United States of Colombia

It is easy to mistake this for a copy of the United States of America, since the name was borrowed and the vocabulary of states, union and constitution was borrowed with it. The resemblance is superficial. The Union of 1863 was much weaker than the American federation had become by the 1860s, and the Colombian states were stronger than American states ever were: they coined or chartered money, kept standing forces, negotiated with foreign creditors and wrote and rewrote their own constitutions — forty-odd of them in twenty-three years. The federal government lived off the customs houses and the post office. When it needed an army it had to ask the states for one.

The republic matters for three reasons. It was the fullest experiment in classical Liberal government that Latin America attempted — free trade, a free press, secular schooling, the separation of church and state, and an explicit attempt to keep the state small and the citizen armed. It ended in a civil war and in the 1886 constitution, a Catholic, unitary, centralist settlement that governed Colombia, with amendments, until 1991. And the quarrel it contained, between a state that is one nation and a state that is a pact of regions, is still the quarrel of Colombian politics: the constitution of 1991 elected the governors and handed the departments a share of national revenue, which is in part a return to Rionegro by other means.

At a glance

What did the constitution of Rionegro create?

The convention met at Rionegro, a town in the hills east of Medellín in the state of Antioquia, on 4 February 1863, and sat until 8 May. Its sixty-odd delegates had been elected by direct vote, but only Liberals could win: the Conservative party had been beaten in the field, its leaders were in exile or in prison, and the opposition press had been shut down. The constitution they produced was signed on 8 May, exactly three years after Mosquera's decree of 8 May 1860 had declared the state of Cauca separated from the Granadine Confederation. Nobody in the room was in doubt about who had won.

The document had 93 articles in 13 chapters, and it was deliberately a partisan statute. It opened with the nation, went on to the bases of the union, the property and charges of the union, citizens and foreigners, the general government, the legislature, the executive, the judiciary, the public ministry, elections, miscellaneous provisions, amendment, and the transitional regime. It did not invoke God. Its two governing ideas, as the Colombian constitutional historian Diego Uribe Vargas put it, were unlimited autonomy for the states and an absolute character for individual rights and liberties. The Liberals had spent the previous decade watching Conservative presidents use the central state against them; they were not going to leave the instrument in place.

Hence the odd machinery of the presidency. The president served two years, could not succeed himself, and was chosen indirectly: the voters of each state chose according to their own state's law, the winner in each state took that state's single vote, and a candidate needed an absolute majority of the nine; failing that, Congress chose from the leading candidates. The states kept the right to elect their own governors and assemblies, to raise their own taxes, to maintain their own forces and to legislate on education, land, justice and religion. The Union kept foreign relations, war and peace, the customs houses, the post, and the few properties the constitution listed as national. Everything else had been given away on purpose.

The name was part of the programme. The union that Bolívar had built and lost, Gran Colombia, had been a single centralised republic; this was to be the opposite. Mosquera had renamed the country the United States of Colombia by decree in November 1861, when he was provisional president and the war was still being won; the convention adopted the title formally, and the constitution fixed it. The model was American, but the intention was Colombian: to make the union a league of self-governing provinces rather than a state with provinces inside it.

Who were the nine sovereign states?

The nine were Antioquia, with its capital at Medellín; Bolívar, at Cartagena; Boyacá, at Tunja; Cauca, at Popayán; Cundinamarca, at Bogotá; Magdalena, at Santa Marta; Panamá, at Panamá; Santander, at Socorro; and Tolima, at Ibagué. Eight of them had been created between 1855 and 1857 out of the old provinces; Tolima was Mosquera's own creation, carved out of western Cundinamarca in 1861 and rewarded for having backed him. Bogotá itself was cut out of Cundinamarca in 1861 and administered as a federal district.

They were wildly unequal. Cauca covered 668,400 km² — the whole Pacific slope from the Ecuadorian border to the Chocó, plus the Patía valley and a huge slice of the Amazon — and had 435,078 people in the returns of 1874. Cundinamarca had 206,400 km² and 409,602 people; Boyacá, small and dense in the highlands, had 482,874, the largest population of any state; Santander had 425,427 on 42,600 km²; Panamá had 221,052 on 86,200 km², strung along a transit route that mattered more to the world than the state did. Magdalena, with 85,275 people, was the smallest and emptiest. Antioquia, with 365,974 people on 59,000 km², was the richest in gold and the most dynamic in settlement.

