THE IMPACT OF PATTERNS OF INDUSTRIALIZATION AND OF POPULAR SECTOR INCORPORATION ON POLITICAL REGIME TYPE: A CASE STUDY OF COLOMBIA

JONATHAN HARTLYN Vanderbilt University A

central question confronting developing nations today is that of the possibility for humane development, combining economic growth with political liberty and social equity. In Latin America, with the turn to repressive military governments in the 1960s and 1970s by many of the more advanced industrial states--Brazil, Argentina, Chile and Uru g u a y - t h e optimistic view that earlier had linked increased modernization and industrialization in Third World countries to democracy had to be reassessed. In addressing this issue, some analysts have focused particularly on issues relating to the continent's colonial heritage and political culture (see Wiarda, 1980). Others have paid special attention to external factors such as the growth of the accumulation of capital on a world scale (e.g., Petras, 1981). A third group has stressed the interaction of external factors, i.e. the world capitalist system, with internal factors, i.e. national political economies, in attempts to determine those structures that constrain or preclude processes such as democratization. A number of analysts within this last approach which might be termed a "historical-structural" perspective (cf. Bennett & Sharpe, 1982) have argued that the possibilities for democratic rule become increasingly limited, as economic growth and industrialization proceed within a capitalist framework in Latin America and other Third World countries (e.g. Evans, 1979; Hewlett, 1979). Those writing within this perspective provide similar arguments regarding the kinds of factors they emphasize. However, there still remains large room for disagreement regarding such issues as the extent to which external factors condition internal ones, the degree to which economic factors limit'political outcomes and the nature and strength of structural constraints vis-a-vis the possibilities and maneuverability and political choice (Bennett & Sharpe, 1982: 642-6). One particularly influential argument within this "historical-structural" perspective that has combined historical, political, and economic factors to provide an explanation for the rise of repressive political regimes in the more advanced countries of Latin America is the bureaucratic-authoritarian model (first presented in O'Donnell, 1972).

30

Studies in Comparative International Development / Spring 1984

This model explains the rise of a new type of repressive authoritarian regime, termed a bureaucratic-authoritarian regime. What distinguishes this new type of regime from other regimes is the combination of rule by the military as an institution (rather than by caudillo figures) with a very sharp restriction of political liberties. In particular, the state apparatus is employed to depoliticize organized labor. The model was originally employed to explain the rise of new forms of military rule in Brazil (1964) and Argentina (1966). Following the rise of brutal military dictatorships in Chile and Uruguay in 1973 (and again in Argentina in 1976), some felt the model was potentially generalizable to other countries in Latin America, sparking efforts to test and refine the model (Collier, 1979). Although many aspects of the model have been criticized or modified, certain central considerations remain. Two elements in particular appear to play a central role in generating contradictions that lead to the emergence of bureaucratic-authoritarian regimes: the pattern of industrialization and the socio-political legacies of initial incorporation of the popular sector. By exploring the case of Colombia, a country which has not had a bureaucratic-authoritarian regime, in the context of these two central elements, this paper intends to serve as a contribution to the development of a more comprehensive theoretical framework that will permit a better understanding of the nature and extent of constraints imposed by economic factors and of the range of political options for Latin American countries. In particular, this paper argues that, as the bureaucratic-authoritarian model suggests, the ways in which industrialization proceeded and initial popular sector incorporation occurred have played an important role in determining Colombia's political regime type. However, the patterns by which these evolved in Colombia differed from those posited in the model in ways that promoted the viability of a more liberal political regime, primarily due to particular international, trade, and domestic political factors not sufficiently stressed in the model. In order to clarify this argument, it is necessary to consider briefly the major arguments presented in the bureaucratic-authoritarian model. The model captures the contradictions generated by import-substituting industrialization and popular sector, incorporation in the image of a "performance-demands" gap caused by the poor economic performance and the high popular sector demands that beset some of these countries. Poor economic performance is viewed as an effecl of the timing and

A Case Study of Colombia

31

staging of import-substituting industrialization (IS]). In summary form, this is a result of 1. Disincentives for agro-export sectors due to national politics and erratic international prices. 2. Import-substituting industrialization becoming increasingly import-intensive; because of politics related to 1SI, small, highly protected inefficient firms with low export potential develop, firms which must be "shaken out" as opening up to the world market occurs. 3. Foreign-exchange gaps arising and balance of payments crises occurring. 4. Increasing inflation, caused by a combination of foreign exchange shortages in conjunction with increased popular sector demands and consumption. 5. Attempts to shift to more capital-intensive, foreign technology dependent intermediate and capital goods industrialization (capital deepening) (O'Donnell, 1972: 57-8, 60; 58, 68-9; 60if; 64; 58-9, 70).

High popular sector demands, in turn, are viewed as a result of urbanization, initial industrialization, and the pattern of initial popular sector incorporation in an earlier populist period. This pattern: 1. Creates an urban popular sector (composed primarily of working class and some middle sector (groups) with organizational forms, political allegiances, and ideological tendencies hostile to established sectors and capable of pressing their demands. 2. Helps develop "prematurely high" consumption expectations by organized urban sectors. 3. Induces patterns of politics of out-bidding by populist politicians seeking urban electoral support (O'Donnell, 1972: 62, 70-1: 102if; 79ff).

This "performance-demands" gap sets the stage for the formation of a coup coalition across institutions and actors within the military, the international bourgeoisie, and government economic decision making groups. This coalition is based on shared perceptions of popular sector activation representing a serious threat to the political regime, on the need for a "predictable" economic environment for industrialization, and, in general, on increased frustration and a sense of mutual recognition in a community of shared interests (O'Donnell, 1972: 102if; 79fl). Although the model represents a brilliant synthesis, a number of its aspects have been criticized. In particular, some have argued that the experience of Latin American countries with populism has varied tremendously and that links between populist governments and initial incorporation of the popular sector and between populist governments and the early stages of import-substituting industrialization are not

Studies in Comparative International Development / Spring 1984

32

clear. In addition, although both the fact and the perception of serious economic problems appear to have played a role in precipitating the coups, the presumed link between the rise of bureaucratic-authoritarian regimes and a need for capital deepening (see point 5 above) was largely found to be missing except for the case of Argentina in 1966 (Kaufman, 1979; Serra, 1979; Cavarozzi, 1982). Indeed, many other factors have been suggested as contributing to the rise or non-appearance of bureaucratic-authoritarian regimes (for a list of 16 of them, see Collier, 1979b: 377n). Yet the central elements of industrialization and popular demand-making addressed by the model continue to be viewed as important (Hirschman, 1979: 82; Kaufman, 1979: 248; cf..Bennett and Sharpe, 1982). For over two decades Colombia has had a limited democratic political regime, as industrialization has proceeded (see Table 1 for comparative economic data). Of the many additional factors scholars have considered as important in understanding the relationship between capitalist economic growth and political regime types in Latin America, the Colombian case adds three that are of particular importance in explaining differences in Colombia's trajectory in comparative perspective. These three are the differential impact of the international system, the country' s initial export structure and the forms of channelling social and political demands in the country. Many scholars have argued the international system has an inhibit-

Table 1. Basic Economic Indicators for Selected Countries

Average Annual 8rowth rate (%) GDP GNP/pc (1978 US$)

Manufacturing as % of GDP

(1960-70)

(1970-78)

(1960)

(1978)

Argentina

1,910

4.2

2.3

31

37

Brazil

1,570

5.3

9.2

26

28

Chile

1,410

4.5

0.8

23

20

850

5.1

6.0

17

20

1,610

1.2

1.9

21

26

Colombia Uruguay

Source:

World Bank, 1980.

