SciELO - Scientific Electronic Library Online

 
vol.37 número112POLITICAL HABITS AND THE POLITICAL ECONOMY OF ECONOMIC REPRESSION AND REFORMLA ECONOMIA POLITICA DE LAS REFORMAS ECONOMICAS EN LA ARGENTINA índice de autoresíndice de materiabúsqueda de artículos
Home Pagelista alfabética de revistas  

Cuadernos de economía

versión On-line ISSN 0717-6821

Cuad. econ. v.37 n.112 Santiago dic. 2000

http://dx.doi.org/10.4067/S0717-68212000011200003 

ON THE CRISIS HYPOTHESIS OF ECONOMIC REFORM:
COLOMBIA 1989-91*

SEBASTIÁN EDWARDS**
ROBERTO STEINER***

ABSTRACT

On July 1999 Colombia entered into an extended facility agreement with the IMF. Clearly, problems currently affecting Colombia go well beyond economics, as political and social tensions have intensified. Many of the policies contemplated in the IMF program had already been addressed in the late 1980s and early 1990s. Indeed, during the final years of the Barco administration an economic reform program was initiated. This was intensified during the first two years of César Gaviria's presidency. The purpose of this paper is to analyze, from a political economy perspective, the launching of the Colombian market-oriented reforms in 1989-91. We are specially interested in investigating whether the so called "crisis hypothesis" applies to Colombia. This hypothesis is based on the observation made by a number of scholars, that a major economic crisis almost always precedes the launching of a reform effort. We argue that although Colombia was not facing a major economic crisis, it was facing a severe political and institutional breakdown. The profound attention devoted by key actors to political issues gave President Gaviria ample room to maneuver in the economic reform front. As a result, his first six months in office were truly remarkable in terms of economic reform. However, following the enactment of a new Constitution, interest groups negatively affected by the reforms were able to regroup, and started to actively lobby to stall the modernization process. As a result, the reformist effort slowed down significantly. Worse, many of the constitutional mandates placed Colombia on an unsustainable fiscal path.

Key words: Colombia, structural reforms, political economy,
JEL Classification: P16

1. INTRODUCTION

On July 1999, and for the first time ever, Colombia entered into an extended facility agreement with the IMF. This was the inevitable result of a sharp deterioration in most economic indicators, particularly in the fiscal front. In a matter of a few years, Colombia had gone from being a regional stellar performer to having lost its reputation for fiscal prudence. While in 1998 it was one of the few Latin American countries that could boast "investment grade", in 1999 it became evident that foreign financing would only be forthcoming if the Colombian authorities went back to the drawing board, under the tutelage of the IMF1.

By now it is clear that the problems currently affecting Colombia go well beyond economics. During the 1990s political and social tensions intensified, to the point that in 2000 the authorities asked the international community to come to the rescue, by providing resources to finance programs aimed at eradicating illicit drugs, and establishing a safety net for those most affected by civil strife. In 1999 around 5.000 civilians died as a result of an un-declared civil war between the armed forces, several leftist guerrilla organizations and right-wing paramilitaries. As many as 1 million Colombians have been displaced as a result of the internal conflict, and at any given time 3.000 persons are kidnapped either by the guerrilla's or by common criminals2.

Within the context of the IMF program, the authorities have agreed to implement changes in order to restore fiscal sustainability and to introduce an ambitious structural reform program. The latter include tax and pension reform -both at the national and state level-, as well as a redefinition of the system of inter- governmental transfers. There is even talk of labor reform and of a reform to the law that defines expenditure responsibilities among different levels of government.

Interestingly enough, many of the policies contemplated in the IMF program had already been addressed in the late 1980s and early 1990s. Indeed, during the final years of the Barco administration (1986-90) an economic reform program was initiated. This was intensified during 1990-91, the first two years of César Gaviria's presidency, when an ambitious reform program was launched with great enthusiasm. At the time of these reforms a number of observers were impressed by the scope of the program3. As it turned out, however, this reformist effort was incomplete and many initiatives never materialized. In fact, with the benefit of hindsight it is possible to argue that many of Colombia's current ills have their origins on this frustrated reform initiative. In particular, a number of provisions in the Constitution enacted in 1991 as part of the overall reform program, have generated serious fiscal imbalances and important rigidities and distortions4.

The purpose of this paper is to analyze, from a political economy perspective, the launching of the Colombian market-oriented reforms in 1989-91. We are particularly interested in three questions: First, why was this modernization effort initiated at that particular time in Colombia's history. Second, which were the main actors behind the launching of the reform process. Here we are particularly interested in understanding the role played by the so-called technopols. Third, what were the views of the more important interest groups regarding market-oriented reforms.

In this paper we are especially interested in analyzing whether Colombia's experience conforms to the accepted paradigm on the conditions that determine the launching of a major reform. More specifically, we investigate whether the so called "crisis hypothesis" of the political economy of reform applies to the case of Colombia. According to this hypothesis, a major economic crisis almost always precedes the launching of a market oriented reform effort.

In an early contribution Lal (1987, 274) argued that "the typical stabilization cum adjustment program is launched when the country is in a `crisis,' usually an incipient or actual balance of payments crisis that necessitates a reduction in the level of current expenditures." Bates and Krueger (1993) are, possibly, the strongest advocates of this "crisis" view of reform initiation. According to these authors, "[t]here is no recorded instance of the beginning of a reform program at a time when economic growth was satisfactory and when the price level and balance of payments situations were stable. Conditions of economic stagnation...or continued deterioration are evidently prerequisites for reform effort (page 454, emphasis added)." Likewise, Rodrik (1994, 63, emphasis added) explains that, "The reasons for the free trade bandwagon are more or less unique and derive from the intense, prolonged macroeconomic crisis that surrounded developing countries during the 1980s... which overshadowed the distributional considerations...." After examining thirteen episodes of economic reforms, Williamson and Haggard (1994, 564) reach a more cautious conclusion: "Crisis is clearly neither a necessary nor a sufficient condition to initiate reform. It has nevertheless often played a critical role in stimulating reform."

The paper has 5 sections, including this introduction. The second section is a brief discussion of the most important aspects of the literature on the political economy of reform, with special emphasis on the "crisis hypothesis." In the third we provide some background on the Colombian reforms. Emphasis is placed on the salient features of the political system and on the most important actors in the reformist effort. The fourth section deals with the initial social, political and economic conditions faced by the Gaviria administration. We argue that although at the end of the 1980s Colombia did not face severe macroeconomic imbalances, it was certainly facing severe institutional problems. The fifth section has to do with the launching of the first wave of reforms, in late 1990. We argue that this was a rather swift process, given that almost everybody was concerned with matters not related to the economic arena. In particular, a Constitutional Assembly was being conformed, and most of the attention was focused on how it would deal with topics such as drug trafficking, guerrilla's and political reform. In that respect, the 1991 Constitution is a breaking point in the reformist effort. Finally, section 6 presents a very brief conclusion.

2. THE CRISIS HYPOTHESIS OF THE POLITICAL ECONOMY OF REFORM

Recently there has been a proliferation of works dealing with the political economy of reform. Most of this literature has emphasized the role of interest groups, the media, initial economic and political conditions, and the role of the external environment. In this section we provide a very brief discussion of the most important aspects of this literature, and present a list of key hypotheses on the dynamics of the political economy of reform. The discussion in this section is deliberately brief, and concentrates on those issues that are particularly important for the analysis of the launching of Colombia's experiment with economic reform5.

Some Basic Issues

Political economy approaches to economic reform in general, and trade liberalization in particular, have traditionally focused on distributive conflicts. The basic framework is based on some variant of the Heckscher-Ohlin theory of international trade, and considers a finite number of actors in the reform process6. Some groups will be hurt by the reform and will oppose it, while those that benefit will support it. This approach has been implicitly followed by Rodrik (1994), who considers three groups of actors: (i) import substituting industrialists; (ii) holders of import licenses; and (iii) users of imports, including producers that rely on imported inputs. Depending on the underlying model of an economy (i.e., whether factors of production are industry-specific or not), it is possible to add to this list any number of groups with special characteristics. In this particular set-up the political support for the reform effort will be proportional to the difference between redistributed income and net efficiency gains from the reforms; what Rodrik (1994) calls the "political cost-benefit ratio."

Reforms are seldom restricted to one area of the economy. In order to have a better understanding of the overall process, and in particular of the interaction between different interest groups, a broader constellation of players and sectors should be considered. For example, if financial sector reform is being considered, the role of banks, speculators and debtors should be incorporated into the analysis. Moreover, as long as privatization is a key component of the reform agenda, the role of state-owned enterprises (SOEs) in the reform process has to be considered explicitly. Commenting on the political economy of privatization, the World Bank has argued that the reform of SOEs can cost a government its support base, since reform generally involves eliminating jobs and long-established subsidies. Also, to the extent that a reform program includes the reform of public sector institutions, the role of public sector unions and the bureaucracy should be taken into account in the analysis.

Bates and Krueger (1993, 456) argue that one possible explanation for the failure of interest groups to prevent or shape economic reforms is that, "in the context of comprehensive economic policy reform, it is difficult for particular groups to calculate where their interests lie." It is possible to argue that the various components of policy packages can act as compensation mechanisms that can help reduce opposition and raise support for a program that includes trade liberalization.

In recent policy discussions it is now customary to distinguish between "first" and "second" generation reforms (Naim, 1995). First generation policy reforms represent a change in economic policies that alters some basic aspects of the economic structure of the country in question. In addition to trade liberalization, these policies include fiscal reforms and adjustment, the elimination of exchange controls, financial liberalization, the implementation of minimal social safety nets, deregulation and some privatization. In contrast, second generation, or institutional reforms aim to drastically change the institutions of the state. These reforms are politically and technically more difficult to implement, as they entail changing the functioning of fundamental institutions.

In sum, recent analyses of the political economy of economic reforms have gone beyond pure economic issues and have emphasized five interrelated aspects: (i) distributive conflict among different groups; (ii) the country's political organization and structure, including the system of checks and balances and the number and nature of political parties; (iii) the role of the bureaucracy; (iv) the role of ideas and of professional and intellectual groups; (v) external actors, including the international multilateral institutions. This broader perspective recognizes the relevance of a rich list of actors. In this framework, coalitions play an important function, as do politically-motivated compensation schemes.