Sovereignty was not a rhetorical flourish. Each state had a constitution of its own and amended it when its politics changed, which was often: the period produced something like 42 state constitutions. Each had a legislature, a governor, courts applying its own civil code — Cundinamarca's code of the 1860s became a model copied across the country — and its own school system. Each raised revenue from land, commerce, spirits and stamps, and each owned the mines and salt deposits inside its borders, which is why the constitution of 1886 had to take the mines and salt mines back by name. Several states chartered banks that issued notes. All of them kept armed forces.

The result was a country with nine capitals and no single political market. A Liberal in Socorro and a Liberal in Popayán might share a party label and nothing else; a Conservative in Medellín lived under a government he had voted against and could, within the letter of the constitution, do something about it. The state was the unit of political life, and the union was the arena in which the states bargained.

How did the federation govern itself?

The federal government was small, cheap and chronically short of money. Its revenue came almost entirely from import duties, collected in a handful of customs houses — Barranquilla, Cartagena, Santa Marta, Riohacha, Buenaventura, Tumaco, and the Isthmus ports — and from the post. It levied no direct tax on citizens, because the constitution had left that to the states, and it therefore had no fiscal relationship with the people it governed. When customs receipts fell, as they did whenever a war closed a port or a foreign market sagged, the Union had nothing to fall back on.

It also had no army worth the name. The constitution left the states their forces, and they used them: historians of the period count some forty regional civil wars and one national war in twenty-three years. What kept the union together was not force but the machinery of elections and the patronage that went with them. Because the presidency was decided state by state, every national election was a negotiation among nine state machines, and the president governed by keeping enough of them satisfied. The Radical leaders — Manuel Murillo Toro, Santiago Pérez, Aquileo Parra, Eustorgio Salgar, Salvador Camacho Roldán, Felipe Zapata — were not a party in the modern sense so much as a network of regional clients who shared a programme and a dislike of the Conservative hierarchy. The arrangement is usually called the Olimpo Radical, the Radical Olympus, and the name stuck because it described the way a small group of men rotated the presidency among themselves.

Elections were staggered, which made politics continuous: the states did not vote at the same time, so there was always a campaign somewhere, and a presidential election could turn on which state had last changed hands. An amendment in 1876 tried to synchronise the state elections so that the presidential succession would be less arbitrary. It helped a little. It did not create a national electorate, because there was no national register, no national constituency and no national party organisation outside Congress.

The judiciary had a Supreme Court, and the constitution gave it the power to decide cases arising under the constitution and the laws of the Union. It had no way of enforcing a judgment against a state that chose to ignore it. That gap — a court without a sheriff, a president without an army, a congress without a tax — is the whole institutional history of the United States of Colombia in a sentence.

Who was Mosquera, and why did he fall in 1867?

Tomás Cipriano de Mosquera was born in Popayán in 1798, into the family that dominated the Cauca valley, and made himself the most consequential Colombian of the century after Bolívar. He was a soldier of independence, a diplomat in Lima, Santiago and Washington, president of New Granada from 1845 to 1849, and the man who commissioned Agustín Codazzi's chorographic commission and pushed through the abolition of the tobacco monopoly, the metric system and the first steam navigation on the Magdalena. He was also, by temperament, a caudillo: a great estate owner who believed in order, progress and himself.

He had begun as a Conservative and a centralist, and the switch came late. As governor of Cauca from 1858 he gathered the Liberal factions around him, and on 8 May 1860 he declared his state separated from the Granadine Confederation and went to war on the Conservative government of Mariano Ospina Rodríguez. He took Bogotá on 18 July 1861, made himself provisional president, and used the next two years to break the Church: the desamortización of ecclesiastical property, the expulsion of the Jesuits, and the tuición, the licence the civil power claimed over the exercise of worship. Those measures were his, not the convention's, and the convention softened some of them.