A Case Study of Colombia

33

ing effect on the possibility for democratic politics, in part because of the actions or perceived requirements of transnational corporations (e.g., Gereffi & Evans, 1981). Without denying the constraints the international system imposes on the distribution of material benefits in capitalist Latin American countries, it is still important to note how world market and political conditions can have various effects across countries, effects that may have an impact on the possibilities for political liberty. For example, world trade conditions may affect countries exporting the same product very differently. In addition, shifts in world market and trade conditions for commodity and manufactured products in recent times have meant that slight changes in the timing of industrialization as well as in its stages (regarding, for example, when export promotion is introduced), may lead to important differences in national experiences with industrialization. International political events may also generate important economic support, as occurred for example with the creation of the Alliance for Progress after the Cuban Revolution. The second factor, the nature of the initial export structure, comprises such elements as the different mix of major export products, the extent of initial foreign presence, the degree of concentration in the production of exports, the patterns of distribution of the benefits from the sale of exports, and the consumption patterns of these products within the countries themselves. These elements may all vary significantly from case to case, with important political consequences. The third factor of central importance is how major political and social conflicts are channelled in a country, including a country's political party structure. For the case of Colombia, an understanding of the strength and nature of its political party structure is central to a comprehensive view of how the country has sustained a form of liberal democracy. Since the mid-19th century politics has revolved around the interaction of the principal factions of the country' s two major political parties, the Liberal and the Conservative. The history of the country has been a history of hegemonic one-party rule, violent civil wars, and coalition government, essentially assuring that the two political parties would be the only significant vehicles through which political demands could be successfully channelled. The two political parties were crucial to the creation of the limited democratic "National Front" political regime in 1958, even as their shared rule restrained its democratic nature. 1 The National Front emerged after a long, tortuous process, from the last and bloodiest of

34

Studies in Comparative International Development / Spring 1984

the country's civil wars, known simply as la violencia, which had led to a regime breakdown in 1949. Urbanization, industrialization, and class conflict all played contributing roles in fomenting la violencia, but in the end it resulted from the inability of some party leaders and the lack of desire in others to agree on conflict-regulating mechanisms in the face of the other party's violent activities (see Hartlyn, 1981: 117-20; Molina, 1977: 251-3; Pollock, 1975). A brief period of oneparty hegemonic rule ended with a military coup by General Gustavo Rojas in 1953. As Rojas attempted to prolong his stay in office in a context of economic disarray and with continued violence in the countryside potentially representing a serious threat to the state, leaders of the two political parties agreed to join forces in opposition to him. In 1957 Rojas was ousted by an interim military junta which handed power over in 1958 to the first of several bipartisan National Front governments. Because of the intensive partisan polarization, a return to party rule was inconceivable without extensive mutual guarantees. The country has been ruled by two-party coalition governments since 1958. The next section will examine the pattern of Colombia's early industrialization. It will be followed by a section examining the initial incorporation of the popular sector and the early organizational history of the working class in Colombia. Subsequently, the timing and staging of Colombia' s recent industrialization, in contrast to the pattern posited by the bureaucratic-authoritarian model, will be analyzed. Although the importance of the political party structure will be noted at key places, the thrust of the presentation is to demonstrate the importance of the differential impact of the international system and of Colombia's initial export structure on its experience with industrialization and populism, and on the possibility for the formation and survis, al of the current political regime. INITIAL INDUSTRIALIZATION AND THE AGRO-EXPORT SECTOR

Import-substituting industrialization and growth via the domestic market came to replace growth propelled by the export of primary products for the larger countries in Latin America from the Depression period onward, and even earlier in some cases. The export sector, in the case of countries examined by O'Donnell, buffeted by erratic international prices, was further hurt by government policy moves aimed explicitly at aiding industrialization. In Colombia, for reasons related both to the international context in which trade in its primary export crop took

A Case Study of Colombia

35

place as well as to its initial export structure, no such concerted effort occurred, though industrialization was clearly favored by certain measures. A central element in Colombia's drive toward industrialization was the prior "export-propelled growth" generated primarily by trade in coffee. Yet, in Colombia no populist coalition against the agro-export sector ever successfully formed, and foreign presence was too insignificant to create a focus for strong nationalist movements (see O'Donnell, 1973: 55-60). In this section, we will explore why initial industrialization occurred with a relatively low level of sectoral (agriculturalindustrial) conflict. This, in turn, helps explain why initial patterns of labor organization led to a relatively weak divided labor movement. A sustained coffee boom in the late 1800s and early 1900s set the stage for industrialization. Significantly, it incorporated groups from both the Liberal Party and the Conservative Party into the export trade, even as its characteristics helped block the development of a more radical nationalist politics. Coffee maintained itself as the predominant export item, as foreign trade almost quintupled in value between 1914 and 1929. Representing 49.4 percent of total exports in 1914, coffee increased to 79.8 percent of all exports in 1924, but declining to 54.5 percent in 1930 (Urrutia and Arrubla, 1970: 209). Throughout this period, coffee production in the latifundia of primarily Liberal landowners eventually came to be challenged by production in small family-owned farms established in the Western highlands by colonizers from the predominantly Conservative departments of Antioquia. By the 1920s most coffee in Colombia was grown on small and medium size farms (Palacios, 1979: 345-92, based on 1932 coffee census data; see also Parsons, 1968; Machado, 1977: 108-9). Many of these landowning small growers were much less receptive to radical ideologies which had made inroads in other countries where agrarian wage earners were more prevalent, or in which the major landowners were foreigners (Rojas, 1970: 109). The fact that the major export product was largely in local hands (even though much commercialization was initially in foreign hands) 2 further inhibited the development of nationalist radical political movements built around opposition to foreign penetration. Economic growth and social differentiation picked up tempo in the 1920s. Total foreign trade grew tYom 57 million pesos in 1914 to 268 million pesos in 1929. United States private investment grew from a paltry US$21.5 million in 1914 to US$124 million in 1929. Led by the coffee boom, the first installment of the US indemnification for

36

Studies in Comparative International Development / Spring 1984

Panama in 1923 of US$25 million and a rapid increase in foreign loans, the gross domestic produce grew at an average annual rate of 7.3 percent between 1925 and 1929, primarily as a result of investments in public works and transportation (McGreevey, 1964; Urrutia and Arrubla, 1970: 128; Dix, 1967: 79). 3 These dramatic economic and social changes had political repercussions. An incipient working class began flexing its muscles. Socialist ideas became popular among intellectual circles. A sharp drop in world trade and in financial flows caused by the Depression caught Conservative government unprepared. Conservative Party hegemony in place since the beginning of the century finally unraveled by 1930, and a Liberal candidate won the election with bi-partisan support by a plurality of votes. Important steps to accommodate the dramatic changes in the world economic situation brought on by the Depression were taken by the new government. These included large devaluations of the peso, instituting a new protective tariff and a law enabling the creation of stock corporations. Major industrial growth and organization and incorporation of the working class came during the period of President Alfonso L6pez Pumarejo (1934-1938). Yet, in part because the prior level of industrialization was so low in Colombia, the Depression years did not bring a massive project of industrialization or creation of a powerful industrial class. Furthermore, the bi-partisan participation in the coffee trade was reflected as well in industry. Although many industrial interests identified with the Liberal Party, much of the country's incipient industrialization occurred in Antioquia, a region traditionally Conservative in politics (Poveda Ramos, 1967). The initial industrializers were closely allied to the dynamic exporting sector and accepted the legitimacy of its demands. The coffee exporters, organized in the Federation of Coffee Growers (FEDERACAFE - Federacion de Cafeteros de Colombia), were unquestionably the single most important actor in the private sector. Although the Federation began as a weak organization, by the mid-1930s it had consolidated itself as a powerful representative for large coffee producers and exporters. Throughout the 1930s the Federation and national private interests with government assistance continued to garner greater control over commercialization of coffee. The Federation's major efforts were oriented toward rationalization of the market and creation of an internal market structure that could operate with some autonomy vis~-vis the external market. Both internal and external factors helped determine that the Federa-

A Case Study of Colombia

37

tion would remain relatively independent from the state and that conflict with industrial interests would be muted. Because coffee was such a central element of the economy, the bi-partisan export sector was interested in preventing sectarian partisan interference. Thus, an essential guarantee of freedom from partisan manipulation was independence from the state (see Machado, 1977: 80-1; Oquist, 1978: 183226). The fact that coffee was produced by so many growers also made politically (and electorally) more difficult any attempt by the state to control the trade and transfer surpluses to the industrial sector. An external factor also helped Colombian coffee-growers. They could take advantage of the efforts by Brazil to increase world coffee prices. Because of the Brazilian efforts, Colombia was never forced to curb production of its coffee sector. Brazil held practically a monopolistic position in world coffee trade during these years, producing three-quarters of the coffee traded on world markets in 1901-15 and two-thirds during the 1920s. Unlike Colombia, in Brazil coffee was produced primarily on large estates. Because of Brazil's near monopoly situation and because production could continue expanding in Brazil, the planters encouraged first the state governments and then the federal government to institute both price support and supply restriction programs. The former succeeded and the latter failed both domestically and abroad. World prices were kept high enough to encourage more and more production from other countries such as Colombia. Colombian growers benefited from the price-support efforts of their Brazilian counterparts (Furtado, 1963: 193-213; Bergsman, 1970: 21-2). The Colombians, producers of a milder coffee which sold at a fairly substantial premium over Brazilian coffees, strengthened their market position vis-h-vis Brazil during the 1930s. By 1932 Brazil's strenuous coffee defense policy, prohibiting new planting and punctuated by the destruction of 57 million (60 kg.) bags of coffee between 1931 and 1937, caused the price differential between the two coffees to narrow. In that situation demand shifted to the better value and Colombian coffee exports expanded at Brazilian expense. In 1933 the Brazilian federal government through the National Coffee Department took full charge of the coffee support program. Finally, in late 1937 Brazil abandoned the price-defense policy (Wickizer, 1951: 78-84). Brazilian investment resources were directed from coffee through the state to other sectors, chiefly manufacturing (UN/ECLA, 1957:263; Hirschman, 1968: 11-12). Both in Brazil and Colombia the exporters were sufficiently strong

38

Studies in Comparative International Development / Spring 1984

to insure that a precipitous drop of over 50 percent in world coffee prices during the Depression years did not drastically affect their real income. In Colombia, by 1935 a series of devaluations caused the value of the Colombial peso to decline by 70 percent (Poveda Romas, 1976: 58). One group of economists noted: During the period under consideration [the immediate post-1930 years] the external sector's income in pesos at current prices seems to have increased far more than internal prices and costs so that the Government's policy provided a real stimulus to production for export, despite the depression of the world markets (UN/ECLA, 1957: 42).

