Limits of Power: Elections Under the Conservative Hegemony in Colombia, 1886-1930

ON February 9, 1930, a Liberal, Enrique Olaya Herrera, defeated at the polls a divided Conservative Party in Colombia, bringing down a regime that had been in power since 1886, the so-called Hegemonía Conservadora, or Conservative Hegemony.¹ Olaya Herrera’s victory was an extraordinary event in itself. The transfer of power from Conservative to Liberal rule that followed was all the more remarkable considering that after 1930, several Latin American governments in quick succession were forcibly overthrown amid the political turmoil and social discontent of the worldwide depression.²

It was not the first time in Colombia an opposition party had come to power through elections.³ Yet because the Liberals in opposition had been unable to reach the presidency since 1886, and because they were largely unrepresented in Congress, particularly before 1904, the perception persisted that the Conservative regime was electorally unassailable. “The party in power here always carries the elections,” wrote the U.S. minister in Bogotá in 1891.⁴ In 1912 an anonymous statement, attributed to “one of the greatest Liberal leaders,” expressed the view that “a suffocating centralization is the basis on which the whole machinery of government runs. As a consequence of this condition and the frauds constantly practiced, it is very difficult or rather impossible to have fair electoral laws enacted.”⁵

Once in power, after 1930, the Liberals passed new electoral legislation and reintroduced universal male suffrage while claiming that theirs was the party of clean elections. In a 1936 message to Congress, President Alfonso López Pumarejo distinguished the new regime from the previous one: the Conservative Hegemony was upheld by the use of force and electoral fraud; the Liberals had suffered a half-century of electoral arbitrariness.⁶

The Conservative Hegemony, particularly the last two decades of its rule, remains one of the relatively less studied periods in Colombia’s political history.⁷ The few existing works covering the whole period either pay little attention to the electoral process or, when dealing with elections, tend to focus on the anomalies of the system, echoing Liberal complaints. The latter is, for example, the general line of interpretation that emerges from the brief but influential essay by Jorge Orlando Melo, “La república conservadora,” originally published in 1975.⁸ According to Melo, the Liberal opposition was left without “possibilities of reaching power through electoral means” after 1886. While he acknowledges some of the changes introduced into the political system after 1910, Melo emphasizes the electoral manipulation exercised by those who controlled the executive, the influence of the clergy in curtailing the Liberal vote, and the power of Conservative landlords over rural peons. Politics remained stagnated; the Conservative Republic was supported by “electoral coercion.”⁹ Still, Melo is not directly concerned with the electoral process. As with most overviews of the period, his picture of elections is impressionistic, although influenced by the hitherto dominant Liberal interpretation.¹⁰

Closer examinations of elections during the Conservative Hegemony are indeed exceptional. In a detailed analysis of the 1897 presidential campaign, Charles Bergquist highlights the fundamental differences between Liberals and Conservatives, who “represented divergent economic interests.” In addition, he suggests, after 1910, elections “were held in relative calm and freedom”; the 1910 reforms “assured meaningful representation to both traditional parties in the legislative bodies of the nation.”¹¹ From a different perspective, Helen Delpar pays greater attention to issues of partisan organization while showing that Liberal electioneering was vigorous during the last quarter of the nineteenth century.¹² Neither Bergquist nor Delpar covers the whole period, however; the crucial two decades after 1910 are still largely untouched. Although their different analyses suggest meaningful ways of interpreting electoral competition, moreover, they both remain far from tackling the development of electoral practices in a systematic fashion.

Other, more recent works also shed light on some of the complexities involved. Patricia Pinzón de Lewin makes interesting observations regarding the army’s limited role in elections, while Medófilo Medina shows how a divided church was an inefficient electoral agent in 1930.¹³ The work of Malcolm Deas is also suggestive. He has stressed the central role of electoral politics, and his picture extends beyond stereotypes to alternative lines of enquiry; for example, the growth of genuine participation, even of those officially excluded from the franchise; and the relative weakness of both the central government and the Conservative Party leadership in dealing with local politicians. Deas also suggests a degree of cautiousness and skepticism when interpreting Liberal anticlerical rhetoric, and the need to look more carefully into the role of the opposition.¹⁴

In spite of all these valuable contributions, the subject of elections remains understudied. To what extent and under what circumstances the Conservative regime was able to manipulate the electoral process is a question that merits further consideration. The traditional Liberal interpretation of elections under the Conservative Hegemony, moreover, still looms large in the literature, as reflected in recent essays by both historians and political scientists.¹⁵

The purpose of this article is therefore to reassess the nature and meaning of Colombian elections between 1886 and 1930. It does not aim to cover all aspects of elections; rather, it concentrates on a set of questions related to the central issue of electoral competition.¹⁶ How decisive was governmental influence in determining electoral outcomes? What was the opposition’s degree of involvement in electoral politics? Did elections encourage political mobilization and awareness? Above all, how competitive were elections; how much were they genuine contests and how much mere charades, reinforcing social control? Taking a fresh look at the competitiveness of Colombian elections, this essay argues that many were indeed significant contests whose various aspects deserve further attention from scholars. Many questions concerning, for example, the issues debated during the campaigns and the interests of the contending parties and their candidates remain open for future research.

The history of Colombian elections in general, not just those held under the Conservative Hegemony, has been largely neglected. That neglect has long been noted by historians such as David Bushnell and Malcolm Deas, who have repeatedly drawn attention to the importance of the topic.¹⁷ Without denying the anomalies of the system, both of them have emphasized the early expansion of Colombian suffrage, the relatively high participation in certain periods, the intensity of competition, and the long-term impact of frequent electioneering. By readdressing the meaning of elections between 1886 and 1930, this essay also aims to suggest the need to incorporate more fully the history of elections in any broader study of the development of Colombian political culture. Colombian electoral traditions, strongly rooted in the early republic, not only persisted during the Conservative Hegemony but were reinforced by intense electioneering, and by a commitment to suffrage that grew to involve substantial sectors of Colombian society.

The Colombian case also contributes to the study of electoral history elsewhere. The meaning of Latin American elections in general is still an obscure topic, despite some recent scholarly advances.¹⁸ The assumption that these elections were mere theatrical events, in which those in power paraded flocks of unaware and indifferent voters to the polls without much difficulty, runs deep in the historiography.¹⁹ Until recently, the same assumption has also prevailed in the study of other unreformed electorates, including that of England. Frank O’Gorman, for one, has challenged such traditional interpretations of electoral life before the Reform Act of 1832 and has found new ways of examining the nature of political control amid habits of widespread electoral involvement —approaches from which students of other countries still have much to learn.²⁰

The period covered here has great significance in the history of suffrage in both Europe and the Americas. Male universal suffrage was generally adopted in Western Europe, for example; but in some countries, such as Italy and Spain, these years were the golden age of caciquismo and electoral management. In the United States, this was also the golden age of machine politics, though it witnessed the challenge of Progressivism as well.²¹ Similarly, the coronéis, or local political bosses, of Brazil’s Old Republic delivered voters for the most powerful party, while the Argentines experienced the impact of electoral reform after the 1912 Saenz Peña Law. Although this essay deals almost exclusively with Colombia, it was written in the belief that electoral practices and behavior ought to be properly considered in the context of comparative history.

