 alone, and an additional 36 in l992. He also used other formal powers, like the line item and total veto.

The chief executive ' s commitment and his powers to act alone are most relevant for measures that are administratively simple and can be put into operation quickly. As noted earlier in this paper, complex institutional reforms require broader consensus, because more agencies and groups must co-operate to put them into effect, and because implementation takes a long time. Therefore, even where the chief executive has substantial decree powers, building consensus within the government itself is a step that cannot be by-passed. For instance, even under Pinochet in Chile, a coherent and committed change team put years of effort into recruiting allies and developing support within relevant government agencies for proposed health care finance reforms. More obviously, most complex reforms require legislative action, and are subject to some degree of judicial review. At these stages they may encounter a wide range of institutional and procedural obstacles.

Some of the constraints on executive authority take the form of " veto gates " : constitutionally or legally determined points in the policy process where decisions and actions can be blocked by individuals or groups. Democracies vary widely with respect to the veto gates built into their political institutions. The structure and procedures of legislatures (unicameral versus bicameral; powers of the speaker and of committee chairmen; etc.) and the rules governing relations between the chief executive and the legislature (most broadly, in parliamentary versus presidential systems) are among the most obvious determinants of veto gates. The extent of judicial review of laws and executive decrees permitted or required by the constitution creates additional veto gates, and can be particularly daunting where the constitution is extremely detailed. The division of powers between the national government, provinces, and municipalities also establishes veto gates. Where citizens have the constitutional authority to demand a referendum (as in Uruguay) or to recall legislators under certain conditions (as in some U.S. states), these too constitute potential veto gates, though they are cumbersome and therefore tend to be used rarely.

Veto gates in a country ' s institutions affect political leaders ' estimates of the prospects for any specific reform, the costs of trying to push it through, and the tactics that can be used and are likely to be effective. Sometimes political leaders try to alter institutions, temporarily or permanently, so as to limit the scope or reduce the number of veto gates. Menem, for instance, won legislative approval for a law increasing the number of Supreme Court members from five to nine, thus ensuring that he appointed four (with the agreement of the Senate); a later resignation gave him the opportunity to appoint a fifth judge. In Jamaica in December l983, Prime Minister Seaga called a snap election which the opposition decided to boycott, creating a temporary one-party Parliament. In Bolivia in l985, President Paz Estenssoro split his cabinet into economic and non-economic segments, in effect making reformist Minister of Plan Gonzales Sanchez de Lozada prime minister for economic affairs.

However, political leaders tailor their measures to pass veto gates much more often than they try to change the institutions. For example, the design of pension reforms in many countries has been carefully crafted to meet anticipated court challenges regarding constitutionally guaranteed " acquired rights " of individuals. Still more obviously, officials promoting reform must assemble sufficient legislative votes to pass the necessary laws, and therefore must ensure support within their own party or coalition, and to the extent possible from other parties, interest groups, and public opinion. Political institutions and procedures not only define veto gates, but also establish the schedules, rules, and many of the incentives of political activity more broadly. The electoral cycle and rules governing re-election determine politicians ' time horizons and the length of time during each government ' s term of office available for legislative initiatives. Electoral systems -- proportional representation, single member districts, and the like B powerfully affect the number and discipline of political parties, the incentives driving legislators ' behavior, and the character of relations between congress or parliament and the executive. For example, Brazilian electoral rules and traditions generate a fragmented party system and undisciplined parties, forcing the president to negotiate with individual congressmen to build issue-specific majorities for each piece of major legislation.

A large part of the extensive writing on the politics of economic reforms focuses on the interest groups that win or lose, and the tactics reformers can use to neutralize the losers and mobilize the winners. Javier Corrales points out in a recent article that most of this analysis neglects the impact of market-oriented reforms on parties with deeply ingrained statist biases, and the strategies that reformist presidents have used to try to overcome their resistance.

Corrales argues, further, that in many Latin American countries " .. the key conflict to resolve is between the executive and the ruling party. " Whether and how this conflict is managed will affect the government ' s capacity to carry through reforms.