These devaluations also had the effect of aiding and protecting established industries, producing for the domestic market. Much of this incipient industrialization was initiated by large landowner-merchants. Industrialization proceeded at a rapid rate. Its annual growth rate between 1933 and 1939 averaged 10.8 percent. Domestic production of import substitutes became more profitable, as the prices of imported manufactured goods became more expensive because of devaluation and high tariffs. The period was marked paradoxically by expansion of industrial production and relative stagnation of productive capacity. The .increase in production came primarily from more efficient utilization of available capacity (UN/ECLA, 1957: 262). During these years, however, industrialization proceeded from a very low initial level and diversification was limited. Unlike the situation suggested by the bureaucratic-authoritarian model, there was no "rupture" in Colombia between industrialists and export-oriented landowners during the 1930s. One reflection of this low sectoral conflict was the industri-

alists' ambivalent relationship with labor which inhibited creation of a clear populist coalition. INITIAL PATTERNS OF LABOR I N C O R P O R A T I O N

The ability of labor to press its demands upon a political regime depends upon the nature and extent of its organization. The bureaucratic-authoritarian model envisions the development of an organized popular sector during the populist period, encouraged by specific public policies of a regime that considers labor an important ally. That popular sector, either ideologically motivated or mobilized by populist politicians, continues to press its demands upon the system even when the country's economic performance declines. Eventually, the popular sector which is "incorporated" into populist regimes is "excluded" in bureaucratic-authoritarian ones (O'Donnell, 1973: 53-95).

A Case Study of Colombia

39

In Colombia no clear pattern of such inclusion or exclusion can be traced. Initial sponsorship by the Liberal Party, followed by Conservative Party repression and Catholic sponsorship of "apolitical" unions, led to a labor movement which was relatively independent from the state, linked to the political parties, and never powerful or autonomous. In the early years of unionization labor was not an important center for radical activity. This was due in part to low rates of immigration and to low levels of foreign investment capital in these years of initial industrialization. The first major labor federation, the Confederation of Colombian Workers (CTC), was founded in 1936 with important Liberal, Communist, and Socialist factions. Its creation was aided by the Communist Party's policies of collaboration and accommodation with the Liberals in the late 1930s and early 1940s (for a "self-criticism", see Partido Comunista de Colombia, 1960). However, the broad support from the government of President Alfonso Lrpez Pumarejo (1934-38; 1942-45) was essential. Indeed, between 1935 and 1939 only 44 of 218 labor-management conflicts were solved without government intervention which was usually on behalf of labor (Urrutia, 1969:118-21, 160). Even during this initial period of state sponsorship of labor, the union movement never became fully dependent on the state (see Pecaut, 1973: 149-54). President L6pez's sponsorship of labor organization met intense opposition from Conservatives as well as within his own party. By the end of his first term in office L6pez himself called for a "pause" in reform measures which in effect occurred in the 1938-42 period. In those years Liberal President Eduardo Santos attempted, with partial success, to divide the labor movement, weakening its Communist wing. Lrpez rode the crest of popular support back to the Presidency in 1942 in an election that bitterly split the Liberal Party. This second term was a period of accommodation, wartime speculation, and economic difficulty, rather than one of continued vigorous reform. L6pez resigned under pressure in 1945, after surviving a coup attempt and after enacting significant new labor legislation. His replacement, the rising Liberal politician Alberto Lleras, formed a bi-partisan government and began the repression of labor which was to intensify in subsequent years. Lleras broke a strike by the Federation of River Workers, destroying what had been the strongest labor union in the country (Pecaut, 1973). Repression of the Liberal labor federation, the CTC, intensified

40

Studies in Comparative International Development / Spring 1984

when the Conservative Mariano Ospina Prrez was able to assume office in 1946 due to a serious division within the Liberal Party. In these elections most of the labor leadership supported the "official" Liberal candidate, while the rank-and-file supported the Liberal populist Jorge Eliecer Gaitan. Subsequently, the CTC suffered further internal dissension, as the Cold War ended the previous unity between Liberals and Communists. In 1950 the Communists opted to withdraw from the CTC in an unsuccessful attempt to form their own federation. Under Conservative rule a new labor federation, the Union of Colombian Workers (UTC - Unirn de Trabajadores de Colombia), founded in 1946 with Jesuit assistance, flourished. Formed on strict principles of collective bargaining and with a much more centralized organization, the UTC was not interested in state syndicalism. Given these characteristics, the UTC prospered under President Ospina, who saw this as an alternative to the Liberal and Communist oriented CTC. The UTC's initial successes in organization came in the manufacturing sector and in the creation of peasant leagues in the countryside. The growing conflict between the Conservatives and the Liberals over control of the state apparatus, of which repression of the CTC formed a part, increasingly escalated into open violence, particularly in rural areas, taking on different characteristics and intensity throughout the country (Oquist, 1978). In 1949 the political regime broke down: a state of siege was declared and Congress was closed with political and civil liberties sharply curbed (for details, see Hartlyn, 1981: 93-140; Wilde, 1978). The Conservative Laureano G6mez became President by means of elections boycotted by the Liberals. As violence continued to spread, labor activity practically ceased under Gomrz's increasingly repressive rule. With most major Liberal leaders in exile, opposition to Gomrz had to come from within his own party or from the military. In effect, the Ospina faction of the Conservative Party supported a coup by General Gustavo Rojas Pinilla in 1953. Rojas came to power on the behest of the political parties and elements of the military. These groups sought an end to the civil conflict which they felt Gomrz's rule was exacerbating. Most party leaders expected an eventual, if not prompt, return to civilian rule. Yet the process was to be more complex than they had first imagined, as General Rojas tried to consolidate his position in power. Rojas' overtures toward labor, in sharp contrast to the experience of Grtulio Vargas in Brazil (especially 1937-1943) or Juan Domingo Perdn in Argentina (1946-1955), were largely unsuccessful. The organizational space already occupied by the political parties and existing labor organizations

A Case Study of Colombia

41

as well as domestic implications of the international environment weighed against his success. The CTC, still linked to the Liberal Party, had been nearly decimated by the governments of Ospina and G6mez, and was not anxious to be incorporated into a state structure dominated by the military and by Ospinista Conservatives. The Catholic-inspired UTC grew substantially during the Rojas years as the General courted Church support. Although some UTC leaders did support Rojas, due to the confederation's orientation toward collective bargaining and its independent revenue base it was basically not open to state incorporation.. Rojas tried to establish a new National Workers' Federation (CNT-Confederacion Nacional de Trabajadores) in 1954. It was soon attacked by the church hierarchy in Antioquia, seat of its greatest strength, as "Peronist and anti-Catholic" (Fluharty, 1957: 245-9). Actually, the CNT was linked to the Peronist movement by its rhetoric and by its formal (and probably financial) ties with the Argentine-sponsored Group of Syndicated Latin American Workers (ATLAS -Agrupacion de Trabajadores Latinoamericanos Sindicalizados). Although the Colombian Church had its own reasons to be cautious in its ties with Rojas, Rojas's close connections with Per6n further inclined Colombian Church officials to oppose him. Furthermore, both the UTC and the CTC remained linked internationally during this period to the United States-dominated ORIT (OrganizaciOn Regional Inter-Americana de Trabajadores ) which bitterly opposed the Peronist ATLAS. By August, 1955 Rojas was forced to order the CNT dissolved (Moncayo and Rojas, 1978: 163-70). A subsequent attempt to create a government-sponsored umbrella labor organization also met with failure. In the end Rojas's ties to labor were so weak that both the UTC and the CTC supported (though they did not help initiate) the political movement to depose Rojas and create the bi-partisan National Front. In part, this reflected organized labor's penetration and co-optation by the political parties and its relative weakness (for more recent developments, see Hartlyn, 1982.) In sum, in Colombia the labor sector, atomized, ideologically divided, and already functioning from within the system, was not readily available for state syndicalism. In contrast to the experience of other Latin American countries such as Argentina or Brazil, presented in the bureaucratic-authoritarian model, the post-Depression and post-World War II periods did not spawn an urban popular sector with organizational forms capable of pressing its demands, or with political allegiances and ideological tendencies hostile to the central political and

