Duverger case study: The United States in the 19th Century

Duverger's Law predicts that Single Choice Voting in single-winner elections will produce two dominant parties or camps. The United States offers the canonical historical demonstration — compressed into two consecutive presidential elections, 1836 and 1840, separated by a single four-year lesson.

The two-party system is often treated as a natural feature of American democracy. It is not. It was designed — deliberately, by one man, Martin Van Buren — as a rational response to the coordination problem that single-choice voting creates. The two elections that followed proved him right.

The first U.S. presidential election in 1789 was radically different from what American elections would become. There were no official candidates, no campaigning, no political parties, and no nominating conventions. The Founding Fathers disapproved of "factions", viewing organized parties as dangerous to the public interest.

In the early 1790s, informal alignments emerged — Federalists backing a strong central government, Democratic-Republicans favouring states' rights and agrarian interests — but these were schools of thought, not organized parties. The Federalist–Republican rivalry dissolved after the War of 1812, leaving the Democratic-Republicans as the only national political grouping. By 1820, President James Monroe ran almost unopposed — a period historians call the Era of Good Feeling, when the collapse of organized party rivalry made it briefly seem as if America might escape the party system altogether.

The surface unity masked deep fractures. Four distinct economic regions — the industrial North, the plantation South, the frontier interior (today's Midwest), and the expanding cotton territories further south and west — had conflicting interests on tariffs, the national bank, land policy, and slavery.

The Electoral College does not always produce a winner on its own. The Constitution requires a candidate to win an outright majority of electoral votes — in 1824, that meant 131 out of 261. When no one reaches that threshold, the decision passes to the House of Representatives, where each state casts a single vote regardless of its population, and a candidate must win a majority of states to take the presidency. In the early decades of the republic, with multiple regional candidates and shifting coalitions, this fallback was a real possibility every four years — and exploiting it would become the centrepiece of the Whigs' 1836 strategy.

When those fractures erupted in the 1824 election, the coordination failure was total: four candidates from the same Democratic-Republican tradition split the vote, no one won an Electoral College majority, and the House of Representatives handed the presidency to the candidate who had come second in both the popular vote and the Electoral College. It is the last time, to this day, that a presidential election has been decided in the House.

That failure produced the man who would solve it. Martin Van Buren, Senator from New York, understood — without yet having the vocabulary of Duverger — that Single Choice Voting punishes the side that divides its vote. His solution was the modern political party: disciplined, coordinated, loyal above all to its own continuity.

In a letter to a leading Virginia newspaper editor in January 1827, Van Buren laid out the strategy with unusual candour. America needed permanent, disciplined parties built on a clear ideological divide. The old Federalist-vs-Republican framework — elites versus the people — would supply the organizing principle. Andrew Jackson's enormous personal popularity would supply the glue. Without that structure, Van Buren argued, Jackson's supporters would again fragment across regional interests, just as they had in 1824, and lose again.

He built the Albany Regency — a tightly disciplined party machine in New York — on the principle that party loyalty must override individual judgment. This was not corruption. It was the rational response to the voting method's arithmetic: the side that divides loses. Van Buren engineered the solution before anyone had formally stated the problem.

The Democrats: one convention, one candidate

By 1836, Van Buren's Democratic Party had mastered party discipline. The 1835 Democratic National Convention nominated Van Buren as the single national candidate — Andrew Jackson's handpicked successor. The party ran one ticket, everywhere, consistently.

The result: Martin Van Buren, 50.83% of the popular vote.

The Whigs: a deliberate split

The Whig Party's response was not disorganization — it was strategy. Unable to agree on a single national candidate from their loose coalition, the Whigs made a calculated gamble: run regionally strong candidates in different parts of the country, deny Van Buren an Electoral College majority, and force the decision into the House of Representatives — where, under the Constitution, each state casts a single vote regardless of its size — hoping to negotiate a win there.

The combined Whig popular vote was approximately 49% — nearly equal to Van Buren's. The strategy came terrifyingly close to working. The decisive state was Pennsylvania, worth 30 electoral votes: Van Buren won it by only 4,222 votes, a margin of 2.4%. Had Harrison carried Pennsylvania, Van Buren's electoral total would have fallen to 140 — eight votes short of the 148 needed — despite winning more than half the national popular vote.