Continuities and Changes Under the Conservative Hegemony

It may be useful to start by clarifying the nature of the Conservative Hegemony, a somewhat misleading term. It was certainly a “conservative” regime that was installed in power in 1886: “conservative” by the very definition of a centralist constitution, which restored the privileges of the Catholic church and strengthened the authority of the state, while “liberty” lost priority to “order.” Although the regime brought the Conservative Party back into the national government, the Conservatives returned to power indirectly, in a process led by a dissident Liberal, Rafael Núñez. Núñez sought to amend the Liberal, ultrafederalist Constitution of 1863 and implement a positivist program of order and progress, known as the Regeneración²² As the leader of his own movement, the Independientes, and as president of the republic for the second time since 1884, Núñez had quelled a Liberal rebellion in 1885 with the Conservatives’ support. Thus the regime that emerged out of a civil war in 1886 was originally based on a coalition between dissident Liberals (Independientes) and Conservatives.

As the Conservatives gained positions of power, some of those who had embraced the dissident cause returned to the mainstream of the Liberal Party. A few remained loyal Nuñistas nevertheless and, together with leading Conservatives, such as Miguel Antonio Caro, attempted to create the Partido Nacional, while disaffected Conservatives claimed for themselves the name Históricos. The division between Nacionales and Históricos was muted in the face of another Liberal revolution in 1899, while the Independientes as such vanished altogether by the turn of the century. Yet the Liberal-Conservative dichotomy was again challenged during the second decade of the twentieth century, when a new movement, the Unión Republicana, at one point succeeded in electing a president, Carlos E. Restrepo.²³

To appreciate the nature of electoral life under the Conservative Hegemony, therefore, the idea of “party competition” first needs to be clarified. Given the loose political organizations that existed, the Liberal-Conservative division —if, indeed, it was valid in the long run —needs to be complimented by a picture of the rival factions.²⁴ These factions, which at times attempted to become independent parties, represented alternatives in the struggle for power during elections. In addition, a newly founded Socialist Party also entered the electoral arena.²⁵

Politics thus did not remain static. Order —the “scientific peace” sought by the leaders of the Regeneración — supposed, in principle, a degree of political demobilization, as reflected, for example, in a less intense electoral calendar. Caro and his fellow Conservatives, however, found it necessary to excite partisan feelings to counterbalance a surviving active Liberalism.²⁶ All hopes for peace were ultimately frustrated by the Guerra de los Mil Días (War of the Thousand Days, 1899-1902), which ended in dismemberment of the country’s territory with the loss of Panamá. The war also generated an extraordinary level of political mobilization, leaving behind a legacy of popular Liberalism that would later reemerge, with radical undertones, in regions such as the Santanders and the Upper Magdalena.²⁷

Paradoxically, the war’s immediate wake brought political accommodation. President Rafael Reyes (1904-1910) incorporated Liberals into his cabinet and pursued constitutional reforms, though these were later threatened by his own dictatorial aspirations. The civilian movement that helped to remove Reyes from power, the Unión Republicana, passed significant reforms in 1910, including laws that guaranteed minority representation for the Liberal Party. Led by President Carlos E. Restrepo (1910-1914), the Unión Republicanas primacy was short-lived, and was followed by the return of mainstream Conservatism with the election of José Vicente Concha (1914-1918). Marco Palacios refers to the period 1914-1930 as the “Conservative Hegemony” proper. Rut even in this case, as Palacios himself observes, “the term hides the weakness of governments in the face of Congress, the complexities of their relations with the church, and their adaptation to liberal and capitalist values.”²⁸

If politics did not stagnate, socioeconomic changes were even more visible. Economic growth went hand in hand with the expansion of the coffee trade: from a relatively slow period during the last quarter of the nineteenth century, coffee exports grew rapidly after 1903. Oil and bananas added to an export sector that, in spite of its limitations, had significant effects on other areas of the economy. Industries that could trace their origins to the end of the 1870s, following the earlier tobacco boom, now expanded into cities such as Bogotá, Medellin, and Barranquilla. They further benefited from the protectionist measures taken by the Reyes administration. The cattle industry on the Atlantic coast grew to meet an increasing demand for beef in some Andean departments.

Imports also grew, and customs revenues increased, allowing for the improvement of infrastructure: railways, roads, and port facilities. This picture, however, does not represent steady progress. The slow recovery after the 1880s was followed by a severe depression, accompanied by monetary crisis at the end of the century. After the Guerra de los Mil Días, however, Colombia enjoyed a period of unprecedented prosperity, which reached a peak in the 1920s, when the country received $25 million from the United States as an indemnity for the loss of Panama. In addition, the reorganization of national finances following the establishment of the Banco de la República in 1923 allowed the country to enter the international capital markets in an extraordinary fashion. Enormous sums in foreign loans were contracted between 1923 and 1928.²⁹

To what extent these changes affected politics, and electoral practices in particular, is not easy to tell, and is beyond the scope of this article. As Deas has observed, “economic and social changes in this half-century in some ways had less obvious political effect than one might expect from their magnitude.”³⁰ Some suggestions are plausible, nevertheless. Economic growth allowed for larger fiscal resources, and therefore gave more opportunity for clientelistic politics, which relied on the control of state resources.³¹ The size of the state remained relatively small, however, in an economy that was, by and large, in private hands. Above all, economic growth produced geographical mobility, which was most evident in the colonization of the new coffee frontier but also in the banana zone of northern Colombia and in oil exploration areas. It was also evident in the growth of major ports and cities, where modern industries developed.

How this mobility affected the relations between politicians and the electorate has yet to be properly studied. The relevant questions include where and why deferential attitudes prevailed, where and how a more modern clientelism replaced old patronage bonds, in what relationships venality flourished, and in which areas traditional partisan loyalties held firm or gave way to new forms of political mobilization. In any case, as this essay will show, parties (like politics in general) did not stand still. An electorate expanding more rapidly than the population had to be won over. Party rallies, especially during heavily contested elections, grew more numerous. So did campaign tours, particularly after 1910, helped by transportation improvements. By 1930, politicians were making good use of the airplane for campaign purposes.³² Political communication was also affected by the introduction of the radio and the modernization of the press.³³

This gradual adaptation to changing circumstances did not lack conflicts. Social unrest was manifest in areas of recent settlement or rapid growth, such as the banana zone or major ports and cities.³⁴ An emergent working class was steered by Socialists and anarchists, whose activities are vividly recorded in the memoirs of the union leader Ignacio Torres Giraldo. But as Torres Giraldo himself acknowledged, the leadership was working with an “electorate on loan”; in Magdalena, for example, where the most serious strike took place in 1928, “revolutionary socialism was a Liberal stimulant.”³⁵

Liberal discourse, through popular leaders such as Rafael Uribe Uribe, Benjamín Herrera, and later Jorge E. Gaitán and Alfonso López Pumarejo, consciously targeted the labor movement. Whatever gains the socialists achieved were lost to the Liberals, first in the 1922 elections and later in the 1930 campaign.³⁶ Conservative discourse also addressed some laborers’ concerns, albeit far more timidly.³⁷ Colombian politics was still characterized by the multiclass nature of its parties. Their social outlook did not differ much; they recruited most of their activists and some of their leaders from those middle orders described by Rufino Gutiérrez.³⁸ Nevertheless, Liberal electoral gains became more obvious in major cities and in areas of large labor concentration, such as the banana zone, while Conservatives preserved some control over traditional rural areas, with the clergy’s support. Indeed, relations with the church remained one of the most divisive issues between Liberals and Conservatives.