42

Studies in Comparative International Development / Spring 1984

economic actors. 4 Both the economic and political legacies from this period were substantially different in Colombia than in its more industrialized southern neighbors. FROM IMPORT-SUBSTITUTION TO EXPORT PROMOTION

Another central element of the bureaucratic-authoritarian model is the nature of import-substituting industrialization, particularly in its more advanced stages. It argues the very import-intensity of the industrial process and poor export performance generate foreign exchange scarcities, leading to slowdowns in economic growth. These slowdowns are increasingly problematic, as popular sector protests increase inflation and unemployment. This process becomes acute, as the country needs to provide a stable environment to induce foreign investors to become involved with the import-substitution of intermediate and capital goods (and durable consumer goods). Colombia's experience with import-substituting industrialization shares a number of characteristics with its continental neighbors, including some features described by the bureaucratic-authoritarian model. As elsewhere, import-substituting industrialization was constrained at times by the import intensity of the process and by balance of payments constraints. It was promoted by tariff barriers as well as other restrictions on imports such as licenses, quotas, and prior deposits (which were also employed to manage situations of tight foreign exchange). This protection helped generate direct foreign investment which eventually dominated a number of the most dynamic and fastest growing branches of industry. In addition, because of the small national market which was further fragmented by Colombia's mountainous geography, most industrial sector activities were dominated by a small number of firms (Arango, 1978; Matter, 1977; Misas, 1975). Yet, important differences are evident. Because of the timing of industrialization and because of the structure of coffee production, the relationship of the state with the industrial and agro-export sectors was not as conflictive. This, in turn, eased the shift in policy from one centrally focused on import-substituting industrialization to one which also emphasized export promotion, leading to a different staging of the industrial process in Colombia than suggested by the model. In contrast to arguments posed by the model, this meant that, in conjunction with the different socio-political legacy of populism, Colombia's problems with inflation and balance of payments were less serious and

A Case Study of Colombia

43

that the country's economic growth performance was good for most of this period.

Industrialization: Import-intensity and Protectionism Industrial growth in the post-war period was uneven, although somewhat greater than for the economy as a whole. In 1950 manufacturing contributed 14.1 percent to GDP; in 1975 this was up to a still comparatively low 19.2 percent (Poveda, 1976). Relatively rapid growth rates in industrial value added occurred in the periods 1950-56 (10.2 percent) and 1968-73 (8.0 percent). Average to low rates of growth in the late 1950s and early 1960s (4.2 percent in 1956-60 and 5.4 percent in 1960-68) reflected balance of payments constraints imposed both by low market prices for coffee as well as by emphasis on import-substitution combined with export pessimism, of which we will have more to say in the next section (figures from Am6zquita & Fernandez, 1977: 183-4). During the 1950-75 period there was a considerable shift in the composition of industrial production, reflecting the impact of importsubstituting industrialization (see Table 2). Foreign investment in manufacturing also expanded. By the end of the National Front industries which contained some foreign investment contributed close to half of total value added in manufacturing. The sectoral breakdown of foreign investment in Colombia is given in Table 3. It shows that much of that investment is concentrated in areas of import-substitution, particularly in chemicals, oil products, rubber and plastics, metal products, machinery and equipment, and paper products. One phenomenon which import-substituting industrialization has tended to generate across countries is a demand for new and different types of imports. In displacing the imports of the goods which are now produced domestically, it often creates a demand for a whole series of new imports, especially of capital equipment for the plants and of materials, parts, and components used in these plants. Many economists agree that import substitution "tends, paradoxically enough, to increase the economy's dependence on imports" (Little, Scitovsky and Scott, 1970: 59). Goods countries need to import become crucial for continued economic growth. For that reason, balance of payments crises and import restrictions in more advanced stages of import-substitution, by generating unemployment, work stoppages, or a decline in manufacturing, may create significant socio-political con-

44

Studies in Comparative International Development/Spring 1984 Table 2. Composition and Distribution of Industrial Production, 1950-74

A.

Composition of total industrial supply:

Per cent nationally produced.

1950-2

1955-7

1960-2

1965-7

1972-4

93.7

95.5

96.9

97.5

96.6

61.0

64.3

76.0

81.8

78.3

Durable Consumer and Capital Goods (3)

16.3

20.3

31.7

41.8

51.8

TOTAL

80.1

77.7

80.8

84.8

83.2

Basic Consumer Goods (i) Intermediate Goods (2)

B.

Distribution of domestic production of industrial goods (per cent) 1950-2

Basic consumer goods (i) 82.2 Intermediate goods (2) 16.0 Durable consumer and capital goods (3) 1.8 TOTAL

i00.0

1955-7

1960-2

1965-7

1972-4

72.0

61.8

58.5

55.0

25.1

33.0

35.4

36.2

2.9

5,2

6.1

8.8

I00.0

i00.0

I00.0

i00.0

(i) Includes food, beverages, tobacco, textiles, clothing and shoes, wood furniture, leather and miscellaneous manufacturing. (2) Includes wood, paper and paper products, printing, rubber, chemicals, oil and coal products, nonmetallic minerals, basic metals and metal products. (3) Includes non-electrlcal machinery, electrical machinery and transportation equipment. Am~zquita and Fernandez, 1977: 211-2, 214. Because of high levels of aggregation in industrial sectors, the figures should be taken as approximate guides.

Source:

flict of the type discussed in the bureaucratic-authoritarian model. The extent to which this process of import-substitution has occurred in Latin America has varied substantially from country to country. In some countries import-substitution went much further in a wide variety of industries than in others. One indication of the extent to which countries have followed extreme policies of import-substitution is the degree of protection offered the industrial sector. Inefficient domestic industries require high rates of protection to defend themselves from

A Case Study of Colombia

45

foreign competition. A s u m m a r y figure e m p l o y e d b y economists to provide an overall view of the degree of protection is the effective rate of protection. One study that e x a m i n e d rates o f protection for a variety of less-developed countries found extremely high overall levels of protection for industry. Most of the protection measures resulted from quantitative controls of imports that were initially imposed primarily to manage difficult balance of p a y m e n t s situations. The authors of the study conclude: The upshot is that the very great variability of protection creates a strong presumption that e i t h e r some industries, or parts of industries or plants, in these countries should never have been set up at all, because there was no hope of their ever yielding a good social return to the country; o r policies have been such that they were set up in the wrong way, or such as to prevent their achieving adequate efficiency (Little, Scitovsky & Scott, 1970: 190). Table 3. Foreign Investment in Colombia, 1966-74 Accumulated to end of 1966 (Thousands (%) us $ )

Basic Consumer Goods Food 29,124 Textiles, leathers 17,215 Other 2,351

Accumulated to end of 1974 (Thousands (%)

us $ ) 9.6 5.6 0.8

35,700 21,920 2,991

9.0 5.5 0.8

Intermediate Goods Wood 1,084 Paper 39,797 Chemicals, oil products, rubber and plastics 137,288 Nonmetallic min. 21,985 Basic metals 7,406

0.4 13.0

5,649 44,473

1.4 11.2

44.9 7.2 2.4

182,502 28,720 8,727

46.2 7.3 2.2

Durable consumer and capital goods Machinery and equipment 49,048

16.1

64,782

16.4

SUB-TOTAL MANUFACTURING

305,368

70.2

395,464

67.7

Financial sector 42,415 Insurance 5,526 Commerce and hotels 49,597 Other 32,224

9.7 1.3 11.4 7.4

81,946 7,024 59,000 41,040

14.0 1.2 i0.i 7.0

i00.0

584,474

i00.0

TOTAL Source:

435,130 Arango, 1978: 141.

46

Studies in Comparative International Development / Spring 1984 Table 4. Effective Protection in Manufacturing for Selected Countries

Industry Group

Argentina (1958)

Processed food Nondurable consumer goods Beverages and tobacco Consumer durables CONSUMPTION GOODS (i)

164

Intermediate I Intermediate II INTERMEDIATES

(i)

167

TOTAL MANUFACTURING TOTALMANUFACTURING (i) PRIMARY PRODUCTION

162

Mexico (1960)

Colombia (1969)

255

20

-0.2

218

300

45

37

285

123

85

94 113

230

22 105 195

68

79 93 -26 133

Chile (1961)

92

115 187

Construction materials Machinery Transport equipment CAPITAL GOODS (i)

Brazil (1966)

25 56

26 23

34

154 97 -65

31

-5 38 30

7 31

55

127

158

32

29

118 18

64

27 -3

i

Sources: For Brazil, Chile and Mexico, Balassa et. al., 1971: 54. Effective rates estimated using Corden method employing free-trade input-output coefficients. For Argentina, Brazil and Mexico (marked (i)), Little, Scitovsky and Scott, 1970: 174, using Corden method. For Colombia, Wogart, 1978: 31, citing Hutchenson, 1973: 68, 77.