Crisis and the Launching of Reform

Each episode of economic reform is historically unique. Nonetheless, recent work on the political economy of reform have been able to detect some regularities followed by most reform episodes (see Table 1). As was explicitly shown in the introduction to this paper, several authors have argued that a major economic crisis almost always precedes the launching of a reform effort.


It should be noted, however, that it is not only evident that crisis do not always lead to reform. In some cases a crisis may actually result in the implementation of more distortive and command-type policies. Rajapatirana et al. (1997), for instance, have argued that this has been the case in many Latin American episodes. Indeed, after analyzing the evolution of trade policy in several countries, they have concluded that on several occasions severe macroeconomic imbalances have resulted, if anything, in the "tightening" of trade policies. Furthermore, though countries may not be experiencing a major crisis -in terms of hyperinflation or balance of payments problems- a severely distorted economic environment can certainly propitiate a reformist effort. In fact, it might very well be the case that a country has averted balance of payments problems precisely because it had adopted a highly distortive and protective tariff structure. Finally, not all crises are necessarily of an economic nature. Olson (1982) argues that a military defeat or a natural disaster can disrupt existing political coalitions, thereby opening room for the (reform-minded) initiatives of new actors.

The Colombian reformist experience of the early 1990s is particularly interesting from a political economy perspective precisely because, as we will show in section 4, while there were no major macroeconomic problems at the outset, the economy was highly distorted and, more importantly, there was a severe gobernability crisis, with rampant violence resulting from confrontation with drug-traffickers and left-wing guerrilla organizations.

Following Nelson (1994), it should be pointed out that disentangling the type of crisis -for example, institutional as opposed to macroeconomic- might not be very important in order to understand the launching of a reform process, as reforms can be propitiated from many different angles. However, that disentanglement might very well be critical in relation to the sustainability of the structural reform process itself. Put differently, it is one thing to advance structural reforms in the context of severe macroeconomic imbalances, quite a different story to do it as a reaction, at least in part, to institutional problems not directly linked to economic issues. Among other things, the support that a reform program can elicit when it is launched from a reasonably acceptable economic situation can be highly unreliable.

While Edwards (1996) and Edwards and Steiner (1999) elaborate on several issues regarding the political economy of structural reforms, in this paper we concentrate on "initial conditions" and on the "crisis hypothesis". In essence, we will challenge Williamson's (1994, 478) statement according to which "a number of countries were indeed stimulated to reform by crises … But there were also cases, such as Colombia and Portugal, where reform occurred without the stimulus of a crisis."

According to this "crisis hypothesis," in the midst of an economic crisis, highly trained social scientists are called by politicians and asked to help forge a way out of the crisis. All of a sudden, the incoming technopols' ideas -usually based on the Anglo-Saxon economic tradition and consistent with the views of the multilateral institutions- become highly influential7. Certainly, ideas are not only important in formulating the reform plan, but also in implementing the actual liberalization program. At the implementation stage the technopols usually find out that the realities of politics conflict with the simple world of economics. Their ability to understand political trade-off's and to design sellable strategies may determine the success of the reform effort.

Although the actual way in which reforms are implemented varies across countries, the early phases of the process are often concentrated on basic economic issues, and in particular on macroeconomic stabilization and trade liberalization. The coordination of macroeconomic and trade reform policies becomes crucial at this time. The political effect of these early reforms, particularly the support they will elicit from the population, will depend on a number of factors, including: (i) whether the government is able to put in place an effective public relations effort; (ii) the performance of the economy; (iii) the government's capacity to put in place compensation schemes that will defuse opposition to the reforms; and (iv) the government's ability to forge broad political coalitions.

Successful reformist governments are able to form broad coalitions and, more importantly, to compensate and neutralize (potential) losers. As Haggard and Webb (1994 a, b) have argued, most successful compensation schemes have been rather complex, and have not been based on simple monetary transfers to those that are directly hurt by the reforms. Actually, there are several distinct ways in which a reformist government can compensate potential losers and/or opposers. The first one, called direct compensation, includes the traditional mechanisms used by governments to `buy' the support of certain groups directly affected by the reform process. An example of this type of mechanism would be the distribution of shares in privatized companies to workers in those companies. Indirect mechanisms rely on making policy adjustments in areas unrelated to the reforms proper, in order to compensate (potential) opponents; cross compensation mechanisms, on the other hand, try to `buy' the support of groups not directly affected by the reforms. Exclusionary compensation mechanisms try to deflect opposition by certain groups, by not including them among those affected by particular reform measures. The final category, political compensation, includes the appointment of reform skeptics into key government positions.

Exporters are usually among the early supporters of reform-oriented governments. They benefit directly from the reduction of import tariffs and from the exchange rate depreciation that often takes place during the early stages of an stabilization program. Producers of import-competing goods are usually among the opponents of trade reform initiatives, but are often at least partially compensated by the real depreciation of the currency. Losers of trade liberalization can also be compensated through other schemes. In particular, if the reform process is seen as a package, some import-substituting groups may support the reforms if they perceive that they will directly benefit from privatizations or financial liberalization, for example.

Support from the multilateral institutions -either in the form of technical assistance or through the provision of funds- may help the reform effort, once it has been launched. However, there is significant evidence that the multilaterals have not usually played a fundamental role in the initiation of the reforms (see Edwards, 1997). According to Lal (1997, 150), in order to encourage the reform government to undertake its tasks, "sweeteners which ease its fiscal problems, in the form of soft loans or grants from multilateral and bilateral foreign governments, may be desirable. Beyond that the role of foreign assistance seems limited."

How well the reform effort fares from a political economy perspective will also depend on the political structure of the country. For example, countries with a two party system and a low(er) degree of political polarization may find it easier to forge a national project and implement reforms with a long pay-off period than those countries with a highly fragmented political system. More generally, countries with a high degree of political conflict may tend to get easily bogged down in political struggles.

With time, and as the original crisis subsides, reform skeptics and opponents are able to regroup and to challenge the reform itself8. Their efforts to slow down the modernization process -and even to reverse it- can be successful if the reforms have failed to generate significant improvements in economic conditions, including higher real wages and lower unemployment. While in some cases nostalgia and the populist temptation are strong enough to bring the reforms to a standstill, in others the modernization forces are able to continue moving forward until the reform process is consolidated. The way in which a reformist government reacts to the populist challenge is also an important determinant of the fate of the modernization effort. For example, in order to regain some public support the authorities may reduce the pace of key reforms, or -as is becoming increasingly common- they may relax the public sector budget constraint to face a political challenge.

3. COLOMBIA: POLITICAL BACKGROUND AND MAIN ACTORS

In this section we analyze the political and economic conditions faced by Colombia before the market oriented reforms of 1989-90 were launched. This analysis provides useful background information for understanding why the Colombian reformist effort took the form it did. The section is organized in three parts. We first deal with the nature of the party system, highlighting the dominance of an entrenched bi-partisan coalition. Secondly, we discuss the implications of the polity in terms of economic and institutional stability. Finally, we make particular reference to some key actors relevant in the reformist effort. In the next section we will discuss the economic, social and political conditions prevalent at the time of the introduction of the structural reform program.

Thirty Years of Power-Sharing9

The party system has changed less in Colombia than any other country in the Western Hemisphere, probably with the exception of Mexico. According to Dix (1990, p. 100), "what does distinguish the Colombian case is the failure of the country's political institutions, and notably its party system, to manifest changes comparable in magnitude to the new levels of social mobilization." On average, during 1945-86 the two traditional parties commanded over 97% of the vote for the House of Representatives. Only once (in 1970) did their combined total at the polls fall below 90%. Between 1931 and 1982 barely 12% of municipalities switched party allegiance. Almost without exception, every time dissident factions appeared, they invariably went back to their original partisan allegiance.

The non-emergence of a third party is surprising in light of the fact that, since the late 1950s, the two traditional parties have converged both in terms of policy and of ideology. Furthermore, Colombia's poor distribution of income and wealth has made it an ideal place for the emergence of a class-based party. However, the workforce has not been heavily employed in sectors that are prone to produce a militant, class-oriented activism. Instead, coffee, the country's main product, is grown on thousands of small, family-operated parcels. This has fostered conservative, individualist values rather than radical or collectivist ones (Dix, 1990). In addition, coffee has been entirely owned by Colombian nationals, as have other key industries. This might explain the lack of nationalistic attitudes, a frequent breeding ground for popular movements. Of course, the exclusionist nature of the bi-partisan coalition might in part explain the emergence and maintenance in time of a very active left-wing guerrilla organization.

Since 1958 Colombia can be characterized as a case of a consociational democracy, unique in that its subcultures do not originate in religious, ethnic, or class differences (Dix, 1980)10. Regarding the two traditional parties, their exclusivist nature and their familial and communal roots mark them as the functional equivalent of subcultures. For the most part, the two traditional parties have been composed of prominent individuals, bound together through clientelistic relationships, with weak organizations and barely any programmatic content. They are nonetheless multi-class, as elites have sought popular support in their partisan competition for the perquisites of power.

The two traditional parties' hold on power was enhanced in the late 1950s through a political coalition (National Front or Frente Nacional). After several years of party violence, the leaders of the two parties reached an arrangement by which both would share power for sixteen years. Several factors led to this agreement. First, the usual conflict between the parties was increasingly superseded by popular resentment toward them. Second, the military's repressive actions and political ambitions became more uncomfortable than a coalition with a partisan foe. Third, the Liberals came to accept Colombia's status as an officially Catholic country. Finally, both parties agreed that the military would not be persecuted and that the Minister of Defense would come from its ranks.

The National Front consensus was based on three main provisions. Perhaps the most prominent was that the presidency would alternate between parties for the duration of the coalition. Throughout the political system, seats in Congress, departmental assemblies and municipal councils, all cabinet and government offices, as well as positions in the Supreme Court, would be divided evenly between the two major parties to the exclusion of all others. To avoid partisan stalemates, non-procedural decisions in all elective bodies required a two-thirds majority.