Socioeconomic and political changes were, of course, uneven in a country characterized by regional diversity. The consolidation of the new coffee frontier in western Colombia marked the rise of Antioqueño politicians to national power, best symbolized by the elections of presidents Restrepo, Marco Fidel Suárez (1918-21), and Pedro Nel Ospina (1922-26).³⁹ After the opening of the Panama Canal, the traditional orientation of the export economy toward the Caribbean ports was diverted to the Pacific. This new integration of the national economy caused some resentment, encouraging a degree of regionalism, particularly in the Santanders and the northern departments.⁴⁰ Fears spread in Bogotá that these provinces would secede after the formation of the Liga Costeña in 1919; but as the 1922 election results showed, partisan bonds were stronger than regional feelings. President Suárez also feared separatist sentiment in Valle and Cauca; these departments, he believed, resented the Antioqueño monopoly on the presidency.⁴¹

Electoral behavior differed from region to region. Clerical influence over electoral politics, for example, was stronger in Boyacá than in Magdalena. But even in Boyacá, where anticlericalism was also stronger, the influence of the clergy varied from locality to locality. The paucity of research does not allow, at this stage, for generalizations about regional variations in electoral practices. Regional developments, moreover, took place within a national framework, shaped by the dominant political culture of Bogotá.⁴² The analysis that follows aims to offer a national perspective on electoral politics, acknowledging that further research on local politics is necessary to complete an already complex picture.

Voters in an Unreformed Electoral System

The regime that came to power in 1886 centralized the electoral system. Each of the nine states that had formed the union since 1863 had hitherto given themselves independent and diverse electoral laws, such as those regulating the qualification of voters. New legislation established two categories of voters and, for some posts, introduced a two-tier system of elections. All citizens—that is, every adult male, with the exception of vagrants — had the right to vote for concejeros municipales and diputados, municipal councilors and deputies in departmental assemblies.⁴³ Literacy and property requirements (a yearly income of $500 or an estate valued at $1,500) were established for voting to elect representatives for the lower chamber. This same restriction was imposed on voting for electores—those who, in turn, elected the president of the republic. Finally, senators were elected by the diputados.

The electoral system did not go unchallenged. Pressures for change, particularly from the Liberal Party, were geared toward measures to guarantee fair electoral procedures. In 1898, Miguel Samper singled out the opposition’s main desires: an independent body to supervise elections, representation of minorities, and effective guarantees for the exercise of suffrage.⁴⁴ All attempts to secure such measures had failed by mid-1899; and Liberal frustration with the lack of electoral reform was one cause of the outbreak of a civil conflict that lasted more than three years.⁴⁵

An agreement was reached only after that war, when the Reyes administration called for a constitutional assembly; this body, in turn, secured representation of minorities through a system of limited voting in multimember constituencies. This reform was not implemented until 1910, when other significant measures were also introduced: direct presidential elections, reduction of income and property requirements for qualified electors (three hundred pesos and one thousand pesos, respectively), and the removal of the executive’s right to appoint electoral juries. Pressures for change did not stop there. At least ten major electoral reform projects —aimed mostly at a fairer electoral procedure—were discussed in Congress over the following two decades.⁴⁶ No proposal received more attention, however, than the introduction of proportional representation, which Congress finally passed in 1929.

Nearly absent from the discussion of reforms was the extension of suffrage.⁴⁷ Yet this issue had caused one of the most heated debates during the constitutional assembly in 1886; Miguel Antonio Caro, the Conservative architect of the constitution, who advocated extending the franchise, was accused of demagoguery. His opponents, mostly the Independientes led by José María Samper, considered previous experience with universal suffrage in Colombia to have been disastrous, the “origin of all the wrongdoings of the social order.” An extended suffrage, according to Samper, only favored the “aristocracy of the gamonales [caciques].”⁴⁸ In the end, Samper’s opinion prevailed, although an open franchise, as already noted, was approved for elections of concejales and diputados.

It would be wrong to assume that the literacy, income, and property restrictions imposed on the franchise excluded the popular sectors from the electoral process. For one thing, male universal suffrage was accepted for local elections. Manuel Serrano Blanco may have exaggerated when he said that these elections raised the “most vehemence and political passion” among the populace in Bucaramanga, but their importance is not in doubt. The opposition parties, including the emerging Socialists in the 1920s, usually participated in municipal elections, even when they routinely abstained from national contests.⁴⁹

The restrictions on voting in national elections, moreover, were often overlooked, or simply overcome by the passage of time. As in Chile, literacy qualifications were sometimes reduced to proof by a signature.⁵⁰ The income requirements also eventually lost their original meaning. Whereas the 1886 requirement of five hundred pesos a year excluded most farm laborers and all domestic servants, it was not beyond the means of middle-income groups, such as schoolteachers, clerks, shopkeepers, and miners; or even people of lower income, such as some railroad workers and artisans—ironworkers, plumbers, masons, or tailors. By the turn of the century, moreover, given the depreciation of paper money, the agregados and chapoleros of the Hacienda Jonás, for example, were earning much more than the required sum.⁵¹

Circumstances may have changed after the financial and monetary measures taken by the Reyes administration, but in any event, the income and property requirements were also reduced after 1910. By the 1920s, most factory workers in Cartagena would have been enfranchised, according to the income test. Urban labor had long been involved in electoral politics; its participation was a tradition that, according to the union leader, Torres Giraldo (who regretted the trend), hampered the organization of a proper labor movement.⁵²

With the available data, it is impossible to describe the social composition of the electorate as has been done systematically for late nineteenth-century Buenos Aires, Chile, or, of course, England. The evidence suggests, however, that the electoral process was far from the exclusive patrimony of the few. Contemporary commentators, such as Enrique Santos, complained that in Boyacá “unfortunately the most educated abstain from the polls.”⁵³ Social pressure to widen the franchise — like that in European countries, such as Belgium, which suffered serious strikes at the turn of the century — was generally absent from not only Colombia but most other Latin American countries, for one simple reason: large segments of the popular sectors, particularly in the urban areas, were already in possession of the vote.⁵⁴