The effective rates of protection from this study and from others for a number of Latin American countries are listed in Table 4. They suggest that Brazil, Chile, and Argentina followed extreme policies of import-substitution. There is little question that by the late 1950s Uruguay was also burdened by an extensively overprotected and inefficient industrial sector (Gutidrrez, et al., 1979: 24-5). As Table 4 suggests, the Mexican case is different. This is due in part to the fact that its long and permeable border with the United States made an overvalued currency and extremely inefficient industries impractical. From early on Mexico significantly expanded her non-traditional exports. Estimates of effective protection for Colombian industry in Table 4 show rates that are substantially lower than those of Argentina, Brazil and Chile. Although these comparisons are somewhat problematic due

A Case Study of Colombia

47

to varying assumptions built into the calculations, the large differences suggest that Colombia avoided some disastrous experiences in importsubstitution. This is probably due to a combination of factors. Because Colombia's industrialization proceeded somewhat later than that of these other countries, it only began to install a capital goods industry o n a modest scale in the 1960s, whereas the other countries began that process, and the heavy protectionism it entailed, a decade or more earlier. Additionally, Colombia moved toward liberalization at approximately the same time as Brazil moved more vigorously in that direction (in the late 1960s). Therefore, Colombia made the shift at a much earlier relative stage of its industrial development. Because of this significantly different pattern, the country appears to have side-stepped some of the major problems related to import-substitution specified in the bureaucratic-authoritarian model, problems that were beginning to appear in the late 1950s and in the 1960s.

Industrialization, the Agro-Export Sector and Export Promotion As industrialization proceeded in Colombia, foreign exchange requirements continued to intensify. Throughout the 1950s and most of the 1960s coffee remained the central element of Colombia's export sector, and industrial growth was tied to coffee's performance on world markets. As Table 5 shows, the volume of coffee exports in the 1950-75 period was relatively stagnant, much as the primary export products of Argentina, Brazil, and Uruguay were in the 1950s and 1960s. But in Colombia there was a dramatic increase in the export of other goods and commodities in the late 1960s and the value of coffee exports grew appreciably due to price jumps in the 1970s. Colombia's experience with coffee shared certain similarities with the Brazilian case. In the post-war period Brazil maintained an overvalued exchange rate which represented a clear bias against agricultural exports. From 1946 to 1964 the volume of both coffee and noncoffee exports were largely stagnant. The Brazilian government essentially did not try to increase non-coffee exports, and the quantum of coffee exports showed no steady increases during this period (Bergsman, 1970: 99-101). The overvalued exchange rate favored by entrepreneurs in import-substituting industries (who wanted cheap import prices for their capital goods, not by those whose products faced stiff competition without other means of protection) was also favored by the government because of the nature of the world coffee trade. Since foreign demand for coffee was perceived as relatively inelastic

48

Studies in Comparative International Development / Spring 1984

Table 5. Coffee and Minor Exports for Selected Series, 1948-78 YEAR

New York price (US cents per ib)

Exports (thousands 60kg bags)

Recorded c o f f e e e x p o r t s (US$ millions

Coffee exports Z total exports

Recorded minor exports (USqMII.)

1948 1949 1950

32.6 37.6 53.2

5,588 5,410 4,472

225.2 242.3 306.4

73.4 72.3 77.8

36.3 34.7 22.4

1951 1952 1953 1954 1955 1956 1957 1958 1959 1960

58.7 57.0 59.9 79.9 64.4 74.0 63.9 52.3 45.2 44.9

4,794 5,032 6,632 5,754 5,867 5,070 4,824 5,441 6,413 5,938

359.4 379.9 492.3 550.2 487.4 413.1 390.1 354.7 363.4 333.5

74.3 78.6 81.3 82.2 81.7 74.9 76.3 77.0 76.8 71.8

50.9 31.6 36.9 43.1 47.8 67.4 48.7 40.9 36.3 51.1

1961 1962 1963 1964 1965 1966 1967 1968 1969 1970

43.6 40.8 39.5 48.8 48.5 47.4 41.9 42.6 44.9 56.4

5,561 6,561 6,134 6,412 5,651 5,566 6,094 6,588 6,478 6,509

307.9 331.8 303.0 394.2 343.9 328.3 321.5 351.4 343.9 466.9

70.8 71.6 67.8 71.9 63.8 64.7 63.0 62.9 56.6 63.5

58.7 70.8 66.5 78.9 107.0 108.7 127.2 170.6 206.8 210.1

1971 1972 1973 1974 1975 1976 1977 1978

49.3 56.7 72.7 77.9 81.7 157.7 240.2 185.2

6,569 6,528 6,766 6,906 8,175 6,289 5,323 9,034

399.7 428.1 595.5 623.1 680,5 996.0 1,512.6 2,026.8

58.3 52.2 50.6 44.0 47.2 56.1 65.4 68.9

235.1 392.1 580.7 792.5 762.0 777.7 799.9 914.9

Sources;

Col, I, Avramovlc, 1972: 265 (1948-1969), Bdlr, 1979 (1974-78); Col. 2, Palacios, 1979 (1948-1972), Bdlr, 1979 (1973-78); Col. 3, Diaz-AleJandro, 1976: 35 (1948-72), Bdlr, 1979: 1896 (1976-78); Col. 4 & 5, Dlaz-AleJandro, 1976: 35 (1948-72), Bdlr, 1976: 1505 (1972-75), Bdlr, 1979: 1895 (1976-78).

with regard to price and since Brazil was the major world supplier, an overvalued exchange rate could be employed to support world coffee prices (Left, 1968: 14). Naturally, the planters were opposed, since this reduced their cruzeiro earnings for a given world market price (Left, 1968: 23). Throughout the 1950s Brazilian coffee growers converted their dollar earnings at an exchange rate lower than the rates paid by importers. This constituted a severe "export tax" which when

A Case Study of Colombia

49

coupled with a policy of relative price supports, significantly lower than elsewhere, enabled the state to appropriate much of the surplus for itself (Left, 1968: 25-6). Brazil's recurring balance of payments problems were not caused simply by import substitution, but by a combination of world market and trade conditions and the policy view that import substitution and export promotion were incompatible. There was considerable export pessimism which translated into policies inimical to the expansion of non-coffee exports, a view only partially reflective of international export opportunities (Left, 1968: 77-88). This was only to change after the 1964 military coup. In Colombia the major exports in the immediate post-war period, coffee and petroleum, were viewed primarily as limited by world market conditions and thus unresponsive to domestic policy steps. In the 1948-72 period they showed no steady long-term upward trend in terms of dollar earnings and there was considerable industrial export pessimism until the late 1960s (Diaz-Alejandro, 1976: 6; Thoumi, 1980). Some economic analysts viewed Colombia's basic economic problem as a reduced capacity to import, a foreign exchange constraint (Nelson, et al., 1971). As in Brazil, Colombia's exchange rate in the 1950s and 1960s was clearly overvalued. Both the government and the coffee federation were in substantial agreement that the volume of coffee produced in Colombia was in excess of what could be profitably marketed. Domestic coffee support prices have been much higher in Colombia than in Brazil relative to world prices. As in Brazil, Colombia utilized the exchange rate as a form of taxation, maintaining a differential between the effective rate of exchange on proceeds from coffee exports and the main rate (see also the discussion in Hirschman, 1981: 82-3). No massive transfers of funds from agriculture to industry occurred, though import-substituting industry was clearly favored by the overvalued exchange rate (see Kalmanovitz, 1978: 136-83). The contrast of the Colombian situation with those of Argentina and Uruguay is even starker. In Colombia the sluggish performance of the major export crop was not due to lack of local incentives or expanded domestic consumption, but was perceived to be due to circumstances outside of national control. One goal of coffee policy in the 1960s was to increase domestic consumption, while cutting production, in order to be able to reduce the size and cost of surplus stocks) In Argentina and Uruguay major export goods such as wheat and beef are central items of consumption for the urban sector. As the working class and

50

Studies in Comparative International Development / Spring 1984 Table 6. Annual Change in Cost of Living for Selected Countries, 1951-81

Years

Argentina

Brazil

Chile

Colombia

Uruguay

1951-55

19

18

49

3

13

1956-60

41

28

25

i0

23

1961-65

27

63

29

15

38

1966-70

19

28

26

9

60

1971-75

73

21

276

19

71

1976-78

265

42

115

24

51

1979-81

122

80

29

26

55

Sources: For 1951-67, Little, Scitovsky & Scott, 1 9 7 0 : 4 6 8 (using IMF, International Financial Statistics, various issues); for 1968-72 (and Uruguay, 1967), IMF, 1973: 35; for 1973-74, IMF, 1975: 29; for 1975-78, IMF, 1979: 45; for 1979-81, IMF, 1982: 51. Five year and three year averages derived from annual averages.

popular sectors became organized, they pressured the regime for increased domestic consumption at low prices. This not only reduced exports, but provided disincentives for increased production (see O'Donnell, 1977: 523-54). Yet, as world coffee prices fluctuated, Colombia did not totally escape the recurring balance of payments crises that affected its southern neighbors, particularly from 1957 to 1967, when coffee prices were low. Stabilization programs and devaluations with differing effectiveness were imposed in 1957-58, 1962, and 1965. The country's precarious economic situation in the early and mid-1960s generated political crises which in other countries may have led to regime breakdown. However, as Table 6 shows, differences in inflation rates between Colombia and the four South American countries that succumbed to bureaucratic-authoritarian coups, are dramatic. The Colombian political regime was saved by the strength and resiliency of the political parties and by the relatively lower level of popular demand-making, evidenced by the relatively weak labor movement.