Though it can be criticized for having impeded social change, the National Front must be credited with greatly reducing inter-party violence and returning the military to the barracks11. The two parties stopped quarreling for the control of the state, while reaping the gains of economic growth. Resulting policies fostered stable growth and moderate inflation without openly discriminating against any activity.

The National Front formally ended in 1974, but several of its provisions were extended for four years12. Informally, the coalition lasted much longer. Every president until 1990 came from within its ranks, and Congress continued to be dominated by the two traditional parties. A referendum held in 1990 mandated the convening of a popularly elected Constituent Assembly that drafted a new constitution (more on this below). Its provisions were designed to facilitate the creation and consolidation of new political movements, empower citizens, and advance political and fiscal decentralization. Though many analysts have been skeptical of these provisions' practical implications, the power-sharing schemes of the National Front did, for all practical purposes, end in 1991.

Some Economic and Institutional Implications of the Polity

The characterization of Colombia as a consociational democracy, in which two parties have explicitly or implicitly shared power, largely explains the economic policies implemented between the 1960s and the 1980s. According to Haggard (1994), Latin American countries with high inflation histories have been those in which the "urban popular sector" and labor groups have been incorporated into populist parties, in the context of relatively polarized party systems. In sharp contrast, the political difficulties of implementing macroeconomic adjustment programs have been less severe where decision-making has been centralized and less subjected to rent-seeking pressures.

Two features distinguish Colombia from the typical Latin American country, as characterized by Haggard. The most influential business group produces an export (coffee) not linked to import substitution, and export promotion has been the cornerstone of policy since 1967. Post-1967 Colombia seems to fit in Haggard's assessment of Asia in the sense that in that region it has been important "to maintain realistic exchange rates and to shift toward the promotion of manufactured exports" (p. 241).

According to Urrutia (1991), in Colombia there has generally been no "economic populism," at least not in the sense of Dornbush and Edwards (1991). Macroeconomic policy has generally been conducted with the purpose of providing some minimum stability deemed necessary to promote growth, rather than with redistribution purposes. This has been made possible not by "populism", but by "clientelism", which has given politicians an important say in the distribution of the national budget, while at the same time allowing them to reap the benefits of stability, produced by a seldom politicized macroeconomic decision-making process.

As Archer and Soberg Shugart (1997) and Soberg Shugart and Nielson (1997) show, Colombia's electoral rules (in which members of Congress are elected by proportional representation rules applied to each district on the basis of factional rather than party lists) provide incentives for politicians to pursue personal, service-related votes instead of party-oriented votes13. Voters are tied to individual legislators and their vote depends heavily on the elected official's delivery of government services. In addition, parties have almost no control over the composition of their delegations in Congress. Almost four decades ago, Nelson et al. (1971) argued that political practices were supplementing the inter-party political and ideological debate with the politics of pressure groups.

(i) Relative economic stability

Within the Latin American context, and notwithstanding the dismal performance of the last five years, Colombia has been characterized as the least volatile economy. Table 2 shows information on inflation and growth for the region's largest economies. In all sub-periods, Colombia had the third lowest inflation rate. In terms of volatility, except during 1960-69, Colombia additionally experienced the most stable rate of inflation. A similar pattern emerges in the country's growth rate. The table indicates that Colombia's relative stability occurred in the context of macroeconomic results that were not particularly impressive.


Economic volatility depends on several factors, particularly external conditions (i.e., terms of trade and capital flows) and economic policy. Table 3 shows that with regard to the terms of trade, Colombia's "external environment" has not been particularly stable in comparison to other regional economies. Instead, during 1970-92 Colombia had by far the most stable economic policy in the region (IADB, 1995).


(ii) Institutional stability

Table 4 shows the phases of central banking in Colombia. The Banco de la República was established as a private and autonomous entity in 1923, and it remained so until 1963. Up to 1931 Colombia adhered to the gold standard, thereafter maintaining a fixed (and stable) exchange rate until 1949. In 1951 the bank's objectives were expanded to include not only price stability but also "economic development." In the early 1960s the bank was nationalized and in 1963 the Monetary Board was created. Composed solely of members of the government and an appointed governor, the Board was given control of monetary, credit and exchange rate policies.


The 1991 Constitution completely overhauled the central banking regime. The Banco de la República was organized as an autonomous state institution, independent from the government, with the sole objective of achieving price stability.

In 1967 Colombia adopted a "crawling peg" exchange rate system. Whatever is said regarding stability in Colombia's key economic variables has to take into account the very prominent fact that since 1967, and at least until 1991, it maintained basically the same exchange rate regime14. The stability of the exchange rate regime has coincided with remarkable institutional stability. Table 5 shows that since 1960 the central bank has had only six governors, with an average tenure of 6.5 years (and counting), well above the regional average suggested in Table 6, taken from Cukierman (1992). If the tenure of the central bank governor has been noteworthy, it pales in comparison to that of the General Manager of the National Federation of Coffee Growers (for Federación Nacional de Cafeteros). Since 1937, the Federation has had only three general managers.



TECHNOCRATS, POLITICIANS AND OTHER KEY ACTORS

As in most other countries, Colombia's socio-economic landscape is extremely complex, full of land mines that any reformist government has to face with extreme care. Table 7 contains a systematic discussion of the most important internal actors, including a brief comment on their general position regarding economic modernization, as well as a detailed discussion on the way they viewed specific aspects of the reform agenda.

Two comments regarding Table 7 are in order. The first is related to the business sector pressure groups, or gremios. For a long time there has been the idea that these groups have had a significant degree of influence (Urrutia, 1983). They have traditionally exercised it in a number of ways: they lobby in a way similar to their US counterparts, they finance political campaigns for members of Congress, and they finance think tanks that develop the intellectual rationale for their positions. Although the gremio formed by coffee producers (FEDECAFE) is the better known, the long list of prominent gremios in Table 7 illustrates that what in analytical models is generically referred to as "owners of capital" is, in most instances, a highly heterogeneous group. Many of these gremios had antagonistic views regarding critical aspects of the reform program.

TABLE 7
MAIN ACTORS IN THE POLITICAL ECONOMY OF COLOMBIA'S REFORMS: DEGREE OF INFLUENCE AND VIEW REGARDING
CORE ASPECTS OF THE REFORMS
TABLE 7 (Cont.)
TABLE 7 (Cont.)
TABLE 7 (Cont.)

The second important aspect captured by Table 7 refers to labor. Broadly speaking, Colombia's labor movement does not appear to be very important. While in 1965 close to 16% of the labor force belonged to a labor organization, that number dropped to 9.3% in 1984 and to only 7.8% in 1990. Affiliation numbers are small, and have barley increased (Table 8). In 1990 the economic activity with the highest percentage of workers belonging to a labor union was "utilities" (generally publicly owned), with an affiliation rate of 42%. Affiliation rates were as low as 1.5% in agriculture and 8.2% in manufacturing.


The importance of the labor movement is thus restricted to the public sector, and even there percentages have been declining (Table 9). Traditionally, Colombia has been characterized by the existence of a dualistic labor movement. While, in general, unions have been weak, public sector unions -and especially teachers', health workers, and state owned enterprises unions- have been very powerful. In this regard, Wiesner (1997) argues that the rent-seeking practices of organized labor in key public sector services have played a crucial role in retarding much needed structural reforms.


As a result of the political and institutional crisis that had reached a critical level during the 1989-90 electoral period (more on this below), members of Congress were decisively in the defensive, as there was an increasing feeling among the public that they were unable to pass meaningful legislation. Voters rejection of the political status quo peaked in March 1990, when in an almost spontaneous way they overwhelmingly voted for convening a constitutional assembly (see section 4). Drug traffickers, which during the 1970s had become increasingly influential in all spheres of Colombia's public life, had also taken a defensive stance — albeit a very aggressive and violent one —, as the Barco administration had stepped up the war against them15. And finally, the guerrilla movement, which for years had waged a war of attrition with the government, had accepted to enter into (what appeared to be) serious negotiations.

In Colombia, as in most developing countries during the late 1980s and early 1990s, policy discussions were also affected by the views of the multilateral institutions. Possibly the most important "external" actor was the World Bank, which had a fruitful long-term relationship with Colombia. The IMF, on the other hand, had significantly less influence, as Colombia had tried (successfully) since the late 1960s to avoid entering into IMF-sponsored programs16.

In terms of actors, an important question refers to the role played by "technocrats" in the launching of the reforms. We believe that this is such a key issue that in this section we discuss it in some detail. On a general basis, several authors have emphasized the long-lasting role of "technocrats" in Colombia's decision-making process. Edwards (1995b) has highlighted their role in explaining Colombia's relative economic stability: "Fedesarrollo (a private research institution) has been the intellectual breeding ground of an amazingly large number of those in charge of economic policy. It could be argued that Fedesarrollo's bipartisan and non-ideological positions are a good reflection of Colombia's implicit pact according to which Liberals and Conservatives share power." Of its nine Directors, five have been Liberal, three Conservative, and one independent (Table 10). In all instances but one, the Director has been affiliated with a different party than that of the government in power. The same individuals act sometimes as government officials and at other times as independent analysts. It can be argued that the institution is a facilitator for achieving consensus, not because of its intellectual power, but rather because it credibly internalizes policy options that are politically viable17.


The specific group of technocrats that was active during the Gaviria administration deserves a special mention. Although the notion of opening up the economy had been accepted by several policy makers by early 1990, most influential figures were skeptical about the bulk of the other reforms. The group of economists that designed the program was, to some extent, outside of Colombia's economic mainstream.

There is little doubt that from the beginning the reform program was as much candidate Gaviria's, as that of his closest associates, including Rudolf Hommes and Armando Montenegro, who would later become Finance Minister and Director of the National Planning Department, respectively18. César Gaviria was trained as an economist at Universidad de los Andes, Bogotá's premiere private university. According to those that knew him during his years as a student, Gaviria developed from very early on a critical view regarding CEPAL's import substitution development strategies, then in vogue in most of Latin America19.