Some Colombian elections, such as that of President Miguel Abadía Méndez in 1926, could be described as the “private act of a few public employees,” but they were the exception.⁵⁵ Electoral participation varied considerably from department to department and province to province. Unfortunately, figures for local elections are unresearched and hard to find. The results of presidential elections, however—particularly those of 1914, 1918, 1922, and 1930 —show a significant expansion of the electorate: it almost doubled between 1914 and 1922 and increased by about 30 percent over the following eight years. Heavily contested elections, such as those of 1922 and 1930, attracted the largest proportion of voters: approximately 48 percent of Colombian adult males on each occasion. In a period when male universal suffrage was not the rule, even the degree of participation in the 1914 and 1918 elections —28 and 30 percent of adult males, respectively— was significant.⁵⁶

To interpret the level of electoral participation in terms of the effective number of votes, moreover, would be to take a narrow approach. Elections involved a larger number of people than those who actually cast their ballots. At election time, politics permeated the social atmosphere. Tertulias in the cafés, such as the Cigarra in Bogotá in 1930, revolved around elections.⁵⁷ Whole families helped the candidates print their ballots and distribute them among potential voters.⁵⁸ Canvassing was often a communal occasion not to be missed. So was election day.⁵⁹

Women may not have possessed the vote, but that did not exclude them from the electoral process. In 1911, the photographer for the newspaper El Gráfico in Bogotá captured the image of a group of devout women who “cheered the priests as they went to the polls.”⁶⁰ During the 1922 contest, Liberal women “signed letters of support [for the Liberal candidate], contributed to campaign expenditures, organized or chaired public meetings, addressed the people.”⁶¹ Conservative women could be even more passionate. They contributed to some dramatic moments when, before departing for the polling place, “the Conchitera mothers . . . with their eyes full of tears told their sons, as they handed them their machetes: “return with them or on top of them.”⁶²

Even those who resorted to abstentionism sometimes did so in a participatory way. In 1888, a dissident group of Conservatives in Mompox decided to abstain from voting in protest against alleged electoral malpractices by local authorities. While polling went on in town, about three hundred partisans moved to the neighboring island of Kimbay, where they enjoyed lunch and a party that continued until late that night, accompanied by music, “joined in the dance by partisan women . . .; by repetitive gun salutes and fireworks; and by cheers to the Conservative Party, the Conservative people, and to Dr. Núñez.”⁶³

Certainly some elections, such as that of 1914, took place with “an entire absence of party strife and feeling.”⁶⁴ In general, however, they were occasions of public excitement, particularly in small villages where electoral politics often was the only disruption to the monotony of daily life. When Rufino Gutiérrez visited Tuluá, Valle, in 1918, feelings were running high in the midst of the presidential campaign: elections were the sole topic of conversation in town.⁶⁵ Even the 1892 presidential election, in which only the vice presidency was contested, was described as a “struggle of life or death” by the Antioqueño Carlos E. Restrepo.⁶⁶ Electoral campaigns that sometimes lasted more than a year prolonged the state of tension.⁶⁷

The Conservative Hegemony attempted to move the country away from the “biennial electionary fever” that had characterized the previous federal regime, but the electoral calendar remained nonetheless full.⁶⁸ Between 1886 and 1930, nine presidential elections took place. The spacing of congressional and local elections was slightly less congested between 1886 and 1910, but after that, hardly a year passed without a campaign in progress. Every other year from 1911 on, elections took place for diputados in February, representantes in May, concejales in October. The frequency of elections in itself, with the passions they raised, could be a measure of their significance in the political system. But it was the lack of certainty about the results, despite the control exercised by those in power, that stimulated the development of a meaningful electoral life.

The Limits of Patronage, Fraud, and Coercion

In light of that uncertainty, the U.S. observer’s assertion that the “party in power here always carries the election” requires careful examination. For a start, the assumption that a cohesive Conservative Party ruled the country during this period, as already suggested, is misleading. In addition, dissidents were common, and they often performed successfully at the polls. Dissident movements sometimes took the form of separate entities struggling for power under individual banners: Independientes, Nacionales, Históricos, Republicanos. Thus in analyzing electoral results, the traditional bipartisan dichotomy needs to give way to a more complex picture.

This section attempts to describe a more competitive scenario in Colombian electoral politics than has hitherto been accepted for this period. In this competitive scenario, even a candidate supported by the government was far from possessing an automatic ticket to victory. Successive regimes might have used different mechanisms to control the political process (including the help of the clergy), but the system contained no well-organized or well-established electoral machinery that made it unchallengeable from within.

Between 1886 and 1930, three out of the nine presidential elections were apparently uncontested: those of 1892, 1910, and 1924. Yet the 1892 contest was peculiar in its struggle for the vice presidency, which led to a serious dispute between the Nacionales and Históricos; and the 1910 election took place amid the circumstances of restoring constitutional order. Liberals abstained from presenting their own candidates in 1904 and 1918, but some of them decided nevertheless to support one of the two Conservative candidates on each occasion. In both these years, the elections became fierce intraparty competitions, with wider implications for national politics. In 1914 the Liberal vote split among the Conservative and Republican candidates and a minor Liberal figure. Liberals did field their own ticket in 1898, 1922, and 1930. Thus, not all presidential elections followed the same pattern, and their outcomes also varied.

It may be argued that there were clear cases in which the “official” candidate carried the elections: in 1892, 1922, and 1924. Yet among these, only in 1922 did the government face a contested election, while the electoral campaigns and results of 1898, 1910, and 1904 merit closer examination before any definite conclusion is reached.⁶⁹ In 1914 the government, then controlled by the short-lived Republican Party, was defeated by the mainstream Conservatives. In 1918, Concha, in turn, handed power to Suárez, a traditional partisan foe. Finally, in 1930, not only the government but the two Conservative candidates suffered defeat at the polls. For any of these contests, breakdowns of the results by departments and municipios are necessary to appreciate where and when the government successfully carried an election.

As for local elections, the picture is even more complex, particularly after the 1910 reforms. In 1922, Liberals defeated both Conservatives and Republicans in various municipios. In Antioquia, for example, they gained control over the city councils of Medellín, Anorí, Caldas, Peñol, Segovia, and Puerto Berrío.⁷⁰ Between 1915 and 1919, the city council of El Banco, Magdalena, was in Liberal hands.⁷¹ By 1921, the emergent Socialist Party had enjoyed significant electoral victories in Cundinamarca, Tolima, and Antioquia, although, as Torres Giraldo himself has acknowledged, Socialist candidates were supported mostly by Liberal voters, who would soon return to the fold to vote for Benjamin Herrera in 1922.⁷² The Liberals lost the national presidential contest that year, but Herrera’s good performance in some departments, such as Santander, Valle, and Bolívar, boosted their morale and helped them fight successfully in the following elections for diputados and representantes.⁷³

It could be argued, furthermore, that the presidential elections, at least, offered hardly any “official” candidates of the kind described, for example, in nineteenth-century Chile.⁷⁴ A study of the campaigns themselves may reveal further aspects of electoral competition that do not emerge from cold statistics. The 1898 presidential contest, for example, illustrates how the development of the campaign determined crucial decisions that affected the final outcome. It also demonstrates the limits of presidential power in influencing elections.