The delegates at Rionegro nonetheless elected him president on 12 May 1863, and he governed until 1 April 1864. In December 1863 he took the country into a short war with Ecuador and won at Cuaspud, which recovered some of the prestige the convention had cost him. He then went to Paris as minister, published a geography of the country, and came back to win the presidency again in 1866, this time as the candidate of the popular Liberals and the artisans against the Radicals who had trimmed his powers. The second term lasted a year. He quarrelled with Congress, ordered an investigation into how the confiscated church property had been sold — an investigation that threatened half the Liberal establishment, which had bought it — and in April 1867 closed Congress. In May the army, or the part of it that mattered, arrested him. He was tried by the Senate, sentenced to three years' exile, and spent them in Lima.

The date of the coup is given as 12 May in the official presidential chronology and as 23 May in most histories; the difference seems to be between the day he was seized and the day the new government was installed. What matters is that the first coup of the federal era was carried out not by Conservatives but by Liberals, against a Liberal president who had tried to govern as a national leader rather than as the manager of a coalition of states.

He came back in 1871, was elected governor of Cauca, sat in the Senate for the state in 1876, and died on his estate at Coconuco in 1878, aged 80. His career explains both the strengths and the weaknesses of the republic: the man who made the federation possible could not live inside it.

What did the Radicals change in the years of peace?

Between 1867 and 1878 the Radicals governed without a national war, and they used the time. The single most consequential measure was the Decreto Orgánico de Instrucción Pública of 1870, which made primary education free, compulsory and secular, put the school system under national supervision, and invited a German mission to train teachers in modern methods. Enrolment in primary schools rose from 43,458 in 1871 to 80,458 in 1876. It is the clearest single number the period offers for what Liberal government meant in practice, and it is also the number that explains the war that followed: the decree took the schools out of the hands of the parish priest.

The same years produced the National University (1867), the Diario Oficial, the national archive and library, and the Banco de Bogotá (1870); the first telegraph line had opened on 1 November 1865, between Bogotá and the neighbouring town of Mosquera. The 1871 monetary law put the peso on a gold footing at five francs, an attempt to attach the country to the Latin Monetary Union; silver stayed in circulation at a legal value above its market price, so the country was bimetallic in practice, and the peg did not survive the war of 1885. Railways were chartered and begun: Barranquilla's line to the sea, the Buenaventura line, the Northern railway from Bogotá, and later the Bogotá–Girardot line. None of them reached the highlands before 1886. The country still moved its coffee and its tobacco on mules and on Magdalena steamboats.

The reforms were not only economic. The Radicals extended the civil power over marriage, cemeteries and the registration of births and deaths, cut the standing army, kept the press free enough that opponents of the government published daily, and tolerated Protestant worship and foreign residents. They also kept the franchise narrow in practice: the constitution of 1863 left electoral law to the states, and the states set it by property, literacy or residence, so that the "democratic" republic was a republic of male householders with a trade, not of adult men.

The cost of the programme was borne by the Church and by the Conservative elites of the highlands, and neither accepted it. The Radicals governed as if the war of 1860 had settled the matter. It had not; it had only postponed it.

Why did the schools question produce a civil war in 1876?

The election of 1876 made the quarrel unavoidable. The Radicals blocked the candidacy of Rafael Núñez, who had begun to attract Liberals tired of the Olympus, and Congress declared Aquileo Parra elected in February. Núñez's supporters called it a fraud. The Conservatives, who had been waiting for a national leader and a national grievance, found both in the school laws: the state was teaching children to read without the catechism, and the parish priests told their parishes that this was a sin against God and the Virgin.

The war began in Cauca on 9 July 1876, when Conservative irregulars from Antioquia and Tolima attacked government troops at Tuluá; the president of Cauca, César Conto, declared the public order disturbed three days later. Antioquia's president, Recaredo de Villa, and Tolima's, Antonio Cuervo, declared for the rebellion in August, and Parra responded by raising the federal army to 30,000 men and dividing it into four campaign armies: Julián Trujillo in the south, Santos Acosta in the west, Fernando Ponce on the Atlantic coast, and Joaquín Reyes in reserve. The Church blessed the rebellion, and priests recruited for it; the government called it a war of religion and treated captured guerrillas as bandits.