Throughout most of 1989 and early 1990 a group of economists, mostly associated with the modern wing of the Liberal Party -a wing increasingly associated with Gaviria's name-, began meeting in a Bogotá restaurant, the Club Suizo, to discuss the country's economic future. The de-facto coordinator of the group -which quickly became known as the Club Suizo group- was Rudolf Hommes, who had advised slained leader Galán20 and later became Gaviria's main economic advisor.

What began as informal meetings, derived into a formal proposition for a new economic policy for the 1990s. This program was published in March of 1990 -two full months before the presidential elections- in the monthly magazine Estrategia Económica y Financiera. The article, which was written by Hommes (1990) and synthesized the views of the Club Suizo group, was titled "An Economic Proposal for the Nineties" (Una Propuesta Económica para los Noventa). The importance of this document is that it is one of the very few that allows us to know what Gaviria's economic team had in mind in terms of economic policy, before the elections. In fact, many of the most important elements of this document were later reproduced in the government's official National Development Plan (La Revolución Pacífica).

An interesting question -and one that has some bearing on the interpretation of the political economy of the Colombian reforms- is who was an actual member of this group. Cepeda (1994), for example, lists 11 people, most of which had been associated with the economics department of Universidad de los Andes. The "Proposal" article, on the other hand, in its opening footnote lists 26 names. Many of those in these lists went on to occupy very senior positions -including at the cabinet and sub cabinet levels- in the Gaviria administration. Two interesting features of these lists are worthwhile noting: first, the vast majority of those involved in the Club Suizo deliberations had graduate degrees from top US or European universities. In addition, almost none of them had participated actively in politics. In that regard, this group constitutes the core of the Colombian technopols -the technocrats turned reformers and politicians- which according to Williamson and Haggard (1994) and Domínguez (1997), among others, have played a key role in the launching of every one of the Latin American reforms. A second characteristic that deserves some attention is that most of the members of Colombia's economic establishment are absent from it. Although the members of the group were far from being unknowns, it may be argued that they were (at least partially) outsiders, with what appeared to be new, and even puzzling, ideas. Another interesting feature of the Club Suizo group is that many of its members had been associated with the Banco de la República and, as is usually the case with central bankers, had a relatively low profile.

Although candidate Gaviria participated in only a few of the meetings, this manifesto clearly represented the core of his campaign's economic ideas. He was, of course, aware of the Club Suizo group activities and received periodical briefings on the different issues being discussed. As the campaign progressed, however, Gaviria started to develop a number of new ideas of his own, many of which had not been discussed by the Club Suizo group - or, if discussed, had been discarded as ineffective. Perhaps the most important of these ideas referred to the abolition of the Capital Controls provision of Decree 444, that for more than twenty years had governed Colombia's exchange rate policy.

4. THE CRISIS HYPOTHESIS: POLITICAL AND INSTITUTIONAL PRE-CONDITIONS FOR REFORM

During the turbulent 1980s, Colombia was, together with Chile, the best performer among Latin American economies (Table 11). While during this "lost decade" average GDP growth was negative in Argentina and Venezuela, and barely positive in Peru and Uruguay, it reached 3.4% in Colombia. The same can be said in terms of per capita GDP. Its average rate of growth was negative in most countries, except Brazil (0.95%), Colombia (1.29%) and Chile (2.76%). Other macroeconomic indicators also made Colombia look well by regional standards. For example, it was the only major regional economy not to restructure its foreign debt, as a result of the debt crisis that ensued following México's 1982 moratorium. On the other hand, inflation never got out of hand.


Social indicators also fared reasonably well. Between 1978 and 1991 poverty fell, both in urban as well as in rural areas (Table 12). According to CEPAL, during the 1980s Colombia was one of the few countries where income distribution actually improved, albeit from a very uneven starting point (Table 13).



Evidently, not all was well in the economic front. Following the problems associated with the debt crisis, towards the end of the 1980s it became evident that as a consequence of a period of "stop and go" policies on several fronts, the Colombian economy had come to exhibit numerous distortions.

As a result of two external shocks -the cut of external financing and the worsening in the terms of trade- Colombia was forced to implement a major adjustment in the first half of the 1980s. This was based on the implementation of a real exchange rate depreciation -through an acceleration of the pace of mini devaluations -, a timid fiscal correction, and a major hike in the degree of protection. Average (nominal) import tariffs went from 26 % in 1980 to 57% in 1984. More important yet, import licenses increased very significantly during this period: while in 1980, 31% of imports required a license, by 1984 99% of imports were either prohibited or required prior licenses! (Table 14).


By 1984 many observers believed that the high degree of protection was suffocating the economy. In essence, several analyst argued that Colombia's acceptable rate of growth was entirely explained by factor accumulation, with total factor productivity steady, if not declining. Certainly, from a regional standpoint, in 1985 Colombia was probably the least open of all Latin American economies (Table 15).


The World Bank initiated discussions with the authorities on the need to implement a trade reform program and in May 1985, after extensive discussions with the authorities, the Bank approved a Trade Policy and Export Diversification Loan to assist Colombia in this task21. An import rationalization process was begun in 1985, with the purpose not so much of exposing domestic production to foreign competition, but of expediting administrative procedures. The following measures are worth noting: (i) The percentage of the tariff universe that could freely be imported increased from 0.5% in 1984 to 36% at the end of 1986. The items included in the prior license regime and in the prohibited import list decreased from 63% and 83% to 1% and 16.5%, respectively. (ii) In 1986 the percentage of rejected applications had decreased to 39%. (iii) The average tariff, that amounted to 42% in 1985, was reduced to 31.2% in 1986.

Distortions were prevalent not only in regard to the trade regime. As has been the case in a number of Latin American countries, Colombia's traditional labor legislation was extremely rigid, imposing very high costs on formal sector firms. With the purpose of protecting workers rights, legislation had severely distorted the labor market, yielding results that were contrary to those that had been expected. Among others, they produced significant employment instability, as it made lay-offs extremely expensive, especially in the case of employees who had been with the same firm for over 10 years. In addition, legislation had become a barrier to the generation of employment in the modern sector as the double retroactive character of severance payments increased labor costs and made them highly unpredictable.

Prior to 1991, a fixed-term labor contract could not last less than one year, except under very special conditions. As a result, firms would use, and sometimes even constitute, temporal employment services under highly unfavorable conditions: (i) wages were lower than those earned by formally employed workers with similar qualifications; and (ii) there were no benefits.

In addition, the law provided that if an employer was laid off without "proper cause" (term which the law did not define), employees would have to pay a penalty equivalent to the wages that were due until the expiration of the contract, if it was a fixed-term arrangement. If it was not, in addition to all legal benefits, the employer had to pay extra wages as a penalty. In the latter case, the employee had to be re-hired (acción de reintegro) and was entitled to receive payment for the period he had been idle (pago de "brazos caídos"). In addition, the employer was entitled to a pension (pensión sanción).

The most controversial aspect regarding labor legislation was the double retroactive character of severance payments. The law stipulated that every year workers were entitled to a bonus equivalent to a month's wage, plus 12% interest. This liability was due upon retirement. Workers could make withdrawals in order to purchase (or improve) housing. Even if the employee had made one or more withdrawals, thereby exhausting the funds he/she was entitled to receive at the time, the employers liability would not be fully cancelled. At the time of the final liquidation -and taking into account the nominal wage increases that occurred between the time of the withdrawals and the time of final liquidation-, only the nominal value of withdrawals was taken into account. Because of this, firms liabilities increased every year, in an uncertain manner, since they were affected by how long employees had been at the firm, by the amount of partial withdrawals and by the return the firm could obtain by investing these funds.

The financial sector was also highly distorted. An important number of banks was state controlled, and ownership of private banks was highly concentrated; interest rates were subject to ceilings; forced credit allocation was very important; securities markets were discouraged; and supervision was lax. Colombia's experience illustrates how even with fairly stable rates of inflation and prudent fiscal policies, private securities markets can deteriorate. By the end of the 1980s less than 100 companies were listed on the stock exchanges as opposed to 400 in the 1960s, and privately issued securities (stocks and bonds) amounted to 2% of total financial system liabilities vs. 20% in 1965.

Not to mention the capital account. Since 1967 Colombia had an extensive system of massive exchange controls. Originally this system had been devised as a way of allowing the central bank to carry out a nominal exchange rate policy based on frequent mini devaluations. The purpose of this exchange rate policy was to avoid real exchange rate overvaluation and to maintain export competitiveness. With time the mini devaluations policy became the staple of Colombia's policy making. Ironically, perhaps, economic agents came to believe that it was a fundamental determinant of Colombia's (relative) economic stability.

Direct foreign investment was also subject to a surrealistic array of controls. The legislation governing DFI had its origins in the Andean Pact infamous Article 24 which, for all practical purposes, was aimed at cutting the Andean nations off the world economy. Likewise, the tax system was seriously distorted. Inflation eroded tax assessments, distributed earnings were subject to double-taxation, evasion was rampant and a number of sectors had been able to obtain significant exemptions.

Although the above listing of distortions amply justify the introduction of corrective measures, it would be an exaggeration to say that they conformed a "critical" situation. At best, one could describe the Colombian economy of the late 1980s as one with an acceptable macroeconomic performance, in the context of several "microeconomic" distortions.

At the onset of the 1990s, and unfortunately still today, what certainly did set Colombia apart from any other Latin American country were severe institutional problems having to do with an un-declared civil war with leftist guerrilla's, attempts by the drug trade to overpower democratic institutions, and levels of everyday violence that made Colombia one of the most dangerous country in the world.

The reformist effort led by César Gaviria in 1990-91 took place in the context of a country that, though economically sound, was in the verge of falling apart because of all the social unrest. On August 15, 1989 Senator Luis Carlos Galán -founder and leader of the dissident Nuevo Liberalismo faction and undisputable front-runner in the liberal party's primary campaign- was assassinated by gunmen hired by the Medellín drug cartel. The months that followed Galán's assassination were particularly bloody and reflected the deep institutional, social and political crisis that the country faced.