Two general approaches are available (other than abandoning reforms) : by-passing the party, or shaping reforms to accommodate and compromise with core party concerns. In the short run, a reformist president ' s scope for by-passing his own party is determined by formal institutional structure and procedural rules, as discussed above. But in the medium term, the beliefs, traditions, and internal structure and incentives of the party itself also affect the degree to which a president can ignore his party without provoking confrontation and perhaps reversal.

For instance, in Venezuela at the beginning of the l990s, President Perez appointed friends and technical specialists to many key positions in the government, tended to disparage the " old regime " and failed to consult Accion Democratica (AD) on major decisions. AD was accustomed to long periods in power, and dependent on rents from statist policies. Moreover, AD > s internal management was powerful and deeply entrenched. Alienated from the government, party leaders attacked the reforms and refused to grant special powers to handle the economic crisis. At the AD party congress in autumn l991, orthodox anti-reform leaders were elected to all the top party positions. Thereafter, AD opposition stiffened into confrontation, leading ultimately to the government ' s downfall.

In Argentina, Menem initially followed a somewhat similar strategy of neglecting and by-passing the Partido Justicialista (PJ), while seeking alliances with the private sector (Bunge and Born), neo-liberal parties, and international actors. The Peronists were outraged, and by late l990 relations between Menem and the JP were severely strained, eroding fragile business confidence. Menem then switched strategies. At a PJ party congress in March l991 he emphasized his fidelity to Peron and interpreted his policies as updated Peronism. He also negotiated a tacit pact with the party, agreeing to submit most of his program to congress and to permit congress to modify bills, but reserving the right to veto unacceptable decisions. That course of action entailed concessions to some key concerns of the party and its union allies. Peripheral provinces (the traditional mainstay of Peronist voting) were largely sheltered from the costs of adjustment through tax and subsidy concessions, while greater Buenos Aires bore the brunt of adjustment. Labor unions won repeated postponement and dilution of measures to liberalize labor markets and reform the Obras Sociales (the largest source of union funds, via Social Security contributions). But the number of strikes dropped dramatically in l991 and the PJ discouraged the most ardent attacks on neo-liberal reforms. As a result, the credibility of the reforms was strengthened, because continuity began to look more assured. Corrales concludes that in countries with strong statist parties, durable structural reforms cannot be achieved by by-passing the president ' s party. There must be substantial accommodation, even at the price of compromising aspects of the reform agenda. That course not only postpones important aspects of the reform agenda, but also is likely to require later revisions in reforms that are approved. But it does make possible some forward motion.

The contrast between Perez ' experience with AD and Menem ' s handling of the PJ hinges in large part on Menem ' s choice of strategies, but also reflects differences between the two political parties themselves. In contrast with AD ' s complacency in Venezuela, the Peronists had suffered two electoral defeats in the l980s, provoking intense internal debates and considerable turnover in leadership. The party ' s tolerance for new ideas, Corrales suggests, was somewhat stretched, though its fundamental stance was not altered. It is also probable that the long decline (from at least the early l980s) in union membership and power, and the shocks of two rounds of hyperinflation contributed to greater flexibility within PJ circles.

Attempts to identify the types of party systems which facilitate reforms have been inconclusive. For example, Barbara Geddes has argued that where parties rely mainly on patronage and clientelism to assemble votes, they will adamantly resist civil service reforms (or indeed any reforms that reduce their command over patronage resources) , except where there are two major parties with roughly equal power. Under those conditions, she suggests, politicians may judge that reforms will hurt their opponents about as much as themselves. The hypothesis has drawn considerable attention, but is not convincingly supported by the evidence. Many countries have adopted patronage-reducing reforms in the absence of evenly matched parties, while such reforms have stalled in some two-party systems. More broadly, a persuasive theoretical case can be made for the advantages of two-party systems over either dominant parties or more fragmented party patterns. Yet the empirical evidence linking different party systems with different records of reform is not convincing. Similarly, years of (speculative) debate over the relative merits of parliamentary versus presidential systems in Latin American circumstances failed to reach compelling conclusions. Analyses of the strengths and weaknesses of different versions of democratic systems are most useful in highlighting the specific causal mechanisms (for instance, the incentives and disincentives for party leaders or for voters) that are likely to result from particular institutional arrangements. How any specific institution actually operates in a country is always powerfully influenced by other aspects of the political system, by the social and economic context, and by national history, traditions, and values. The search for " optimal " institutions in general probably is not helpful, but debates over modifications in a country ' s own institutions can benefit from comparative analyses of the effects (and causal mechanisms for the effects) of various alternative arrangements in other democracies.