A Case Study of Colombia

51

But international assistance was also of central importance. The US intended for Colombia to be a "showcase" of the Alliance for Progress. Throughout the early and mid-1960s, Colombia's precarious balance of payments situation was saved by funds from the Alliance for Progress, the World Bank, and the International Monetary Fund. Between 1961 and 1970 the country received an average of almost 200 million dollars per year, making Colombia, relative to her population, one of the major foreign aid recipients in the 1960s (Berry, 1980: 305). A United States Senate sub-committee report on the effectiveness of the Alliance for Prograss in Colombia claimed that Alliance funds had been continually manipulated for short-term goals of political stability related to budgetary deficits or balance of payments constraints, thus postponing necessary and desired economic and long-term reform: Limitationof financialassistancewouldbe the most powerful short-termtool available to the United States to influenceColombianpolicies with respect to such matters as exchange rates and export promotion, but the use of this tool could at the same time produce shock waves which wouldthreaten political as well as economic stability. (U.S. Senate .... 1969: 20; see also, 153-4). The crises surrounding periods of difficulties with balance of payments, rapid inflation, and large-scale devaluations were managed within the existing National Front political regime. In 1966-67 the Liberal President Carlos Lleras Restrepo (1966-1970) instituted a new framework for trade and foreign exchange (Decree-Law 444 of 1967) that led to significant improvement in the export picture. From the 1950s a number of Colombian and foreign economists had been arguing that a major effort to diversify and promote exports was essential to break-up the tbreign-exchange bottleneck. Since 1956 Colombia had had a modest export incentive program under which the import contents of certain exports, primarily manufactured goods, were exempted from import costs (the Plan Vallejo) (Diaz-Alejandro, 1976: 57-9; Colombia, Fondo . . . . 1977: 1-2; 7-8). Notwithstanding these antecedents, this important policy shift was a result of a nationalist reaction by the new President against another call by the IMF for a large devaluation (Maullin, 1967). It was not entirely obvious at the time that the law had established the means for a partial reorientation of the economy from an importsubstitution model to one of export promotion. One of the major elements of the bill was the establishment of a "crawling peg" devaluation mechanism controlled by the Monetary Board (Junta Monetaria) and the Banco de la Republica by which devaluations would occur

52

Studies in Comparative International Development / Spring 1984

constantly and gradually, rather than brusquely as in the past. Although the real increase in the exchange rate due to devaluation from 1967 to 1975 was only 2.7 percent, this still marked significant progress. In the past significant declines in the exchange rate had been common (Weisner, 1978: 93-4). The same 1967 law established an export promotion fund (PROEXPO - Fondo de Promocidn de Exportaciones) financed by a tax on imports. Its many activities include generous lines of credit, insurance plans, information about foreign markets, and other promotional measures. In the opinion of one prominent economic policy maker, its most important function was probably that of a "pressure group" for export interests within the government bureaucracy in the face of opposing fiscal, monetary, or even social considerations (confidential interview, November 1977). Other incentives to exporters have included generous tax certificates initially provided for all products other than coffee, petroleum and its by-products, and raw cattle hides, worth 15 percent of the value of their exports. These certificates could be employed to pay income taxes as well as sales and import taxes (Diaz-Alejandro, 1976: 55; Colombia, Fondo . . . . 1977). The results, following these policy changes, were impressive. In sharp contrast to the previous period, after 1967 there were no significant exchange crises, even as the rate of inflation grew in the 1970s. The major reason for this in the first part of the 1970s was the dramatic increase in the value of non-traditional or " m i n o r " exports. As is shown in Table 5, they doubled in value from 1967 to 1971. Domestic inducements established by previous policy efforts and particularly by those offered along with the devaluation measure were one major explanation for this growth. Subsequently, the rise in coffee prices and a dramatic expansion of drug trafficking played major roles in improving the country's foreign exchange situation. In addition, this different staging of industrialization in Colombia was aided by the fact that the above policy shifts were enacted at the same time that a number of the new exports benefitted from a booming world economy, from international trade conditions. Subsequent to the Lleras government other moves consonant with a transition to more orthodox policies such as reductions of tariff levels on some goods, increased lifting of import restrictions, and the raising of interest rates have occurred. A greater consciousness of the costs of many import substitution schemes also appears to have developed in the post-1967 era. (Diaz-Alejandro, 1976: 230). Overall, however, these other moves have been quite modest. By the mid-1970s, under the impact of

A Case Study of Colombia

53

the coffee boom and illegal drug trafficking, the country's steady rate of devaluation, a central element in export promotion, was floundering (Morawetz, 1980). Thus, the dramatic shift portended by the 1967 law and its subsequent reforms and their immediate effects have been partially mitigated by more recent changes in national policy in response to changing trade circumstances. Unlike some of its continental neighbors, Colombia has not shifted to extreme forms of economic liberalism. Taken as a whole, the policies have represented "an incomplete transition from secondary import to export substitution, with a good deal of export promotion 'on top of' a relatively unimpaired structure of secondary import substitution" (Ranis et al., 1977: 8). The contrast between the Colombian case and the case of its continental neighbors that succumbed to bureaucratic-authoritarian coups is significant. In Colombia a shift to more orthodox policies encouraging realistic exchange rates and export promotion occurred without the need for a repressive political regime. Colombia, was able to mitigate or avoid several of the major economic problems identified in the bureaucratic-authoritarian model. Pressures on Colombia's balance of payments, inflation and the bitter economic conflict, and potentially explosive political strife associated with sharp devaluations were all experienced to a much lower extent than they were in many of the more industrialized Latin American countries. SUMMARY AND CONCLUDING REFLECTIONS

This review of the Colombian case provides evidence that the existence of Colombia's open political regime can be explained by how the country's pattern of industrialization has evolved and by its experiences of initial popular sector incorporation. It also shows how the evolution of these two elements differed in important ways from the pattern posited by the original bureaucratic-authoritarian model. The Colombian case does not support either culturally or economically deterministic views that overlook past historical differences or views that place excessive weight on a unidirectional impact of the current international system. In addition, it points to three factors--the initial export structure, the differential impact of the international system, and the political party structure--that appear to have contributed significantly to how the pattern of Colombia's political regime type evolved. These factors should be considered in a comprehensive theory linking different capitalist economic growth strategies with political regime types.