In the last four months of 1989 alone, close to 90 bombs exploded in the major cities. The headquarters of a leading newspaper were destroyed, while banks and corporate buildings were subject to several terrorist acts. A powerful bomb all but destroyed the headquarters of the security forces, killing 63 people and wounding more than 600 -though missing its target, the head of the security forces, a controversial maverick in the fight against Pablo Escobar and other leaders of the drug mafia. Maybe the most gruesome element of this relentless terrorist rampage was the planting of a bomb on a commercial jetliner flying from Bogotá to Cali, attack in which over 100 innocent persons perished22.

To be sure, the degree to which violence had escalated went well beyond these spectacular episodes, duly captured by the international press. Only in El Salvador and Guatemala, which were in the midst of outright civil war, was the rate of homicide higher than in Colombia (Figure 1). Contrary to what many analysts now believe, Colombia had not always been a particularly violent country. Actually, there is a strong positive correlation between the "guesstimates" of the value of the drug trade and the evolution of homicides (Figure 2). Not surprisingly, the presence of drugs not only fueled everyday violence. It also financed the dramatic expansion of guerrilla organization's (Figure 3). The causual evidence provided by this graphs has recently received solid econometric support (Sánchez and Núñez, 2000).




More depressing evidence on how critical the situation was in terms of social disruption is provided in Figures 4 and 5, which make reference, respectively, to the total number of massacres, guerrilla attacks and kidnappings. Sadly, of course, is the fact that most of these indicators, which had reached an alarming level in 1990, have actually deteriorated since then. In fact, only the rate of homicides has declined, and experts are still searching for an explanation to this positive development.



When the 1990 presidential elections were finally over on May 27th, three presidential candidates had been assassinated - Galán, Bernardo Jaramillo from the left wing UP, and Carlos Pizarro from the former guerrilla movement M-19. Additionally, Ernesto Samper, one of the leading Liberal Party candidates, had been seriously wounded in an attack aimed at communist leader José Antequera.

Galán had been a relentless proponent of the modernization of Colombia's political system. Starting in 1987, and after returning to Colombia from a year in the U.K., his views on the economy began to change. Increasingly, he called for modernizing the economy and opening up to international competition (Vargas, 1993). On October 20th 1989 Galán's elder son surprised the Nuevo Liberalismo leadership when at his father's funeral he publicly, and emotionally, asked César Gaviria, Galan's campaign manager, to take over as the movement's leader23. A few days later Gaviria officially joined the primaries as the front-runner. In March 1990 he defeated his two liberal rivals - including Ernesto Samper, who would eventually succeed him as president in 1994. When César Gaviria, the 43 years old economist, became Colombia's constitutional president on August 7th 1990, Colombia was facing one of its worst social, political and institutional crises. The economy may have been doing reasonably well, but the sense of institutional disintegration was evident to perceptive analysts and, clearly, to almost everyone living in the country.

5. THE 1991 CONSTITUTION: A BREAKING POINT IN THE REFORMIST EFFORT

In the preceding section we argued that, although in the late 1980s Colombia was not strictly facing an economic breakdown, it was going through major political and social turmoil. In that regard, the Colombian experience conforms to a broad -and, we argue, correct- interpretation of the "crisis hypothesis." Moreover, the scope of the crisis, and in particular its political and social dimension -including the increasingly important issue of drug trafficking- determined the strategy followed by the Gaviria administration. In particular, it had two major effects: First, it allowed the authorities to go well beyond the economic sphere, and to launch a major political reform, with a new Constitution as its centerpiece. And second, since much of the media's attention was on these political reforms, it allowed the government to launch economic reforms that, at least on paper, were deeper than what it would have been otherwise possible.

In this section we analyze in some detail the political economy process surrounding the enactment of the Colombian constitution of 1991. In particular, we argue that during the fourth quarter of 1990 the Gaviria administration wisely seized the opportunity granted by the political and institutional upheaval, in order to advance its economic reform agenda. While the general public was in favor of drastic changes that would put an end to the drug baron's blood bath and to decades of guerilla warfare, Congress was a highly discredited institution. Without much opposition -and with key actors concerned with politics, drugs and guerrillas, rather than with economics- the Gaviria administration rapidly moved forward key economic reforms during its first few months in office. Barely a year down the road, when popular and political support became critical in order to consolidate and advance the reformist effort, the whole process came to a standstill. In that regard, it is possible to argue that the key breaking point is the enactment of a new Constitution in 1991.

Probably the most salient feature of Colombia's economic reform program is that it took place in the context of a much broader reform effort, one that included not only the economy, but also attempts at opening the political system, making peace with the guerrilla's and bringing drug lords to justice (Table 16)24. President Gaviria was well aware that while there were problems in the economic front which justified the introduction of a structural reform program, what was really critical and allowed for no delays was the unstable political situation, which was deteriorating at a very fast pace.

TABLE 16
THE COLOMBIAN REFORMS: AN ITINERARY
TABLE 16 (Cont.)

THE CONSTITUTIONAL ASSEMBLY

In 1990 Colombia had one of the oldest constitutions in the world. Originally enacted in 1886, it had gone through a handful of revisions, prominent among them those of 1936 and 1968. The three presidents that preceded Gaviria in office had attempted to reform the Constitution. Lack of success generally was associated with procedural problems. The last (unsuccessful) reformist attempt had been led by President Barco. After clearing all instances in Congress, the reform collapsed when legislators that were in the mafia's payroll demanded the introduction of an article banning the extradition of Colombian nationals. Mr. Barco was left with no alternative other than retrieving his reform from Congress.

For various reasons, in 1990 the stage was set not only for a full-fledged constitutional reform, but actually for the convening of a constitutional assembly that would draft a new constitution. The general public was fed up with an inept Congress, was terrorized by the drug mafia, and was eager to see major advances in the re-incorporation into civil society of those involved in a long-lasting guerrilla warfare. The drug lords saw a new constitution as the perfect instrument in order to finally ban, for good, extradition. Finally, Gaviria understood the possibility of advancing an important political reform, which would lead to a more efficient and less corrupt Congress, while paving the way for a peace agreement with the guerrilla's25. Explicitly, neither economics nor extradition were priorities in his agenda regarding matters to be addressed by a constitutional assembly. He requested the Supreme Court permission to convene a Constitutional Assembly that would deal with a rather restricted agenda.

The actual process that led to the convening of a Constitutional Assembly began as a spontaneous act led by students who realized that a major constitutional amendment would never make it through Congress. They gathered enough signatures so that on the occasion of the May 1990 presidential election the people would vote for or against the convening of a Constitutional Assembly. While Gaviria comfortably won the election, 87% of voters responded "yes" to the convening of the assembly. Suddenly, Colombia had a new President, with a major mandate for change, and with a powerful instrument to do so.

Still, given the long tradition of Courts overturning constitutional amendments, Gaviria did not take things for granted. In order to minimize the possibility of a judicial setback, he requested the Supreme Court permission to convene the election of members to a Constitutional assembly, which would work within a rather limited, pre-defined agenda (Vargas, 1993 and Silva, 1997).

To be sure, Gaviria also wanted to leave extradition out of the subject matters to be dealt with by the assembly. Of course, the mafia had different ideas. In the months following Mr. Gaviria's inauguration, the mob kidnapped several prominent citizens, including the daughter of a former presidentas well as a junior member of the family that owned Colombia's leading daily.

Under a scenario of great expectation, the Court authorized the President to convene the election of a Constitutional Assembly in December, 1990. To most everybody's surprise, the Court explicitly established that there would be no limits regarding the topics that could be addressed by the assembly. That, of course, put extradition back in everybody's mind.

The results of the election of the members of the assembly represented a big change with respect to what had been traditional in most elections. For once, pluralism was the distinct element. While the liberal party elected 25 members and the conservatives 20 (divided in two factions), Alianza Democrática M-19 (the political movement led by members of the former M-19 guerilla movement) managed to elect 19. Minorities, including blacks and indigenous communities, were also represented.

While there is evidence that the more polarized a Constitutional Assembly is, more weight will be given to short-term priorities as opposed to long-term objectives (Myerson, 2000), another element stemming from polarization is also worth highlighting. With all kinds of diverse political interests represented in the assembly, the government certainly had a key role to play. In an assembly in which the extreme right (the Movimiento de Salvación Nacional, a splinter group of the conservative party) had 11 members and the extreme left 19, the government -elected in a landslide just a few months before- would be the natural powerbroker.

While the assembly was in session, the new administration instrumented a peculiar system by which drug dealers who voluntarily turned themselves in would never be extradited. As a result, the Ochoa brothers -the less confrontational and sanguinary branch of the Medellín cartel- landed in jail. Pablo Escobar, the undisputed capo di tuti capi, would not settle for that. Through members of the assembly that were on his payroll, he managed to introduce the banning of extradition in the topics to be discussed by the assembly. Although it is quite likely that only a few members of the assembly were on the take, when the issue was finally voted, it was broadly supported. On the same day that a key vote banned the extradition of Colombian nationals, Mr. Escobar turned himself in, to be housed in a comfortable prison he had build for himself.

On July 5th, 1991 Colombia had a new Constitution, one that with the guidance of the government had been written by a broad spectrum of Colombians. The sense of relief was understandable. For the first time in decades, meaningful legislation had been written outside an exclusionist bi-partisan coalition. If that were not enough, all-important drug-traffickers were in jail, and, with extradition constitutionally banned, the terrorist nightmare seemed all but over. Most Colombians were in a state of euphoria, and Gaviria could claim credit for having been able to achieve something that his three predecessors were unable to.

The new constitution is a vast document that regulates the main aspects of Colombia's political, social and economic life26. As in many Latin American countries, it establishes rights, regulations and obligations in an incredibly detailed way.

Regarding property rights, the Constitution provides an array of rules, which try to strike a balance between protecting private property rights (including promoting privatization) and promoting state ownership of certain resources.

Articles 64-65 provide a vivid illustration of the way in which the new legal order catered to specific interest groups. Article 64 establishes that it is the state's responsability to ensure that agricultural workers become, gradually, landowners. Additionally it states that it is the state's obligation to promote peasants access to "education, health services, housing, social security, entertainment ..." Not one word on how these services would be financed.