IV. Scenarios for Sustained or Revived Reform

The launch phase of economic reforms is often accompanied by a " negative consensus " : the urgent desire to get a crisis under control, and sometimes the further conviction that old economic strategies and institutions no longer work. Later stages of reform may demand the creation of a more " positive consensus " on the principles and goals that should guide further measures. People begin to ask, " What kind of a society are we creating? " A credible portrayal B a rough vision B of an altered but desirable society may be essential to deepening and consolidating reforms.

The market-oriented Washington consensus of the l980s, with its focus on increased efficiency, certainly did not provide an adequate vision. The broadened agenda of the l990s, with its strong emphasis on reduced poverty and better governance, is more compatible with national visions of better societies, but is little more than a broad sketch of directions. One major element of any such vision must be a recrafted social contract, a redefinition of the scope and form of the state ' s responsibility for social insurance, reduction of inequality, and relief of poverty. A second major element must be a workable plan for the rehabilitation of the state, so that it can effectively fulfill its on-going and crucial responsibilities for regulation, infrastructure, and social services and welfare. In many Latin American countries, these two sets of issues are now at the core of political debate and conflict.

There are no clear technical models or solutions to these issues. Still less are there political formulas for building consensus and overcoming vested interests intent on blocking change. Democracies have always struggled with the tension between decisiveness and participation, between ability to act and necessity to build consensus. (Crises temporarily shift the balance in favor of decisiveness.) Countries with institutions that facilitate decisive action by strong leaders are likely to press reforms faster and deeper than systems which encourage or require broad participation in major policy choices. But in democracies, decisions taken with little consultation are subject to reversal, and major institutional reforms are particularly vulnerable to this risk. In New Zealand, for instance, radical changes in the health sector were introduced by an independent commission in the early l990s, with little consultation with the Ministry and little attention to strong public opposition. Public pressure and administrative and technical difficulties soon caused aspects of the reforms to be modified or abandoned; major features were then reversed after the next election.

Systems with many channels for voice and multiple veto gates run different risks: paralysis in the face of mounting problems, loss of credibility, and ultimately a turn to populist leaders. That downward spiral is not inevitable: a number of factors can intervene. Long-entrenched popular ideas do change. Repeated crises may force such change, but reformers ' persistent efforts to educate public opinion on specific issues also have an impact. In the United States, there is a growing B though still inadequate B public understanding that sky-rocketing costs of medical care and the growing number of uninsured cannot be effectively addressed using traditional finance and delivery arrangements.

New interests and alliances, in part a result of changing economic structure and strategies, can also help break old deadlocks. For instance, open economy strategies have prompted progressive business groups in several Latin American countries to take a strong interest in measures to improve the quality of education. New leadership may be able to capture public imagination and encourage new alliances, in part by offering a credible and attractive vision. And modest changes in political institutions themselves B for example, in the laws regulating campaign finance, or in a major party ' s internal rules of governance B may change the incentives driving politicians and legislators.

None of these interventions is easy, and none is guaranteed to succeed in reviving reform momentum. Evolving public opinion, shifting alliances, leadership and vision, and sometimes modest changes in political institutions themselves are the stuff of A politics as usual @ . There are no obvious B or even plausible B alternatives.

Endnotes