54

Studies in Comparative International Development / Spring 1984

In Colombia the central element of one of these factors, the initial export structure, is the coffee sector. Internally, the coffee sector never experienced high rates of state expropriation of its surplus, in part because--unlike Brazil--there were many small coffee-growing landholdings. Since coffee was not an important commodity for domestic consumption (although efforts were made to expand this consumption), the conflicts and tensions that arose in countries such as Argentina and Uruguay between small groups of exporters, lobbying for high prices and incentives to export, and urban consumer groups, seeking increased consumption at lower prices, did not ensue. Furthermore, the international system did not weaken major coffee growers and merchants, but assured that they would continue to play a central role in the country's economic and political life. Colombia was fortunte that Brazil, the leading coffee producer, led the fight and paid much of the cost of trying to keep coffee prices high in the 1920s and 1930s. The fact that initial labor incorporation came at a time (the 1930s) when the industrial base was very small precluded any effective alliance between workers and industrialists at that time "against" the agro-export sector. There was little sectorat conflict, for common interests between early industrialists and the coffee sector were extensive, and the management of internal economic policy during the 1930s was to their mutual benefit. In particular, the devaluations in this period favored both coffee exporters as well as those industries which were already installed in the country and which expanded in those years. Coffee growers and industrialists were found within both major political parties. If Lrpez's populism in the 1930s was short-lived, Rojas' in the 1950s was stillborn for both internal and external reasons. The central reason for Rojas' difficulties was clearly the existing political party structure, but the external environment contributed to his downfall. Rojas' attempt to emulate Perrn came at a period when the Argentine faced wide opposition from domestic and international business and commercial elites, international labor organizations, and the Church. Finally, in the more recent period the impact of the international system on Colombia also helped sustain the civilian political regime. It is true that in the 1950s and 1960s, as Colombia pursued import-substituting industrialization, it too found itself constrained by the import intensity of the process and suffered from lack of foreign exchange and balance of payments problems. The regime, however, received important international support. Its problems were eased at critical moments by funds from the United States Alliance for Progress, estab-

A Case Study of Colombia

55

lished in reaction to the Cuban revolution. Import-substituting policies had not proceeded as far and, structures inimical to exporting such as heavy protectionism or the creation of inefficient firms had not become solidified, when a shift to moderate export promotion occurred. In this instant the country was somewhat fortunate, because the 1967 policy package of gradual or mini-devaluations and export incentives resulted from a political reaction against a more traditional set of measures advocated by the IMF. The resulting export diversification both in agricultural commodities other than coffee and in some industrial lines eased Colombia's foreign exchange problems until the mid-1970s surge in coffee and illegal drugs. This different staging in the industrial process was aided not only by these important domestic policy shifts, but also by a change in the international economic environment, a boom in world trade. A more critical eye was also being cast on import-substitution across the continent, based in part on the experiences of the Southern Cone countries. Unlike many of these countries, Colombia did not have to dismantle or "shake out" many over-protected inefficient firms. As a result, the political tensions of the economic transition were easier to manage. The third factor, the crucial role of the political party structure in channelling and co-opting political and economic demands in Colombia has been deliberately understressed to focus attention on the importance of the country's initial export structure and on aspects of the international system which have had a favorable impact on the formation and survival of the current limited democratic regime of Colombia. The dominance of the country's political life to date by the two political parties founded in the 19th century is something quite unique to Colombia. The central role of the political party structure in controlling the early organizational history of the working class and in "managing" populism under Rojas is undeniable. This factor and the importance of such occurrences as Brazil's role in maintaining coffee prices in the 1920s and 1930s, foreign assistance in the 1960s, and the policy shift toward export promotion in the late 1960s all suggest that Colombia's experience with democracy is not one easily replicable elsewhere. Some of these variables are not easily amenable to manipulation by policy makers. Other variables such as agreements to limit or circumscribe political conflict and the staging of industrialization might be manipulated under certain circumstances. In addition, there may well be other cases in which because of special circumstances the international system has favored or will favor less authoritarian regimes in the Third World.

56

Studies in Comparative International Development / Spring 1984 NOTES

1 would like to thank Alfred Stepan, Daniel Levy, Jeffrey Freyman, Samuel Morley, Kenneth Betsalel, and an anonymous reviewer for comments on earlier versions of this article. I gratefully acknowledge the financial assistance of the Danforth Foundation, the Social Science Research Council and American Council of Learned Societies, and the Doherty Foundation during the period in which the research for this article was carried out. Responsibility for its contents is mine alone. 1. As it finally emerged, the National Front agreement stipulated that from 1958 to 1974 the presidency would alternate between members of the two parties; all cabinet officers, legislative posts, and government and public employment positions were to be divided equally between the two parties and then proportionately within each party to the different factions. In addition, most measures required a two-thirds votes in Congress to pass. In 1968 this last measure was changed to one of simple majority. At that time it was decided that national elections would be open to all parties beginning in 1974, and the parity agreement in public employment would extend until 1978. From 1978 on, the majority party was required to give adequate representation to the party, receiving the second highest number of votes. Elsewhere, 1 have argued that Colombia is best considered a limited democratic consociational regime. (See Hartlyn, 1981: 367-370; Wilde 1978). In contrast to the bureaucratic-authoritarian regimes, its politically more liberal regime is unquestionable. 2. Palacios (1979: 404) estimated that in 1933 approximately 47 percent of Colombian coffee export was conducted by foreign firms. Thereafter, there was a relatively steady decline: 35 percent in 1940-41,28 percent in 1944, 18 percent in 1954, 15 percent in 1961, 18 percent in 1966, and 12 percent in 1970. 3. Colombian loans floated on the New York bond market grew from US$20.4 million in the 1920-25 period to US$215.4 million in the 1926-28 period. Foreign investors were confident about Colombia because of both the United States compensation payment over Panama and the establishment of the Banco de la Republica (Central Bank), following the Kemmerer mission (see McGreevey, 1964: 204-5). 4. A useful treatment of the Brazilian case which traces how Vargas' more controlled "paternalistic-administrative" populism led to a more "radical populist" politics is found in Erickson (1977: 49-93). For a further discussion of the Argentine case, see O'Donnell (1972:115-99). 5. Bird (1970:211-3, 218), in addition to the above goal, identifies two other goals of policymakers during this period: maintenance of real income of coffee growers (and here the fact that so many small and medium landowners formed part of this group was helpful to coffee merchants and large growers) and the reduction of the Coffee Federation's inflationary borrowing from the Banco de la Reptlblica to purchase the domestic coffee crop (which had acute effects particularly in 1963 and 1964). These somewhat contradictory goals have played a role in creating Colombia's confusing tax structure and its elaborate exchange-rate system prior to the 1967 reforms. REFERENCES AMI~ZQUITA, SAUL and JAVIER FERN,h2qDEZ 1977 "La economla colombiana, 1950-1975." Revista de Planeaci6n y Desarrollo, 9: 1-278. ARANGO, JUAN IGNACIO 1978 Inversi6n extranjera en la industria manufacturera national. Bogot,-i: Departamento Administrativo Nacional de Estadistica. AVRAMOVIC, D. 1972 Economic Growth of Colombia. Baltimore: The Johns Hopkins University Press. Bdlr (Banco de la Republica) 1976 Revista del Banco de la Republica. (Nov.) 1979 Revista del Banco de la Republica. (Dec.).

A Case Study of Colombia

57

BENNETT, DOUGLAS and KENNETH SHARPE 1982 "Capitalism, bureaucratic authoritarianism, and prospects for democracy in the United States." International Organization 36: 633-663. BERGSMAN, JOEL 1970 Brazil:Industrialization and Trade Policies. London: Oxford University Press. BERRY, R. ALBERT 1980 "The National Front and Colombia's Economic Development." In R. Albert Berry, Ronald G. Hellman and Mauricio Solaun (eds.), Politics of Compromise: Coalition Government in Colombia. New Brunswick, N.J.: Transaction. BIRD, RICHARD M. 1970 Taxation and Development: Lessons from Colombian Experience. Cambridge: Cambridge University Press. CAVAROZZI, MARCELO 1982 "Military Retreat and Impediments to Democratization in Argentina." Paper presented to the LASA Conference. COLLIER, DAVID 1979 The New Authoritarianism in Latin America. Princeton: Princeton University Press. 1979a "'Overview of the Bureaucratic-Authoritarian Model." In David Collier (ed.), The New Authoritarianism in Latin America. Princeton: Princeton University Press. 1979b "The Bureaucratic-Authoritarian Model: Synthesis and Priorities for Future Research." In David Collier (ed.), The New Authoritarianism in Latin America. Princeton: Princeton University Press. Colombia, Fondo de Promoci6n de Exportaciones 1977 "Breve resefia sobre los estimulos tributarios dentro de la politica de fomento a las exportaciones." Bogota. Unpublished. DiAZ-ALEJANDRO, CARLOS F, 1976 Foreign Trade Regimes and Economic Development: Colombia. New York: Columbia University Press. DIX, ROBERT H. 1967 Colombia: The Political Demensions of Change. New Haven: Yale University Press. ERICKSON, KENNETH PAUL 1977 The Brazilian Corporative State and Working-Class Politics. Berkeley: University of California Press. EVANS, PETER 1979 Dependent Development: The Alliance of Multinational, State and Local Capital in Brazil. Princeton: Princeton University Press. FLUHARTY, VERNON LEE 1957 Dance of the Millions: Military Rule and the Social Revolution in Colombia, 1930-1956. Pittsburgh: University of Pittsburgh Press. FURTADO, CELSO 1970 Economic Development of Latin America. Cambridge: Cambridge University Press. GEREFFI, GARY and PETER EVANS 198I "Transnational Corporations, Dependent Development and State Policy in the Semiperiphery: A Comparison of Brazil and Mexico." Latin American Research Review 16 (3): 31-64, GUTII~RREZ, ALFREDO, et al. Uruguay: Economic Memorandum. Washington, D.C. The World Bank. 1979 HARTLYN, JONATHAN "Consociational Politics in Colombia: Confrontation and Accommodation in 1981 Comparative Perspective." Ph.D. diss., Yale University. "Interest Associations and Political Conflict in Colombia/' Unpublished. 1982 "Colombia: Old Problems, New Opportunities." Current History 82: 62-65, 83-4. 1983 HEWLETT, SYLVIA A. "Human Rights and Economic Realities in Developing Nations." In Tom Farer 1979 (ed.), The Future of Inter-American System. New York: Praeger.