By early 1896, more than a year before the voting, the electoral campaign was already under way. It included an effort to reelect Miguel Antonio Caro, then serving as acting president. Although Caro later denied that he had any such aspirations, he allowed his name to be discussed in the press. Some newspapers were launched especially to support his candidacy. Public reception of the reelection proposal, however, was disappointing. According to El Repertorio Colombiano in April 1897, even civil servants opposed it: “there is not a single signature of support that could indicate any backing to his name.”⁷⁵ In July, a poorly attended demonstration organized by his followers in Bogotá was a final signal to Caro to withdraw his name from the ballot. “The country has already said what it does not want,” wrote Carlos Martínez Silva, a leader of the Históricos, in a triumphant mood.⁷⁶

Caro also tried and failed to impose the candidacy of his minister of government, a former Independiente. Although a fellow Conservative, Manuel Antonio Sanclemente, was finally elected as Caro’s successor, in these circumstances it could hardly be considered an outright victory for the government; the opposition called it a “moral defeat.”⁷⁷ The election results had further implications. By 1898, the Nacionales had lost control of the asamblea in Cundinamarca. By August, when the new president was sworn in, the opposition clearly held the majority in the lower chamber. By the end of the year, a defiant national congress had passed legislation aimed at dismantling the “so-called fundamental regeneration.”⁷⁸

Some contemporary leaders, such as Caro, had no scruples about openly advocating an active role for public officials during elections.⁷⁹Empleomanía, the thirst for state jobs, was often said to be the main motivation behind politics; public employees might therefore be well disposed to give their support to the government’s candidates. As Martínez Silva acknowledged, however, they formed only “a minority among voters.”⁸⁰ The Conservative regime could also count on the support of the army, the police, and the clergy. Nevertheless, the government’s power to control the electoral process was limited. The so-called electoral machinery was, in theory, controlled from the center, by the president and his minister of government, mainly through their political appointees, the governors. As the Santandereano Manuel Serrano Blanco put it, “Governors were at the heart of all official and political life.”⁸¹ But the question remains to what extent their influence was similar to that of the governors in Argentina and Mexico at the turn of the century, whom François-Xavier Guerra calls “the sole and true Grand Electors.”⁸²

A closer look at the circumstances in which Colombian governors exercised their power suggests that they were far from omnipotent in their departments. For one thing, their authority, particularly in the distant provinces, could be challenged even by their own appointees.⁸³ Governors were under attack from all sides at election time, when “the rule of law is always considered arbitrary and ... all the actions of the authorities, even the most normal, are subject to criticism and quite a few insults.”⁸⁴ Governors in the Llanos, where Liberals often did well at the polls, have been described as a “besieged group” in Liberal territory.⁸⁵

Above all, governors had no official role in setting up the electoral boards, the bodies established to oversee the polls. This was especially true after 1910, when Congress was given the function of appointing the members of the Gran Consejo Electoral, which in turn presided over the pyramidal organization of consejos departamentales, juntas, and jurados electorales.⁸⁶ The relationship between governors and electoral boards in the departments was usually one of recurrent conflict. The government was often powerless in the face of electoral maneuvers by members of these boards; as the governor of Atlántico remarked, “the authority of the executive is reduced to contemplating the abuses of those who control the electoral power with a chain of employees larger and better paid than the government.”⁸⁷

It is therefore possible to distinguish, at some stages, the structure of government from that of the electoral machinery, even if at times they reinforced each other or even looked identical. While the former was centralized around the executive, with the president at the top, the latter followed more closely the vagaries of local politics. Of paramount significance to this process was the election of departmental deputies, for they, in turn, elected senators, and through them could influence appointments to the electoral boards. Thus the so-called electoral power was somehow tied to the results of municipal, departmental, and congressional elections, although the reverse was also true.⁸⁸

This electoral power was never permanently settled. Its composition changed as political loyalties shifted during successive electoral campaigns, when the different directorios of the parties, set up especially to contest the elections, exercised their leverage.⁸⁹ Although the situation varied from province to province, no party ever had an overall monopoly on electoral power. After the 1910 reforms, the Liberals were guaranteed a place on the electoral boards; in some constituencies, such as Atlántico in 1923, they managed to exercise considerable influence. A more typical scenario, however, was that of a coalition between different factions, such as Liberals and Conservatives, sharing power at the local level.⁹⁰

The control of electoral boards was inextricably linked to the question of fraud. In some instances, evidence of fraud and related electoral malpractice is easily identifiable. As the sole Liberal representative in Congress, the Antioqueño Rafael Uribe Uribe denounced the 1896 elections as a “shameless violation of suffrage.”⁹¹ The notorious Registro de Padilla of 1904, the year General Juan Iguarán filled in the blank ballot of his constituency according to his own personal whim, has become part of the folklore of Colombian political history.⁹² Equally notorious is the telegram from Ruperto Melo, a Boyacá gamonal, to President Ospina, sent eight days after the elections of 1922: “An enthusiastic Conservatism keeps on voting.”⁹³ These were cases of outright fraud—that is to say, crude adulteration of the ballot.⁹⁴ The manipulation of electoral results was usually more subtle, however; for example, controlling access to the franchise by drawing up a new register before each election.

Any attempt to understand the nature of electoral fraud and to appreciate how much it determined the outcome of Colombian elections must go beyond a mere description of the anomalies of the system.⁹⁵ Not all elections were fraudulent, and the degree of electoral corruption varied from constituency to constituency. Fraud, furthermore, was not an exclusive practice of any party or of the Conservative government. It was a Liberal, Pedro Juan Navarro, who was accused of distorting the electoral process in Barranquilla in 1923.⁹⁶ Some governments, particularly the Republican administration of Carlos E. Restrepo, made genuine efforts to convince local authorities that it was their responsibility to guarantee a fair and clean electoral process, though such efforts were often frustrated by the resilience of traditional political behavior.⁹⁷

As elections became competitive, the role of the opposition in conditioning the level of fraud became as crucial as that of the government. Those who considered the results unfair could resort to the judiciary, and occasionally their claims would receive a favorable answer.⁹⁸ From this perspective, electoral fraud ought to be considered in the wider context of what was, after all, a competitive struggle. In competitive terms, fraud, when present, was but one of many factors determining the outcome. In most contested elections, until the escrutinios (actual counting of votes) took place, “all parties considered themselves to be the winners.”⁹⁹ The escrutinios were conducted out in the open, as prescribed by law, and generated intense confrontations between the opposing factions. In 1918, during the escrutinios in Santa Marta, news of electoral fraud infuriated the barras (the crowd) witnessing the event. Violence against some of the escrutadores was followed by partisan clashes that left 2 dead and 12 wounded.¹⁰⁰ In circumstances such as these, the authorities were often hapless arbiters of conflicts that were beyond governmental control.