A large number of articles are aimed at protecting labor rights and, more specifically, at protecting certain sectors of organized labor. Article 25, for example, establishes that "[e]very person has the right to a job in dignified and just terms". Articles 39, 43 and 44 provide modern civil rights, while Article 54 goes well beyond this by establishing that it is "[t]he obligation of the state and employers to provide technical and professional training to those that require it." In an effort to protect the rights of retired people, Article 53 indicates that the indexation (of pensions) is a constitutional right. The most interesting constitutional provision regarding labor markets is, perhaps, contained in article 334, which states that "[t]he state would intervene to ensure that full employment is achieved..."

The Constitution established a significant degree of political as well as financial decentralization. Articles 339-344 state that the National Development Plan will have to be elaborated after consultation with sub-national governments. A number of articles establish in a very specific way the magnitudes and modalities of transfers from the national to the sub-national governments, independently of the country's other obligations or of its economic performance. Many of the provisions implicitly protect rents related to specific activities and sectors, as highlighted by Wiesner (1997). For example, Article 361 mandates that funds from royalties from any activity related to the exploitation of non renewable natural resources will have to be spent on new mining projects. This provision constitutes a major concession made by the Assembly -with the government's implicit support- to the oil workers' (Ecopetrol) union.

The Constitution also established a number of economic and political institutions aimed at protecting the private sector and citizens from economic abuse. There is no doubt that the most important of these was the establishment of an independent central bank (art 371-373).

In retrospect, and from the strict perspective of the economic reform process, it is possible to argue that the 1991 Constitution was not particularly helpful. In fact, an analysis of its economic provisions suggests that by providing special concessions to specific interest groups, the constitution promoted rent seeking, and created an economic structure inconsistent with an efficient and impersonal modern economy.

Quite soon it would become clear, both to the people and to the Gaviria administration, that no "nirvana" had been attained with the new Constitution. In this regard, the role played by the "Congresito" (for "small Congress") deserves a special mention. With substantial popular support, the Constitutional Assembly had revoked the mandate of the Congress that had been elected for a 4-year term in 1990. With new elections to take place in December, 1991, a temporary legislative body was designated by the assembly. The duties of the "Congresito" had to do with urgent transitory matters of which the new Constitution was full of and with approving the budget. While the "Congresito" was very efficient in advancing the agenda set out both by the government and the assembly -efficiency that added to the general sense of euphoria-, it was soon to be replaced by a new legislative, which, composed by many of the same members of the revoked Congress, was back, with a vengeance. Having many of the same legislators back also highlights the fact that, at least regarding names and faces, the 1991 Constitution had failed in producing a new, enlightened political class27.

TIMING: PRE AND POST CONSTITUTIONAL ASSEMBLY REFORMS

Since early on in the campaign trail Gaviria and his collaborators had been working in the design of an economic reform program. Once in office, the President decided to move fast, taking advantage of the reform mood then prevalent. During its first six months in office -while working on the details of the convening of a constitutional assembly and of the "política de sometimiento" (surrendering policy) of drug-traffickers- the administration submitted to Congress key legislature on economic matters. In September, 1990 it presented the proposal to reform the foreign exchange statute (Estatuto Cambiario) and in October reform packages dealing with the financial sector, taxes, labor regime, housing and ports. In addition, through Presidential decrees trade liberalization was enhanced.

The significance of the reform proposals cannot be underestimated. This point is highlighted by the fact that most of the initiatives had been unsuccessfully attempted in the past. For example, the 1974 and 1979 financial liberalization programs had to be scaled down, when changes in macroeconomic conditions were unfortunately dealt with mainly through financial repression. As a result, in 1990 Colombia's financial sector was highly inefficient, heavily taxed and substantially in government hands (Barajas, Steiner and Salazar, 1999).

Regarding trade liberalization, the policies of 1989-1990 followed two important failed attempts, one in 1965-66 another one in 1978-82. At the end of the 1980s, as described before, Colombia had one of the worlds most distorted trade regimes: tariffs were high and disperse and quantitative restrictions were prominent. With respect to the labor code, a stance highly protective of worker's rights had been in place, basically unaltered, since the 1960s.

Finally, regarding international financial transactions, it is important to underscore that the Estatuto Cambiario, in place since the introduction of Decree Law 444 in 1967, was an institution revered by most Colombians. Through it, a crawling-peg regime was sustained on strict controls on capital mobility. It was credited with providing Colombia with an orderly and stable balance of payments situation, one in which the promotion and diversification of exports achieved through a competitive real exchange rate took center stage.

Surprisingly, all of Gaviria's reform initiatives swiftly made it through Congress by January, 1991. In fact, the bulk of them was unanimously approved in a matter of eight days (Cepeda, 1994). No wonder the whole process came to be known as "el revolcón" . In a matter of weeks, Colombia, the region's most stable economy and the country least prone to change, had apparently introduced one of the broadest reform programs in Latin America. To nobody's surprise, Gaviria was "man of the year" in the leading weekly magazine.

Certainly, the expediency of the process is partly related to the detailed technical work of Gaviria and his staff. However, other important elements should be kept in mind as well. As was mentioned above, most Colombians were eager for change. Whether they supported market-oriented reforms or not was, to a certain extent, irrelevant at the time. Regardless of the fact that the economic situation was not bad, the political and institutional crisis was so deep, that most everybody wanted a change. The immense support received by the spontaneous initiative of a few students to convene a Constitutional Assembly whose agenda would only become clear several months later supports the claim that most everyone wanted "a revolcón."

Second, the first wave of reforms meant that a very popular administration -Gaviria commanded 85% of popular support when he took office- would have to confront an extremely discredited Congress. The legislative body wanted to show the people that, this time around, it was up to its historical task. Not to mention the distinct possibility that the Constitutional Assembly that was being convened could very well decide, as it eventually did, to revoke Congress's mandate.

Following the enactment of the 1991 Constitution, a new Congress was elected and in deliberation late that year. Many revoked members won their seats once again. In his second year in office, the President would have to face a much less collaborationist legislature. To be sure, some members of Congress were glad to get back at the President, who, after all, had agreed to the revoking of their mandate28. Post Constitutional assembly reforms would have a very different fate as compared with those introduced in late 1990.

A third element that helps explain the swift passage of the first wave of economic reforms is related to the comprehensive nature of the entire reformist effort. In particular, the economic reforms of 1990 were introduced at the same time that important attempts were being made at reforming politics while advancing negotiations with both the guerrilla's and the drug lords. In the period stemming from August 7, 1990 -when Gaviria was inaugurated- to July 5, 1991 -when the new Constitution was signed- public opinion was focused on the government's attempts at reaching an agreement with the mafia in order to have the kingpins surrender, in exchange for the assurance that they would not be extradited. Understandably so, in comparison to eventually having Pablo Escobar behind bars, a possibility over which there were new developments almost every day, the liberalization of foreign investment in the financial sector was a subject that hardly made the newspapers. Gaviria and his economic team certainly took good advantage of this golden opportunity.

Finally, a fourth factor that explains the success of the earlier part of Gaviria's reformist effort is related to the ample use of compensation mechanisms, wisely used to broaden the support for the whole initiative. This topic is dealt with in detail in Edwards (1996) and Edwards and Steiner (1999).

Key reforms introduced in 1992, after the enactment of the new Constitution -many of which actually stem from the Constitution itself- included pension and health reform, a new central bank charter, a second financial sector reform, decentralization and privatization. Attesting to the fact that the second wave of reforms yielded less than expected results, in late 1999 the Colombian government entered into an agreement with the IMF. Critical elements of the agreement include the revision of the system that governs the transfer of resources from the central government to departments municipalities and a major overhaul of the pension system.

Several explanations can be offered to the fact that the post-Constitutional Assembly reforms were not as well designed as those introduced in late 1990. From the outset it should be pointed out that some reforms were all but completely defined within the Constitutional assembly. Although they did require posterior legislative work, their key elements were already written in the Constitution. That is the case of the new central bank charter29 -which came out reasonably well- and of fiscal decentralization, which came out with all kinds of problems30.

In addition, there is the issue of a much less collaborationist Congress, following the October 1991 re-election of several members of congress whose mandate had been revoked only months before. And of course, several events happened in 1992 which substantially diminished Gaviria's popularity and which subsequently weakened him in the political arena. Cepeda (1994) refers to 1992 as Gaviria's "horrible year." His popularity plummeted from 78.8% at the beginning of the year to 34.9% at the end of the first semester. In May, 66.7% of those surveyed complained about "the bad economic situation" and 74.3% said they would never vote for Gaviria again. By the way, presidential re-election was banned in the 1991 Constitution.

Several events typify the "horrible year." First, and partially as a result of a severe drought brought about by "El Nino" and partially as a consequence of severe managerial problems in the energy sector, electrical power had to be rationed throughout most of the year. At its peak, power was rationed for eight hours a day. If having no power wasn't disturbing enough, Gaviria's handling of the affair made matters worse. He went on national television to inform that after a very detailed investigation, managerial problems were explained by the ineptitude of a few low-level bureaucrats in his administration. Of course, people were expecting big heads to roll, as a result of the fiasco.

On July Gaviria's biggest nightmare became a reality. Adding insult to injury, Pablo Escobar -who lived the good life in a so-called jail he had build for himself- escaped during a sloppy operation undertaken in order to transfer him to a real jail, one in which he would presumably stop running his business.

During the second half of his administration, restoring power and re-capturing Escobar became Gaviria's most important priorities. Structural reforms took a back seat, in the context of a relationship between the executive and the legislative that was entirely different from the "honeymoon" the President had enjoyed during his first year in office.

6. CONCLUSION

In this paper we have analyzed the conditions that led, in 1989-1990, to the launching of a major reform effort in Colombia. In particular, we have investigated whether the Colombian experience conforms with the so-called "crisis hypothesis" of economic reform. Throughout the analysis we have argued that although Colombia was not facing a major economic crisis at the outset of the reformist effort, it was certainly facing a severe political and institutional breakdown. The profound attention devoted by the general public and by most key actors -including, prominently, the Congress- to political issues gave incoming President César Gaviria ample room to maneuver in the economic front. As a result, his first six months in office were truly remarkable in terms of economic reform. At the time, the enactment of a new Constitution was seen as one of the most important achievements of the Gaviria reform program.