The fighting was bitter and improvised. Both sides used guerrillas, both extorted money and horses from farms, and the ranks were filled with peasant militia officered by men whose qualification was social position rather than training. Trujillo won the first large engagement at Los Chancos at the end of August, took Pereira in November, and fought the rebel army to a standstill at La Garrapata on 19–22 November, where the Antioquian Conservatives used a machine gun for the first time in the country's wars. The decisive blow came at Manizales, which Trujillo stormed on 23 March 1877; Antioquia's president, Silverio Arango, surrendered at San Antonio the same night, and the Liberal army entered Medellín on 22 April. The last engagement was at El Cocuy on 27 April, and the last rebel commander, Felipe Farías, submitted in Magdalena on 25 May.

The settlement was punitive. A general amnesty was issued on 6 July 1877 for all who laid down their arms, but it excluded officers, bishops and clergy who had taken part, and the government exiled the bishops of Pasto, Popayán, Santa Fe de Antioquia and Medellín. Trujillo, the victor, was elected president in 1878, and the state of Antioquia was placed under a civil and military chief: the Conservatives' richest province was governed for a time as occupied territory.

What did the war of 1876–77 settle?

It settled that the federation could defend itself and that it could not do so without becoming something other than what it had been. To beat the state armies the federal government had to raise a national army, find the money to pay for it, and give the president emergency powers that the constitution had never contemplated. After 1877 the Union had a larger force and a larger debt than the men of Rionegro had intended, and the states had learned that a national election could be reversed by force.

It settled nothing about the schools. The decree of 1870 stayed on the statute book, the Church stayed out of the classroom where the Liberals had the power to keep it out, and the Conservatives stayed unreconciled. Enrolment, which had reached 80,458 in 1876, fell to 68,291 in 1880 and did not recover its pre-war level for years. The war had been fought over the formation of citizens, and on that ground neither side won.

It also broke the Liberal party. The Radicals had won with the votes of Liberals who disliked them, and the "independent" Liberals around Núñez drew the conclusion that the party could not govern the country from the Olympus. When Núñez was elected president in 1880, with Conservative votes as well as Liberal ones, the federal republic entered its last phase: a Liberal presidency that had begun to look for a different constitution.

The human cost is not well recorded. Contemporary estimates for the war of 1876–77 run to tens of thousands of dead, mostly from disease in the camps and among the guerrillas, but no reliable count exists; the figures usually printed are reconstructions. What is certain is that the war was fought across most of the nine states, that it lasted ten months, and that it taught both parties that the next war would be fought for the state itself.

How did tobacco, quinine and coffee pay for the country?

The federation lived by exports, and the exports changed under it. Customs receipts at the ports were the Union's income, so the country's trade was not a background to its politics; it was the fiscal base of the government that the politics was about.

Tobacco came first. The leaf of Ambalema, on the upper Magdalena in Tolima, had been a state monopoly until 1849, and the Liberal abolition of the monopoly turned it into an export crop; by the 1860s it was the country's leading legal export, shipped down the Magdalena to Barranquilla and on to Bremen and Hamburg. Phanor Eder, writing in 1913, noted that Ambalema tobacco "at one time had a reputation almost superior to that of Havana" and that the town still showed the evidence of the wealth of "the good old tobacco days". The good days ended in the 1870s: the soils of the Magdalena plain were exhausted, the fevers of the valley killed labourers, the leaf was of uneven quality, and Sumatran and Java tobacco took the European market with a cheaper and more consistent product. By 1900 Ambalema was making cigars for the home market.

Cinchona bark, the source of quinine, was the second boom. The trees grew in the temperate forests of the eastern cordillera, above Fusagasugá and the Tequendama falls in Cundinamarca, in Cauca, in the Santanders and in Nariño, and from the 1850s Colombian bark supplied a large part of the world's quinine. In 1875 more than £2,000,000 of cinchona bark was exported. Then the Dutch planted cinchona in Java and the British in India, and the price of sulphate of quinine fell from 16s 6d an ounce in 1879 to 3s 6d in 1883 and 2s 6d by 1885. A trade built on the collection of wild bark could not survive a plantation industry, and by the mid-1880s it was over.