58

Studies in Comparative International Development / Spring 1984

HIRSCHMAN, ALBERT O. 1979 "The Turn to Authoritarianism in Latin America and the Search for its Economic Determinants." In David Collier (ed.), The New Authoritarianism in Latin America. Princeton: Princeton University Press. 1981 " A Generalized Linkage Approach to Development, with Special Reference to Staples." In Essays in Trespassing: Economics to Politics and Beyond. Cambridge: Cambridge University Press. HUTCHENSON, THOMAS 1973 "Incentives for Industrialization in Colombia." Ph.D. diss., University of Michigan. IMF (International Monetary Fund) 1973 International Financial Statistics, 26 (Sept.). 1977 International Financial Statistics, 30 (Dec.). 1979 International Financial Statistics, 32 (Nov.). 1982 International Financial Statistics, 35 (June). KALMANOVITZ, SALOMON 1978 Desarrollo de la agricultura en Colombia. Bogota: Editorial La Can-eta. KAUFMAN, ROBERT R. 1979 "Industrial Change and Authoritarian Rule in Latin America: A Concrete Review of the Bureaucratic-Authoritarian Model." In David Collier (ed.), The New Authoritarianism in Latin America. Princeton: Princeton University Press. LEFF, NATHANIEL H. 1968 The Brazilian Capital Goods Industry, 1929-1964. Cambridge, MA.: Harvard University Press. LITTLE, IAN, TIBOR SCITOVSKY, and MAURICE S c o T r 1970 Industry and Trade in Some Developing Countries: A Comparative Study. London: Oxford University Press. MCGREEVEY, WILLIAM P. 1964 "Statistical Series on the Colombian Economy." Dept. of Geography, University of California, Berkeley. 1971 An Economic History of Colombia 1845-1930. New York: Cambridge University Press. MACHADO, ABSALON 1977 El cafr: de la aparceria al capitalismo. Bogot,-i: Punta de Lanza. MATTER, KONRAD 1977 Inversiones extranjeras en la economia colombiana. Medellin: Ediciones Hombre Nuevo. MAULLIN, RICHARD L. 1967 "'The Colombia-IMF Disagreement of November-December 1966: An interpretation of its place in Colombian politics." Santa Monica: The RAND Corporation, Memorandum RM-5314-RC. MISAS A., GABRlEL 1975 Contribucidn al estudio del grado de concentracidn en la industria colombiana. Bogota: Ediciones tiempo presente. MOLINA, GERARDO 1977 Las Ideas Liberales en Colombia de 1935 a la Iniciacidn del Frente Nacional. Tomo 1II. Bogot,4: Ediciones Tercer Mundo. MONCAYO, VICTOR MANUEL and FERNANDO ROJAS 1978 Luchas obreras y polltica laboral en Colombia: BogoN! La Carreta. MORAWETZ, DAVID 1980 "Why the Emperior's New Clothes Are Not Made in Colombia." World Bank Working Paper No. 368. Washington, D.C.: World Bank. NELSON, RICHARD R., T. PAUL SCHULTZ, and ROBERT L. SLIGHTON 1971 Structural Change in a Developing Economy: Colombia's Problems and Prospects. Princeton University Press. O'DONNELL, GUILLERMO 1973 Modernization and Bureaucratic-Authoritarianism: Studies in South American Politics. Berkeley: Institute of International Studies, University of California, Politics of Modernization Series No. 9.

A Case Study of Colombia 1977

59

"Estado y alianzas en la Argentina, 1956-1976." Desarrollo Econ6mico 16: 523-54. OQUIST, PAUL 1978 Violencia, Conflicto y Politica en Colombia. Bogota: lnstituto de Estudios Colombianos. PALACIOS, MARCO 1979 E1 Caf6 en Colombia (1850-1970): Una Historia Econ6mica, Social y Politica. Bogot~i: Editorial Presencia y FEDESARROLLO. PARSONS, JAMES J. 1968 Antioquefio Colonization in Western Colombia. Revised Edition. Berkeley: University of California Press. Partido Communista de Colombia 1960 Treinta Afios de Lucha del Partido Communista de Colombia. Bogot~i: Ediciones Paz Y Socialismo. PECAUT, DANIEL 1973 Politica y sindicalismo en Colombia. Bogota: Editorial la Carreta. POLLOCK, JOHN C. 1975 "'Violence, Politics and Elite Performance: The Political Sociology of La Violencia in Colombia." Studies in Comparative International Development 10: 22-50. POVEDA RAMOS, GABRIEL 1967 -Antecedentes y Desarrollo de la Industria en Colombia." Unpublished. 1976 Politicas econ6micas, desarrollo industrial y tecnologia en Colombia 1925-1975. Bogota: Editora Guadalupe, Colciencias. RANIS, GUSTAV, et al. 1977 "Income Distribution and Growth in Colombia.'" Presented to the Conference on Distribution, Poverty and Development, Bogota. ROJAS RUIZ, HUMBERTO 1970 "El frente nacional: Soluci6n politica a un problema de desarrollo? In Rodrigo Parra S. (ed.), La dependencia externa y el desarrollo politico de Colombia. Bogota: Universidad Nacional de Colombia, Direcci6n de Divulgaci6n Cultural. SERRA, JOSt~ 1979 "Three Mistaken Theses Regarding the Connection Between Industrialization and Authoritarian Regimes." In David Collier, (ed.), The New Authoritarianism in Latin America. Princeton: Princeton University Press. THOUMI, FRANCISCO 1980 "Industrial Development Policies During the National Front Years." In R. Albert Berry, Ronald G. Hellman and Mauricio Solaun, (eds.), Politics of Compromise: Coalition Government in Colombia. New Brunswick, N.J.: Transaction. UN/ECLA (United Nations, Economic Commission on Latin America) 1957 Analyses and Projections of Economic Development, Vol. 3, The Economic Development of Colombia. Geneva: United Nations. U.S. Senate, Subcommittee on American Republic Affairs, Committee on Foreign Relations, 91st Congress, 1st Session 1969 "Colombia--A Case History of U.S. Aid." Washington, D.C.: U.S. Government Printing Office. URRUTIA M., MIGUEL and MARIO ARRUBLA (eds.) 1970 Compendio de Estadisticas Hist6ricas de Colombia. Bogota: Universidad Nacional de Colombia, Direcci6n de Divulgaci6n Cultural. URRUTIA M., MIGUEL 1969 The Development of the Colombian Labor Movement. New Haven: Yale University Press. WIARDA, HOWARD 1980 The Continuing Struggle for Democracy in Latin America. Boulder: Westview Press. WICKIZER, V.D. 1971 Coffee, Tea and Cocoa: An Economic and Political Analysis. Stanford: Stanford University Press. WIESNER D., EDUARDO 1978 "Devaluaci6n y mecanismo de ajuste en Colombia." Banca y Finanzas No. 159: 43-123.

60

Studies in Comparative International Development / Spring 1984

WILDE, ALEXANDER 1978 "Conversations Among Gentlemen: Oligarchical Democracy in Colombia." In Juan Linz and Alfred Stepan, (eds.), The Breakdown of Democratic Regimes: Latin America. Baltimore: The Johns Hopkins University Press. WOGART, JAN PETER 1978 Industrialization in Colombia: Policies, Patterns, Perspectives. Tubingen: Mohr, Kieler Studien No. 153. World Bank 1980 World Development Report. Washington, D.C.

SOCIOLOGISTS STUDY-VACATION TO THE SOVIET UNION HOSTED BY THE SOVIET CENTRAL COUNCIL OF TRADE UNIONS

June 29-July 15, 1984 Moscow Leningrad Kishinev

$1845" MEET VISIT DISCUSS ENJOY INCLUDES

includes everything social scientists, journalists, party and trade union officials, peace and women's groups research institutes, workplaces, factories, collective farms, children's camp, rest home social problems, the media, human rights, economic planning cultural events and historic sights all meals, hotel, airfare and visits Counterpart Tours

250 West 57th Street, Suite 1428 New York, N.Y. 10107 (212) 245-7501 (800) 223-1336 *This trip may be tax deductible Co-Sponsoredby New Political Scienceand LaborResearchAssociation