The Conservative regime counted on support from governors and other officials, nevertheless. In addition, it could rely on the army and the police, two institutions traditionally involved in the electoral process. Not only did their members have the right to vote, but on election day, soldiers guarded the polling places. For the opposition, the army and the police were instruments of the government employed to manipulate the polls. The army’s influence was intermittent, however. In number of votes, its weight diminished over time. In the 1917 local elections, for example, only 91 soldiers from the barracks in Barranquilla went to the polls.¹⁰¹

Not all soldiers were Conservative, moreover, nor did they all follow the government’s instructions. The 1917 vote from the regiment in Barranquilla was split: 31 for the “official” list of the Conservative Party, 22 for the dissidents, and 38 for the Liberals. Half the regiment was actually Liberal. From this and similar instances, the government often feared that it could not count on the loyalty of the troops on the coast.¹⁰²

The situation was certainly different in other regions, but as the army became more professional, the general tendency was to abstain from the polls, a policy particularly encouraged after 1910. The army’s electoral role, as Patricia Pinzón de Lewin has shown, was increasingly, albeit reluctantly, as the arbiter of partisan conflict.¹⁰³ When in the early 1930s a law was finally passed to prevent the army from voting, most officers, according to Ricardo Bayona Posada, did not consider themselves affected because they had “never voted.”¹⁰⁴ The police, who, by contrast, were under local rather than national control, remained highly partisan, although the Republican administration attempted to curb official influence over their voting practices.¹⁰⁵

Next to electoral fraud and coercion, no other charge was raised so often against the Conservative regime as that of its dependence on the church. Commentators both inside and outside the country constantly made references to the “theocratic republic.”¹⁰⁶ Given the significance of past conflicts with the church in the history of partisan divisions, the church was hardly neutral in electoral politics.¹⁰⁷ As polling day approached, bishops issued circulars not only upholding the clergy’s right to vote but also encouraging their flocks to cast their ballots for the “appropriate” candidates.¹⁰⁸ From the pulpit, local priests were more outspoken. In Boyacá, La Linterna complained in 1916, “the clergy declared war against Liberal and Republican candidates.”¹⁰⁹ By 1930 Monseñor Ismael Perdomo, the archbishop of Bogotá, was known as the “elector of electors in Colombia.”¹¹⁰

The clergy undoubtedly was heavily involved in electioneering, in close alliance with the Conservatives. The relationships between the priests and their congregations and between the clergy and the politicians, however, were complex. Under what circumstances could a priest exercise effective political influence over his parishioners? Were the priests in control of the electorate or, as seems to have been the case in nineteenth-century Ireland, did they appear “all-powerful as long as their views coincided with those of the electors”?¹¹¹ And who guided the process, the clergy or the laity?

Although the vast majority of Colombians were Catholic (at least nominally), the existence of sharp partisan loyalties indicates that deference to the priesthood in moral and religious matters did not extend to political affairs. The clergy sometimes felt betrayed: in 1918, Indians around Pasto surprised their priests when they cast their ballots for Guillermo Valencia and not for Marco Fidel Suárez as they had been instructed.¹¹² Again, the clergy’s influence varied significantly from region to region and village to village. Eduardo Caballero Calderón describes how in Boyacá the clergy, far from controlling an all-submissive flock, faced hostile frontiers: “A religious wind blows from Soatá to Tipacoque . . . from Capitanejo, on the contrary, there climbs through the Cañón de Chicamocha a hot air of Voltairean anticlericalism.”¹¹³

While it is possible to identify a clear element of anticlericalism in the Liberal electorate, the relationship between the Conservative voters and the clergy could sometimes be ambiguous. According to El Repertorio Colombiano, the hierarchy’s decision not to back Caro’s candidacy in 1897 interpreted the popular mood. Caro and his followers resented the clergy’s lack of support; in Tunja, a Nacionalista mob assaulted the bishop’s palace.¹¹⁴ By tactlessly meddling in electoral campaigns, the clergy risked alienating Liberals, Conservatives, and Republicans alike. In 1912, President Restrepo warned the Vatican that as a consequence of partisan sermons from the pulpit during election times, “those who do not share the same political opinions as the priests are leaving the churches.”¹¹⁵ The electorate’s reaction of the electorate was sometimes violent: there were riots against the Colegio San Bartolomé in Bogotá during the local elections of 1913.¹¹⁶

Fearing the rise of the Liberals, whose doctrines it vehemently opposed, the church did not hesitate to side with the Conservatives. This did not mean, however, that priests controlled the party. Some bishops reminded the faithful that the Catholic hierarchy did not depend on any “party leaders” or “heads of political circles.”¹¹⁷ The reverse was also true. When it came to elections, moreover, the clergy usually followed the laity’s lead. Priests might have the power to veto a candidate but, as Carlos E. Restrepo observed, the selection process was in the hands of partisan juntas, whose members usually studied “not the Catholic credentials of the candidate . . . but the force of his opinions, his partisan ascendancy, his merits in our civil wars.”¹¹⁸

In 1930, when Monseñor Perdomo decided to back the candidacy of Alfredo Vásquez Cobo, members of the Conservative Party’s national directorate thought the archbishop’s action broke with a longstanding tradition.¹¹⁹ The selection of the presidential candidate was a prerogative of the Conservative parliamentary majority, and in 1930 that majority favored Guillermo Valencia. As it had done in 1897, the hierarchy attempted to remain neutral in the face of a Conservative division; but in general, a divided party meant a divided church. And in these circumstances, as the year’s disastrous election results proved, the clergy’s effectiveness as an “electoral agent” was limited.

The church’s meddling had contradictory implications for Colombian political culture. It contributed to sectarianism, and therefore encouraged partisan conflict and a degree of intolerance.¹²⁰ But it also reinforced electoral traditions in at least two ways. First, it contributed to the “democratization” of politics in the sense described by Alberto Edwards for nineteenth-century Chile: religious questions touched on the lives of most people.¹²¹ Such questions raised issues that were commonly discussed and understood by all social groups. They encouraged, if for no other reason, a politically aware electorate. Second, and just as significant, the clergy’s involvement in elections strengthened a commitment to suffrage that embraced large sectors of Colombian society. Alongside the references to God as the creator of the “civil society” and the attacks against atheists and utilitarians, the bishops’ circulars emphasized “the duties [of the faithful] regarding elections . . . suffrage requires them to deposit their votes whenever necessary.”¹²²

The Opposition: From Bullets to Ballots

Despite all these factors, the Conservatives might be assumed to have managed to stay in power simply by rigging elections. This question, however, cannot be properly addressed without considering the attitudes toward suffrage of the main opposition party. After all, the Conservative regime had been the product of a civil war that had left a defeated Liberal Party strongly divided and demoralized. As the U.S. minister commented, regarding assertions that the Liberals were not protected in their right of suffrage in 1891, “this is not known to be so, because the Liberals have never attempted to exercise it to any considerable extent.”¹²³

The opposition did eventually come to terms with the ballot box. That reconciliation, however, did not exclude a persistent revolutionary mood among the rank and file, also reflected in the various factions struggling for control of the party. The final section of this essay examines how the Liberal Party moved away from revolutionary strategies and became committed to suffrage in the struggle for power. As significant as the politics of influence was in elections, the politics of organization was crucial in explaining electoral behavior during the Conservative Hegemony.