Coffee was the third and the lasting one. It had been grown in the north-east since the 1830s, in the hills of Norte de Santander and then Santander, and it spread into Cundinamarca and, from the 1870s, into the frontier lands of Antioquia, where the colonización antioqueña pushed smallholders up the western slopes and down into the Cauca valley. Coffee needed labour, altitude and transport, and it rewarded the family farm rather than the plantation; it also needed credit, which came from the merchant houses of Medellín and Bogotá that advanced money against the crop and took the consignment. By the late 1880s coffee had passed tobacco and quinine and was the country's leading export. Its great expansion, the one that made Colombia a coffee country, came after 1900: in 1911 the country exported 37.9 million kilogrammes worth $9.47 million.

Gold ran underneath all of it. Antioquia's mines had been the Spanish empire's chief source of gold and remained the one export that never failed, which is why Medellín's merchants had capital to lend when the coffee frontier opened. Hides, cacao, vegetable ivory, dividivi and straw hats filled out the list without changing its shape.

The finances were the weak point. Exports were worth $6,772,017 in 1865, $8,247,817 in 1870, $15,836,943 in 1880 and $14,171,241 in 1885, and the trade was chronically in deficit with imports. The Union borrowed abroad, serviced its debt irregularly, and after 1880 began to experiment with paper: the Banco Nacional, founded in 1880 with the government as its only shareholder, was given the right to issue notes in 1886 without obligation to redeem them in specie, which is where Colombian fiat money begins. In 1883 a financial crisis, produced by the fall in quinine and the low price of coffee and tobacco, pushed foreign exchange to a premium of 20 per cent and reduced the money in circulation in Bogotá, a city of some hundred thousand people, to a reported $200,000. A state that cannot tax its citizens and depends on a single customs house is a state that lives from one harvest to the next, and the federation lived that way for twenty-three years.

Why did Panama become the republic's central diplomatic question?

Panama was a state of the union with a small population, a thin elite and the most valuable strip of real estate in the hemisphere. The Panama railway, built by an American company and opened in 1855, carried travellers and freight across the isthmus and paid the Colombian government a share of its receipts; the treaty signed with the United States on 12 December 1846 — the Mallarino–Bidlack treaty — had given Washington a right of transit and a guarantee of neutrality over the route, which in practice meant a right to land troops to protect it.

Bogotá understood both the value and the danger. In 1869 and 1870 the government negotiated canal treaties with the United States and let them die in Congress, where the American terms and the American demand for a strip of territory looked like a lease on sovereignty. The refusal sent the canal to Europe. In 1876 Congress passed a law authorising a concession, and a contract was signed on 28 May with A. de Gorgoza and General Stephen Türr, acting for a French society; a French naval officer, Lucien Napoléon Bonaparte Wyse, spent five months on the isthmus in 1876–77 surveying routes and then went to Bogotá.

The result was the Wyse concession of 20 March 1878, granted to the Société Civile Internationale du Canal Interocéanique de Darien. It gave the company the exclusive right to build an interoceanic canal through Panama, and it provided that the waterway would revert to Colombia after 99 years without compensation. On 11 July 1879 the Túrr society sold its rights to Ferdinand de Lesseps for 10 million francs; the Compagnie Universelle du Canal Interocéanique was organised on 17 August 1879; the first ceremonial spade was turned on 1 January 1880; and excavation began in March 1881. The French attempt, the sea-level plan, the yellow fever and the money are another story, but the Colombian half of it matters: the concession was a lease on the country's principal asset, granted by a government that needed revenue, to a company that promised a canal and a share of the tolls.

The isthmus then became the place where Colombian politics and American power met. In April 1885, with the country at civil war, the United States landed sailors and marines at Colón and Panama City to protect the transit route, and a Chilean cruiser, the Esmeralda, was sent from Valparaíso to make clear that Chile would not accept an annexation; the Colombians reoccupied Panama City at the end of April. The precedent was not lost on anyone. When the constitution of 1886 was written, it dealt with the isthmus in two clauses: Article 201 placed the Department of Panama under the direct authority of the national government, to be administered by special laws, and Article 202 reserved to the nation the assets the union had held, including the mines and salt mines the states had owned.