As it turned out, however, following the drafting of the new Constitution, interest groups negatively affected by the reforms were able to regroup, and started to actively lobby to stall the modernization process. As a result, the reformist effort slowed down significantly during the second half of the Gaviria administration (Edwards 1996 and Edwards and Steiner 1999). What is more serious, however, is that many of the constitutional mandates have ended up placing Colombia on an unsustainable fiscal path. In this regard, it is telling that the current arrangement with the IMF calls for reforming two distinct elements of the 1991 Constitution: the system of inter-governmental transfers and the social security system.

* The author's acknowledge financial support from the OECD, Colciencias and Banco de la República. At different stages, Daniel Cerquera and Natalia Salazar provided able research assistance. Previous versions of this paper were presented at Universidad de los Andes, Lehigh University, the IMF Institute and at the 1999 meetings of LACEA.
** University of California, Los Angeles. National Bureau of Economic Research and IAE, Universidad Austral (Argentina).
*** Universidad de los Andes (Colombia).

1 The overall fiscal balance went from a 1% of GDP surplus in 1995 to a 6.3% deficit in 1999. The latter reached 7.1% in the case of the central government. Most departments and municipalities were not current on their debt obligations. Growth stalled in 1998 and GDP contracted by 4.5% in 1999, with unemployment reaching 20%. While there are exogenous elements that partially explain the demise of the Colombian economy -including the deterioration in the terms of trade, a complicated international financial situation, problems in some of its major trading partners and an all but un-declared war with the guerrilla's—, it is evident that severe structural problems still abound.

2 To add to this depressing scenario, in the indexes provided by the IBC International Country Risk Guide and by Johnson and Sheehy (1997) Colombia scores very low in terms of the quality of the judiciary, in terms of corruption, and in regard to the quality of the national bureaucracy.

3 A well-known evaluation of the Gaviria administration is that of Cepeda (1994). His assessment of the reformist effort is extremely positive. He argues (our translation, pg., xiii) that "the reform program undertaken by César Gaviria was one of the most integral and successful ever known…. It was impressive because of its depth, scope (economic, political and institutional), speed of implementation and, of course, because of the positive results it is already showing". It should be noted that this evaluation, which was sponsored by the National Planning Department, was published only 3 months after Gaviria left office.

4 According to a Central Bank report, of all the changes introduced with the 1991 Constitution, the one with the greatest bearing in the increase in public expenditure has been the transfer of revenue from the central government to territorial entities. According to Ocampo (1997), the overall permanent fiscal cost of the laws motivated by the new constitution -decentralization, social security, and wage adjustments to the military and the judiciary- amount to 4.2% of GDP, explaining 78% of the increase in total public expenditure. Today, 30% of public sector employees are still covered by a regime in which severance payments are twice retroactive and in which they can retire at an earlier age. Although they account for only 8% of the entire workforce, their pension liabilities are a most serious strain on fiscal policy.

5 Detailed coverage of the analytics of the political economy of reform can be found, among others, in Williamson (1994), Haggard and Kaufman (1992), and Tommasi and Velasco (1996).

6 In spite of their elegance, however, models based on the Stolper-Samuelson framework have a number of limitations. They assume that the interests of workers and capitalists are independent of the sector where they operate initially, and they ignore important macroeconomic considerations, including the potential role of the exchange rate. Extensions of the basic Stolper-Samuelson framework, however, allow for additional actors, as well as for complex relationships between them. A powerful extension assumes that some of the factors (capital, say) are sector specific. In this case, capitalist interests differ depending on which commodities they are producing at the time the reforms are launched.

7 Domínguez (1997,7) defines technopols as follows: "Technopols are a variant of technocrats. In addition to being technocrats... technopols are political leaders (1) at or near the top of their country's government and political life (including opposition political parties) who (2) go beyond their specialized expertise to draw on various different streams of knowledge and who (3) vigorously participate in the nation's political life (4) for the purpose of affecting politics well beyond the economic realm and who may, at times, be associated with an effort to "remake" their country's politics, economics, and society. Technopols so defined may operate in either authoritarian or democratic regimes."

8 Nelson (1994) argues that in general it is more relevant to focus on the sustainability of reform than on its inception. The latter is generally facilitated by the so-called "honeymoon effect."

9 This sub-section and the next draw partially from Jaramillo, Steiner and Salazar (1999).

10 Lijphart (1977, p. 1) distinguished a consociational democracy as one in which "centrifugal tendencies inherent in a plural society are counteracted by the cooperative attitudes and behavior of the leaders of the different segments of the population". By means of proportionality rules, grand coalitions and mutual vetoes, leaders of different subcultures are able to provide reasonable stability to an otherwise polarized society. Other analysts have referred to Colombia as a "modified democracy," with qualifications such as "controlled", "oligarchic" or "bipartisan elitist". Some have called it a "formally authoritarian democracy" or a "constitutional dictatorship" (Hartlyn, 1993).

11 Recent violence has occurred outside of any power dispute between the traditional parties, although many of its causes can be traced back to the exclusivist nature of the National Front. On these matters, the essays collected in Leal (1995) are very relevant.

12 A 1968 Constitutional amendment prolonged parity in cabinet and administrative positions for four years and established that after this extension the party other than that of the President would receive "equitable" bureaucratic participation.

13 The 1991 Constitution maintained these rules for the lower house of Congress. To facilitate matters for political minorities, Senate seats are now assigned by proportional representation rules applied in a national electoral district.

14 It is revealing to note that in 1987 the Central Bank organized a very well attended international seminar to commemorate the twentieth anniversary of the Estatuto Cambiario, the 1967 piece of legislation that laid the foundation for the crawling peg system.

15 On the influence of drugs on the Colombian economy see Steiner (1998).

16 In many ways, Colombia faces a unique external environment, since the Drug Enforcement Administration plays an important role in determining the US attitude towards the country. The DEA advises the president on whether to "certify" different governments as having made progress with respect to the war against drugs, and is very influential in determining the administration's support -or lack of it- for certain programs.

17 For better or for worse, Colombian economic technocrats have frequently been inclined to reach broad consensus on economic policy. This tendency was illustrated by a symposium held in early 1978, during the peak of the coffee boom and in the midst of a huge accumulation of reserves. Wiesner (1978) reports that of the seventeen young economists who expressed their views during the discussions, only one suggested doing away with exchange controls. The voices of these economists were not simply those of academia. Of the seventeen speakers, six would eventually become finance ministers, three would hold ministries in other economic areas, and seven would become either advisors to the monetary board or members of it successor institution, the Central Bank Board. Only in the 1990s did this trend towards consensus begin to dissipate.

18 Interestingly enough, in spite of having been Minister of Finance (1986-87) and Minister of Government (1987-90) during the Barco government, Gaviria did not participate very actively in the formulation of that administration's liberalization program. Gaviria's tenure as Finance Minister was rather short, and he devoted almost every effort to designing -and persuading congress to pass- a substantial tax reform.

19 Some have intimated that Gaviria came from the provinces (Pereira), which affected his economic views. He is supposed to have been horrified by the influence of interest groups and by the multiplicity of deals that were cut between politicians and private sector representatives in Bogotá.

20 Front-runner in the liberal party primaries, he embraced a modernist conception regarding economic policy. He was an outspoken enemy of the drug mafia. When he was slained by the mob in late 1989, he became an instant martyr. On the request of Galán's eldest son, César Gaviria, his campaign manager, took over as leader of his political movement.

21 For more details see Thomas (1986).

22 Details on this sordid period of Colombia's contemporary history can be found in Silva (1997).

23 In his effort to become a presidential candidate that would unify the liberal party, Galán had asked former minister César Gaviria, who did not belong to the Nuevo Liberalismo wing of the party, to become his campaign manager. In this position Gaviria not only gave the campaign a sound organizational base, but also helped Galán develop a reformist view regarding the economy.

24 The specific details of the most important reforms can be found in Edwards and Steiner (1999).

25 While President Barco successfully completed a political settlement with the M-19 guerrilla movement, when Gaviria took office the two main guerrilla factions (FARC, for Fuerzas Armadas Revolucionarias de Colombia and ELN, for Ejército de Liberación Nacional) were very active. They still are.

26 Regarding politics, the 1991 Constitution established the basis for a "pluralist and participative democracy." A key element in this regard is political and fiscal decentralization, which should bring the government closer to the people, enhancing transparency and accountability. New participation mechanisms were introduced, including the referendum, the plebiscite, the revoking of mandates and the "acción de tutela", according to which any citizen can go before any judge and claim for his "fundamental rights", of which the Constitution is full of. In addition, the balance of power shifted away from the President, to the Congress, and particularly to the Courts.

27 When the assembly agreed to revoke the Congress, it was established that assembly members could not run in the forthcoming congressional elections. That meant that Antonio Navarro, the highly visible leader of AD-M 19, and Alvaro Gómez, founder of a splinter group within the Conservative party, could not lead their groups congressional electoral effort. Partly as a result of this, both movements, which had played a key role in the assembly, saw their political power diminish significantly.

28 The revoking of Congress's mandate received ample popular support. The fact that in the new election several revoked members were re-elected is a puzzling development, which merits a detailed analysis that goes beyond the scope of this paper.

29 A slightly critical assessment of the 1991 central bank charter can be found in Alesina, Carrasquilla and Steiner (2000). A self-serving evaluation from an insider perspective can be found in Clavijo (2000).

30 Most assessments of the 1991 decentralization process offer a very negative balance. In some quarters skepticism was made public very early on. See Correa and Steiner (written in 1994, published in 1999).