In 1891, instructing party followers on how to organize themselves for the elections, the Centro Liberal de la República in Bogotá acknowledged that many a Liberal was reluctant to get involved. The Centro Liberal appealed to the revolutionary elements of the party not to miss any chances; they should also take advantage of electoral opportunities.¹²⁴ Throughout the 1890s, the Liberal Party was divided among those who wished for peace, those who called openly for a resort to arms, and those who, while accepting the regime, left the door open for an eventual revolution.¹²⁵ A revolution did break out in 1895, weakening the party and furthering divisions as a result of another military defeat. After the brief interlude of the 1898 presidential campaign, and with feelings of frustration at the results, the bellicose faction gained the initiative within the party. In a widely publicized speech in 1898, Rafael Uribe Uribe voiced the arguments of those who despised suffrage under the Conservatives: any Liberal involvement in the electoral process served only to legitimate an odious regime.¹²⁶

The Guerra de los Mil Días, which followed, led to the third military defeat in a row for the Liberals. It seemed to mark the end of their military adventurousness, which, in any case, now faced an increasing professionalized army. Nevertheless, the Liberals’ revolutionary spirit did not subside.¹²⁷

A nonrevolutionary strategy to oppose the regime, as already noted, was to abstain from voting. According to the newspaper Vanguardia Liberal of Bucaramanga, this was a very effective means of protest.¹²⁸ The Liberal Party resorted to abstention as a political weapon on various occasions, particularly during presidential campaigns. Only in exceptional cases, however, did this policy mean total avoidance of the ballot box. More often than not, it meant refraining from presenting a Liberal ticket in the presidential contest. And in most of these events —1904, 1914, and 1918, for example —some Liberals ended up supporting a Conservative candidate. They thereby further split a party already divided over questions of war or peace, abstention or electoral participation, and now alignment or nonalignment with Conservative factions.

Other opposition parties also adopted a policy of abstention. In 1923, the Convención de Obreros Socialistas declared suffrage a “useless exercise” although, like the Liberals, the Socialists left the door open to contest municipal elections.¹²⁹ At the local level, Conservatives were also ready to abstain from the polls if circumstances demanded.¹³⁰ Abstention, moreover, was not always a protest against the legitimacy of the regime. In what was doubtless an extraordinary event, Liberals and Conservatives in Bucaramanga in 1923 decided jointly to stay away from the municipal elections in order to press the central government to complete the Puerto Wilches railway.¹³¹

As a political strategy, abstaining from the polls was neither consistently utilized nor unanimously backed by the opposition. In the electoral convention that decided to contest the presidency in 1897, delegates warned fellow Liberals that a policy of abstention could lead eventually to the disappearance of the party.¹³² After the turn of the century, and particularly after 1910, concern grew among some Liberals that rejecting the ballot box could be political suicide. Thus, voicing a defense of elections often implied overcoming past beliefs about how to reach power. “Power is not conquered now through the barracks; it has to be conquered through the ballot box,” Ricardo Tirado Macías told an audience of laborers in 1913.¹³³

As suffrage acquired value, the abstentionists became a target of Liberal attack. For Liberal newspapers, such as La Linterna of Tunja, voting was a civic duty.¹³⁴ In Bogotá, El Tiempo led the campaign against abstention, supporting the criticism of ideologues within the party, such as Antonio José Iregui.¹³⁵ To motivate the Liberal electorate, it was necessary to emphasize what was at stake at the polls. In 1925, on the eve of the municipal elections, a manifesto issued by a Liberal junta to partisans in Santa Marta detailed the problems to be solved by the incoming city council: land conflicts, water supply, and housing, among others.¹³⁶

Nevertheless, no argument against abstention was more powerful than the experience of electoral victory. In 1923, Vanguardia Liberal, whose director had led a successful campaign in Santander, abandoned its previous abstentionist line.¹³⁷ Although during the second half of the 1920s, abstentionist and even revolutionary strategies regained some support in the party—particularly among a few civil war veterans —they were rejected by rising figures such as Jorge Eliécer Gaitán and Alfonso López Pumarejo.¹³⁸

Liberal decisions to return to the polls stimulated a fresh electoral rhetoric, accompanied by a concern for party organization. In its 1891 pamphlet of advice to the Liberal rank and file, the Centro Liberal de la República encouraged Liberal cadres in the departments to canvas the barrios, speak on the hustings, and, if possible, distribute propaganda. It also warned party workers to scrutinize the electoral process rigorously. From the composition of the electoral juries to the sites of polling stations, nothing should be left to chance. The members of the Centro Liberal had no doubts about the key to success: it was “a good electoral organization.” A true party had to develop “habits of electoral vigilance.”¹³⁹

It was precisely to the lack of these elements that Enrique Santos attributed the Liberal failures two decades later. Santos did not cease his attacks on the church or his denunciation of electoral fraud, but he thought more blame should be placed on the Liberal Party. No organized party could be fooled or cheated by fraud.¹⁴⁰ Party divisions had indeed been the main factor behind the Liberal defeat in the 1915 Boyacá Congressional elections.

Despite these frustrations, Liberal involvement at the polls was not entirely fruitless. Throughout the period, not only the Liberals but also small opposition parties, such as the Socialists, managed to score significant electoral gains, occasionally even gaining control of local government. More scholarly attention should be given to those campaigns, when Liberal organization posed serious threats to the regime, and to the various circumstances that encouraged the mobilization of the Liberal electorate. Miguel Samper may not have won the presidency in 1898, but in Bogotá, an evident Liberal triumph boosted morale. Samper, in any case, apparently was not popular among certain factions in the party.¹⁴¹

Similarly, when the Liberals decided to back a Conservative candidate, they could not expect to arouse much enthusiasm from their rank and file; but to canvas for a popular leader such as Benjamín Herrera was a different story. Liberals rallied around Herrera in mass demonstrations. The results were later reflected in electoral majorities in most of Colombia’s largest cities: Bogotá, Barranquilla, Cali, Medellín, and Ibagué.¹⁴² Liberal gains in municipal, departmental, and congressional elections, as Alejandro Galvis Galvis noted in Santander in 1923, were also the result of intense electioneering and improved party organization.¹⁴³

When, at the end of 1929, the candidacy of Enrique Olaya Herrera captured the popular imagination, Alcides Arguedas witnessed an extraordinary movement of people in the streets of Bogotá, “a miracle of civic pride.”¹⁴⁴ Far from being a miracle, the mass mobilization that followed at the polls was rooted in a long-established electoral tradition, kept alive in spite of civil conflicts, official malpractice, party divisions, and calls for either insurrection or abstention.