The French company collapsed in 1889; the scandal that followed in France ruined de Lesseps's reputation and left the concession, the excavations and the treaties in the hands of a liquidator. The United States bought the French assets in 1902, and when Bogotá refused the Hay–Herrán treaty in 1903, the republic of Panama declared independence with American warships standing off the coast. The United States of Colombia had spent twenty-three years treating the isthmus as its chief diplomatic asset; the republic that replaced it lost it within seventeen.

How did Núñez and the Regeneration end the federation?

Rafael Núñez was born in Cartagena in 1825, trained as a lawyer, and made his name as a minister of war and then of finance in the 1850s, where he showed an unusual aptitude for the public accounts. He was a Liberal, and a Liberal of the independent kind: he had opposed the federalist ideas of Florentino González in 1853, sat briefly at Rionegro in February 1863, and spent the following decade as a consul in Le Havre and Liverpool, writing essays on politics and society. He returned in 1874 and was elected president in 1879, taking office in 1880.

His first presidency was orderly and constructive. He founded the Banco Nacional, raised tariffs modestly to protect local industry, began the Bogotá–Girardot railway, and promoted arbitration in international disputes — a programme that would have fitted any Liberal government of the period. What was new was his conclusion that the constitution of 1863 could not be made to work. Elected again in 1884, he offered the Radicals a negotiated reform of the charter; they refused, suspecting a trap, and in August 1884 Liberal radicals in Santander, led by Solón Wilches and then by Sergio Camargo, rose against him.

The war of 1884–85 was the last and the most destructive of the federal wars. The rebels took Honda and Barranquilla, which gave them the Magdalena, and besieged Cartagena; the government, with Núñez's encouragement, called in the Conservatives, who provided both troops and the political base for what became the National Party. The decisive battle was fought at La Humareda, near El Banco on the Magdalena, on 17 June 1885, and it destroyed the rebel army. Within days Núñez announced from the balcony of the presidential palace that the constitution of 1863 no longer existed; the phrase is recorded in several forms and dated variously to 17, 20 or 23 June, but its meaning was unambiguous. The war ended in November with the capitulation of the last Liberal generals, Foción Soto and Siervo Sarmiento. The usual estimate is about 10,000 dead.

Núñez then did what Mosquera had done in 1861: he governed by decree and convened a body to write a constitution. The Council of Delegates had eighteen members, nine independent Liberals and nine Conservatives, chosen by the states, and it met in November 1885. The "bases" it adopted on 1 December 1885 were submitted to the municipalities and approved, which was the government's substitute for a constituent assembly elected by the people. Núñez's speech of 11 November 1885 set out the doctrine: the nation was one and indivisible, the states had been a fiction, and the Catholic religion was the moral foundation of the society the state governed.

The Regeneration was therefore not a Conservative restoration in the ordinary sense. Its author was a Liberal who had despaired of Liberal federalism, and its instrument was a coalition, the National Party, in which Conservatives supplied the ideology of order and the Church and independents supplied the administration. That is why the new settlement was so durable: it gave the Church what it had lost, gave the provinces a central government that could be talked to, and gave the oligarchy a constitution that kept the vote narrow and the presidency long.

What did the constitution of 1886 change?

The constitution was promulgated on 5 August 1886 — some texts give 4 August — and it began by invoking God as the supreme source of all authority, in deliberate contrast to the godless charter of 1863. It had 210 articles in 21 titles. Its first article did the essential work in one sentence: "The Colombian Nation is reconstituted in the form of a unitary Republic." The nine sovereign states became departments. Their governors were appointed by the president for three years, were removable by him at will, and answered to him rather than to their assemblies. Their assemblies survived, but they met every two years, had no control over the governor, and could be overruled by him.