REFERENCES

Alesina, A.; Carrasquilla, A. and Steiner, R. (2000), "The Central Bank in Colombia", Fedesarrollo, Working Paper Series No. 13.         [ Links ]

Archer, R.P. and Soberg Shugart, M. (1997), "The Unrealized Potential of Presidential Dominance in Colombia". In S. Mainwaring and M. Soberg Shugart, editors, Presidentialism and Democracy in Latin America. Cambridge University Press. pp. 110-159.         [ Links ]

Barajas, A.; Steiner, R. and Salazar, N. (1999), "Interest Spreads in Banking in Colombia, 1974-96", IMF Staff Papers, 46(2), 196-224.         [ Links ]

Bates, R. and Krueger, A.O. eds. (1993), Political and Economic Interactions in Economic Policy Reform: Evidence from Eight Countries. Cambridge, Mass. and Oxford: Blackwell.         [ Links ]

Cepeda, F. (1994), Dirección Política de la Reforma Económica en Colombia. Bogotá: Fonade.         [ Links ]

Clavijo, S. (2000), "Central Banking and Macroeconomic Coordination: The Case of Colombia", Banco de la República, Borradores de Economía, Nº 159.         [ Links ]

Correa, P. and Steiner, R. (1999), "Decentralization in Colombia: Recent Changes and Main Challenges", published in the Proceedings of a 1994 Conference Held at Lehigh University, C.C. Callahan and F. Gunter (eds.) Colombia an Opening Economy?", JAI Press.         [ Links ]

Cukierman, A. (1992), Central Bank Strategy, Credibility, and Independence: Theory and Evidence, Cambridge, Mass., MIT Press.         [ Links ]

Dix, R.H. (1980), "Consociational democracy". Comparative Politics. 12 (3): 303-321.         [ Links ]

Dix, R.H., (1990), "Social Change and Party System Stability in Colombia." Government Opposition. 25: 98-114.         [ Links ]

Domínguez, B. (1997), International Trade and Finance: New Frontiers for Research: Essays in Honor of Peter B. Kenen. Cambridge; New York and Melbourne: Cambridge University Press.         [ Links ]

Dornbush, R. and Edwards, S. (1991), "The macroeconomics of populism." In: R. Dornbush and Edwards S., editors. The Macroeconomics of Populism in Latin America. NBER and University of Chicago Press.         [ Links ]

Echandía, C. (1999), El Conflicto Armado y las Manifestaciones de Violencia en las Regiones de Colombia, Oficina del Alto Comisionado para la Paz - Presidencia de la República de Colombia, Bogotá.         [ Links ]

Edwards, S. (1995a), Crisis and Reform in Latin America: From Despair to Hope. Washington D.C.: The World Bank.         [ Links ]

Edwards, S. (1995b), "Un Cuarto de Siglo de Fedesarrollo: Una Perspectiva Personal." En: H. Gómez B., editor. Economía y Opinión. Bogotá: TM Editores-Colciencias.         [ Links ]

Edwards, S. (1996), The Political Economy of Incomplete Market-Oriented Reform: The Case of Colombia. Los Angeles: University of California, Los Angeles and NBER.         [ Links ]

Edwards, S. (1997), Trade Liberalization Reform and the World Bank. AER 87(2).         [ Links ]

Edwards, S. and Steiner, R. (1999), "Neo-Liberalism in Macondo: The Political Economy of Structural Reforms in Colombia". University of California and Universidad de los Andes.         [ Links ]

Haggard, S. (1994), "Inflation and Stabilization". In: Meier, G. editor. Politics and Policy Making in Developing Countries: Perspectives on the New Political Economy. San Francisco: ICS Press.         [ Links ]

Haggard, S. and Kaufman, R. eds. (1992), The Politics of Economic Adjustment: International Constraints, Distributive Conflicts, and the State, Princeton: Princeton University Press.         [ Links ]

Haggard S. and Webb, S. (1994a), Voting for Reform: Democracy, Political Liberalization and Economic Adjustment. Oxford: Oxford University Press.         [ Links ]

Haggard, S. and Webb, S. (1994b), "Introduction". In: S. Haggard and S. Webb, editors. Voting for Reform: Democracy, Political Liberalization, and Economic Adjustment. Oxford University Press.         [ Links ]

Hartlyn, J. (1993), La Política del Régimen de Coalición. Bogotá: Tercer Mundo Editores-UniAndes-CEI.         [ Links ]

Hommes, R. (1990), Una Propuesta Económica para los Noventa. Estrategia Económica y Financiera 141.         [ Links ]

Jaramillo, J.C.; Steiner, R and Salazar, N. (1999), La Economía Política de la Política Cambiaria en Colombia. Coyuntura Económica, June.         [ Links ]

Inter-American Development Bank, IADB. (1995), "Overcoming Volatility." in Economic and Social Progress in Latin America. Washington: IDB.         [ Links ]

Johnson, B. and Sheehy, T. (1997), Index of Economic Freedom, Heritage Foundation, several issues, Washington D.C.         [ Links ]

Lal, D. (1987), The Political Economy of Economic Liberalization. The World Bank Economic Review 1,2.         [ Links ]

Lal, D. (1997), The Poverty of Development Economies. IEA, Hobart Paperback 16.         [ Links ]

Lijphart, A. (1977), Democracy in Plural Societies: A Comparative Exploration. New Haven: Yale University Press.         [ Links ]

Leal, F. (1995), En Busca de la Estabilidad Perdida, Bogotá, Colombia: Tercer Mundo Editores/IEPRI/Colciencias.         [ Links ]

Londoño, R. (1986), "La Estructura Sindical Colombiana en la Década del 70." In: H. Gómez, R. Londoño, J.L. and G. Perry, editors. Sindicalismo y Política Económica. Bogotá, Colombia: Fedesarrollo/Fescol.         [ Links ]

Meisel, A. (1996), "Autonomía de la Banca Central e Inflación: La Experiencia Colombiana, 1923-1995." Borradores Semanales de Economía. No. 49. Bogotá, Colombia: Banco de la República.         [ Links ]

Ministerio de Trabajo y Seguridad Social. Boletín Estadístico del Trabajo y la Seguridad Social. Bogotá. several issues.         [ Links ]

Myerson, R. (2000), "Economic Analysis of Constitutions", Discussion Paper Nº 1291, Northwestern University.         [ Links ]

Naim, M. (1995), The Second Stage of Reform. In Louis Goodman et al. Eds., Lessons of the Venezuela Experience. Washington, D.C.: Woodrow Wilson Center Press; Baltimore and London: Johns Hopkins University Press.         [ Links ]

Nelson, R.R., Shultz, T.P. and Slighton, R.L. (1971), Structural Change in a Developing Economy: Colombia's Problems and Prospects. Princeton, NJ, United States: Princeton University Press.         [ Links ]

Nelson, J. (1994). Panel Discussion. In J. Williamson, ed., The Political Economy of Policy Reform. Washington, D.C.: Institute for International Economics.         [ Links ]

Ocampo, J. A. (1997), "Una Evaluación de la Situación Fiscal Colombiana". Fedesarrollo. Coyuntura Económica. 27, 2, junio. Bogotá.         [ Links ]

Olson, M. (1982), The Rise and Decline of Nations. New Haven, Ct: Yale University Press.         [ Links ]

Policía Nacional de Colombia, Revista de la Policía Nacional, several issues.         [ Links ]

Rajapatirana, S.; De la Mora, L.M. and Ravindra, Y. (1997), Political Economy of Trade Reforms, 1965-1994: Latin American Style. World Economy 20(3).         [ Links ]

Rodrik, D. (1994), "The Rush to Free Trade in the Developing World: Why so late? Why now? Will it last?" In S. Haggard and S. Webb, editors, Voting for Reform: Democracy, Political Liberalization, and Economic Adjustment. Oxford University Press.         [ Links ]

Sánchez, F. and Núñez, J. (2000), "Determinantes del Crimen Violento en un País Altamente Violento: El Caso de Colombia", mimeo, Universidad de los Andes.         [ Links ]

Silva, M. (1997), "César Gaviria: Los Años del Revolcón", in Nueva Historia de Colombia", Editorial Planeta.         [ Links ]

Soberg Shugart, M. and Nielson, D.L. (1997), "Constitutional Change in Colombia: Policy Adjustment through Institutional Reform." Comparative Studies (forthcoming).         [ Links ]

Steiner, R. (1998), "Colombia's income from the Drug Trade", World Development, June, Vol. 26, Nº 6.         [ Links ]

Suescún, R. (1992), "Inflación y Devaluación como un Fenómeno Fiscal: la Financiación Óptima del Gobierno a través de la Tributación, el Señoreaje y las Utilidades por Compraventa de Divisas." Ensayos Sobre Política Económica. 22: 7-50.         [ Links ]

Thomas, V. (1986), Macroeconomía y Política Agropecuaria: La Experiencia Colombiana. Bogotá, Banco de la República.         [ Links ]

Tommasi, M. and Velasco, A. (1996), Where Are We in the Political Economy of Reform?, Policy Reform. vol. 1.         [ Links ]

Urrutia, M. (1976), Historia del Sindicalismo en Colombia. Medellín: Editorial La Carreta.         [ Links ]

Urrutia, M. (1983), Gremios, Política Económica y Democracia. Bogotá: Fondo Cultural Cafetero.         [ Links ]

Urrutia, M. (1991), On the Absence of Populism in Colombia. In R. Dornbush and S. Edwards, The Macroeconomics of Populism in Latin America, Chicago: University of Chicago Press.         [ Links ]

Vargas, M. (1993), Memorias Secretas del Revolcón: La Historia Intima del Polémico Gobierno de César Gaviria, Revelado por uno de sus Protagonistas. Bogotá: Tercer Mundo Editores .         [ Links ]

Wiesner, E. (1978), "Devaluación y Mecanismo de Ajuste en Colombia." In: E. Wiesner, editor. Política Económica Externa de Colombia. Bogotá, Colombia: Asociación Bancaria de Colombia.         [ Links ]

Wiesner, E. (1997), La Efectividad de las Políticas Públicas en Colombia: Un Análisis Neo-institucional. Bogotá: Ministerio de Hacienda y Departamento Nacional de Planeación.         [ Links ]

Williamson, J. (1994), In search of a Manual for Technopols. In J. Williamson, ed,. The Political Economy of Policy Reform. Washington, D.C.: Institute for International Economics.         [ Links ]

Williamson, J. and Haggard, S. (1994), The Political Conditions for Economic Reform. In Williamson, J. op.cit.         [ Links ]

World Bank. Global Development Network Growth Database, William Easterly and Hairong Yu, www.worldbank.org.        [ Links ]