Conclusions

On February 10, 1930, Bogotá’s newspapers published the results of the presidential elections held the previous day. “Guillermo Valencia triumphed throughout the country,” ran the headline in El Debate. A different impression was obtained from reading El Nuevo Tiempo, whose pages gave the lead to Alfredo Vásquez Cobo. El Tiempo, in turn, had no doubts: Enrique Olaya Herrera had won by a landslide. Arguedas was taken aback by this curious outcome: “Three candidates for the presidency, . . . and all of them winners; this is unbelievable.”¹⁴⁵ The mood in the streets, however, did not reflect the contradictions in the newspaper’s headlines, and soon the Conservative candidates and the government were conceding defeat. In accepting defeat and handing over power peacefully to the opposition, the Conservatives made the presidential contest of 1930 a landmark in Colombian history.

The events of 1930, however, need to be considered in the wider context of an electoral tradition that had been sustained during the Conservative Hegemony. Significant changes occurred throughout the period, but even before the 1910 reforms, as this essay has shown, the nature of Colombian politics could not be fully appreciated without an understanding of the existence of a meaningful electoral life. The development of an electoral culture in Colombia had its origins in the early republic. At the turn of the century, it was encouraged by a reappraisal of suffrage, a development that should receive further scholarly attention. During the 1890s, there were already signs of an emerging intellectual atmosphere, in which the opposing parties sought compromise and mutual tolerance. Nevertheless, it was only after the Guerra de los Mil Días that a settlement was reached.

During the following decades, the political debate was heavily influenced by a new generation of intellectuals—the so-called Generación del Centenario —which included leading Liberal, Conservative, and Republican figures.¹⁴⁶ While in the opposition, they supported the abandonment of revolution in favor of the ballot. While in government, they attempted to reform electoral practices. What was needed, according to Carlos E. Restrepo, was a pedagogy of suffrage directed at both the authorities and public opinion.¹⁴⁷ Yet for educating society about suffrage, nothing was more significant than the very practice of elections—those prolonged campaigns during which, year after year, a proliferation of speeches, canvasses, leaflets, and even pastoral letters from the clergy served to widen the social commitment to representative forms of government.

As in other countries, the development of electoral traditions in Colombia was a patchy process. The national overview presented by this essay has aimed to show that, between 1886 and 1930, Colombian electoral life was more significant than has hitherto been accepted. Elections were competitive, involving degrees of uncertainty, their outcome conditioned not just by the exercise of fraud, patronage, and coercion but also by much forthright and untrammeled electoral activity. An effective campaign, and an effective candidate, could make a difference.

As in all countries, elections varied significantly from place to place. Julio H. Palacio left a vivid description of electioneering when he went canvassing with his partisans throughout the Atlántico province in 1897. In Soledad, Polo Nuevo, and Baranoa, they had to make speeches and socialize in their attempts to persuade an undecided electorate. In Juan de Acosta and Tubará they did not bother to canvass at all, because they already counted on a majority of supporters through the dominant local families. They did not bother in Santo Tomás, either, but in this case because the town was an impenetrable “Liberal bastion.” In Sabanalarga they risked facing hostile voters, and the fear of a violent public reaction forced them to leave the town quietly at siesta time.¹⁴⁸ Only further studies at the local level would help distinguish how amenable to control the electorate was; the constituencies in which traditional patterns of deference and patronage prevailed or venality and corruption flourished; and where and how electoral competition finally developed.

How the World Votes is the title of a book by Charles Seymour and Donald Frary, published in 1918, in which these two Yale scholars examined “the story of democratic development in elections,” chiefly Europe and the United States. Besides a brief section on Japan, they did, however, devote a chapter to South America. Here Seymour and Frary described elections as “usually a pure sham,” a “farce” in which only “official influence” mattered. It may be a credit to the authors that they included the region at all; they even acknowledged in an afterthought that “a superficial consideration of the workings of democracy in the South American states has led to many hasty and inconsidered judgments, which have bred unwarranted contempt for the conditions of politics in our sister republics.”¹⁴⁹

Electoral life in the history of Latin American politics remains a neglected field, despite some significant advances. Thus, in a major study of patronage and politics in nineteenth-century Brazil, Richard Graham highlights the role of Brazilian elections both as a means of legitimating the ruling order and as a mechanism of social control. In a more recent essay, François-Xavier Guerra refers to the political systems of late nineteenth-century Latin America as “democratic fictions,” in which elections were government-controlled events whose results were needed for the purposes of legitimacy.¹⁵⁰ Graham certainly presents a complex picture of Brazilian politics, while Guerra reveals interesting aspects of the origins of the Spanish American tradition of representation. Nevertheless, to define in advance the meaning of elections by their legitimating role, or by their function as techniques of control, tells little about the peculiarities of Latin American elections. After all, as Frank O’Gorman observes with regard to the English experience, “where there were electors to vote, there were problems of control”; and wherever elections have taken place, they have always played a legitimating role.¹⁵¹

We need to go further and look deeper into the circumstances and conditions under which that control was exercised or that legitimacy obtained. In this context, elections in Latin America may acquire a meaning that most historians have been hitherto unwilling to accept: as in nineteenth-century Ireland, they could be interpreted as a “mechanism for assessing . . . the relative state of power relationships within society as a whole.”¹⁵² Even more significant, their study could be helpful in appreciating where and when, why and how elections succeeded or failed in advancing representative forms of government.¹⁵³

This approach is all the more relevant to Colombian history because that country, as Ignacio Torres Giraldo remarked, “has always lived ‘en razón electoral.’ ”¹⁵⁴ By rescuing the process of electoral competition from oblivion, it is possible to identify how electoral practices, which involved large sectors of Colombian society, strengthened a widespread commitment to representative democracy. It is this tradition that this paper has tried to vindicate; a tradition not without ironies and paradoxes, but a tradition nonetheless, which, between 1886 and 1930, managed to survive both “revolution” and “hegemony.”

I wish to thank Malcolm Deas for his advice, encouragement, and generosity, and for allowing me access to his library. I am grateful to Carlos Malamud, Gustavo Bell, Paula Alonso, Carlos Dardé, and Margaret Lavinia Anderson for reading and commenting on the original typescript, and to Marco Palacios and Margarita Garrido for providing useful documents. I also wish to thank the four anonymous HAHR referees for helpful criticism. Earlier versions of this paper were presented at the Latin American Seminar at St. Antony’s College, Oxford; the Instituto Universitario Ortega y Gasset, Madrid; the Universidad del Atlántico, Barranquilla; the Universidad de los Andes, Bogotá; and the 18th Latin American Studies Association Congress in Atlanta, where I received useful comments.