The presidency became the centre of the system: six years, elected by electoral assemblies, with a vice-president and a Council of State, and with the power to appoint and dismiss governors, to declare a state of siege and to intervene in the departments. The Senate was elected three members per department for six years; the Chamber of Representatives for four. Article 38 made the Catholic, Apostolic and Roman religion "the religion of the Nation", to be protected by the public powers as "an essential element of social order", while conceding that the Church was not and would not be official; Article 41 provided that public education would be organised and directed in accordance with Catholic religion. Article 29 restored capital punishment for a list of ordinary crimes — treason in a foreign war, parricide, murder, arson, gang robbery, piracy and certain military offences — while Article 30 kept political offences exempt from it.

Two details show how much the new men had learned from the old ones. Article 201 put Panama under the direct authority of the national government, administered by special laws: the isthmus was too valuable to be left to a departmental assembly. And Article 202 took back for the nation the mines, the salt deposits and the public lands that had belonged to the states, along with the union's assets as they stood on 15 April 1886. The federation had been a confederation of proprietors; the republic would be a proprietor itself.

The Concordat signed with the Holy See on 31 December 1887 completed the settlement. It gave the Church control of Catholic education, the registration of marriages, the cemeteries and the mission territories, and it made Catholicism the religion of the state in everything but name. The constitution of 1886, with the reforms of 1910, 1936, 1945, 1957 and 1968, governed Colombia until the constitution of 1991 replaced it: 105 years, the longest life of any Colombian charter.

Was the federal experiment a failure?

The Liberals who wrote about it afterwards thought so, in the sense that they thought it had been destroyed by its enemies rather than by its own contradictions. The Conservatives thought it had never been a state at all: for them Rionegro was the institutionalisation of anarchy, nine little republics with nine little armies, and 1886 was the restoration of the nation. Both accounts are partisan, and the modern historiography has been more careful.

On the evidence, the federation did several things well. It kept the peace between 1867 and 1876 and again, more precariously, between 1878 and 1884; it expanded schooling, extended the civil power over marriage and registration, kept a free press and a competitive politics in which parties fought elections they could lose; it produced a generation of administrators — Murillo Toro, Camacho Roldán, Salgar, Parra, Núñez himself — who understood public finance. Exports roughly doubled between 1870 and 1880. The country's elite learned to govern itself through elections, however restricted, rather than only through pronunciamientos.

It also failed at three things that a state must do. It could not tax: the Union had no direct revenue and no fiscal relation with its citizens, so it could not pay for the army, the schools or the railways that its programme required. It could not arbitrate: the Supreme Court had no means of enforcing its judgments against a state, and the Senate had no means of settling a disputed election. And it could not demobilise: because the states kept their forces, every political dispute had an armed option, and the option was taken some forty times. Federalism did not cause Colombia's nineteenth-century violence; it gave the violence a structure and a set of ready-made bases.

The comparison with the United States is instructive precisely because it fails. The American federation worked because the federal government had a tax, a court with jurisdiction over the states, and, after 1865, a monopoly of organised force. The Colombian federation had none of the three. What it had instead was a written constitution, a free press and a political class that preferred to bargain; that was enough to last twenty-three years and not enough to survive a determined president with a foreign-trained army and a party behind him.

The argument did not end in 1886. The constitution of 1991 elected the governors, gave the departments and municipalities a share of national revenue, and created a Constitutional Court with the power to strike down national legislation — the three things Rionegro had wanted and could not build. Colombians are still arguing about whether that settlement is government or the prelude to disintegration, which is the argument the men of 1863 started.

Common myths about the United States of Colombia

Timeline of the United States of Colombia

What the map shows next

At 1863 on the map the map draws the union as it was constituted in May, with the nine states inside one outline and the federal district cut out of Cundinamarca; at 1864 on the map the same territory is shown in the year the census was taken and the Mosquera presidency ended. At 1865 on the map the outline is still labelled the United States of Colombia, and that is where the drawn sequence stops: from 1866 the same borders carry the label Colombia, so the map cannot show the last twenty years of the federation, the war of 1876–77, or the Regeneration.

The states that came before and after are drawn separately: the Granadine Confederation to the west of 1863, and the Republic of Colombia from 1886, which kept the same territory under a unitary government. Reading the two together is the quickest way to see what 1886 changed: not a border, but the location of sovereignty.