And even if the Whigs had succeeded in denying Van Buren an Electoral College majority, their strategy would still have failed at the next stage. The Democrats controlled 14 of the 26 state delegations in the House of Representatives — enough to give Van Buren the presidency there regardless of how the Electoral College had voted. The Whigs were trying to exploit a constitutional mechanism that the same Democratic majority would have blocked one step later.

This is cycle one of the textbook Duverger's Law sequence. One side has unified behind a single candidate. The other side has split — deliberately and rationally — yet still lost, despite representing nearly equal popular support. The lesson was now available for anyone willing to learn it.

The Harrisburg Convention

The Whigs held their first true national nominating convention in Harrisburg, Pennsylvania, in December 1839. Three candidates stood: Henry Clay, the party's founder and ideological soul; William Henry Harrison, a war hero with broad Northern appeal; and Winfield Scott.

Clay was the philosophically coherent choice. He had founded the Whig Party, championed the national bank, and articulated the "American System" of protective tariffs and internal improvements. He led on the first ballot. But the logic of the convention overrode the logic of policy.

Convention rules — winner-take-all by state delegation — rewarded candidates with concentrated majority support within states rather than broad support spread across them. Clay had support everywhere but majorities nowhere. Harrison's support was more concentrated in key Northern states. Several Southern delegations that favoured Clay did not send delegates at all. After five ballots, Harrison won: 148 votes to Clay's 90 and Scott's 16.

This is the first documented instance of a party choosing who it thought it could win with over who its members actually preferred — the electability calculation that would define American party conventions ever after. The Duverger's Law mechanism had migrated inward, shaping the candidate field before a single public ballot was cast.

The Result

Harrison ran as a war hero and man of the people. The slogan: "Tippecanoe and Tyler Too!" The Whigs avoided discussing the issues that divided their coalition — slavery and the national bank — and concentrated on exploiting Van Buren's unpopularity after the financial crisis of 1837. The result speaks for itself:

This is cycle two. The previously split side has unified. It wins. The previously unified side now loses. The sequence — one side splits and loses; learns the lesson; unifies and wins — is identical to the hypothetical example in Duverger's Law. Except this one happened two centuries ago, in the real world, with real consequences that have never been undone.

From this election onward, no major American party would again field more than one presidential candidate. Two decades later, one would try.

From Whigs to Republicans

The Whig Party never managed to settle on a coherent answer to the most explosive question of the 19th century: slavery. By the early 1850s, the party had fractured along sectional lines and collapsed. Northern anti-slavery Whigs, joined by anti-slavery Democrats and former members of smaller protest parties, founded a new party in 1854: the Republicans. Their platform was unambiguous — slavery must not expand into the new western territories.

The mid-1850s briefly looked more fragmented. A third movement — the American Party, better known as the Know Nothings, a nativist organisation built on hostility to Catholic immigration and opposition to both existing parties — had surged in 1854–55, winning governorships and congressional seats. But the slavery question was swallowing everything, and the Know Nothings had no coherent answer to it. By 1856 their moment was passing.

The Republican Frémont came second; the Know Nothing Fillmore a distant third in electoral votes. The Republican Party had inherited the Whigs' role: the second pillar of the two-party system. Duverger's Law had reasserted itself with a different pair. The Whig–Democrat axis had become the Republican–Democrat axis, and the structure was set.

The Democratic Schism

The 1860 Democratic National Convention met in Charleston, South Carolina, in April 1860. Slavery was, again, the issue that no compromise could contain. Northern Democrats, led by Stephen Douglas, supported "popular sovereignty" — letting each new territory decide for itself. Southern Democrats demanded a federal slave code that would protect slavery in every territory, regardless of local preference.

When the convention adopted the Northern position, fifty Southern delegates walked out. The convention deadlocked and adjourned without nominating anyone. It reconvened in Baltimore in June. The Northern wing nominated Douglas. The Southern wing held its own rival convention and nominated John C. Breckinridge, the sitting Vice President of the United States.

For the first time since 1840, a major American party would field two competing presidential tickets. A fourth party — the Constitutional Union Party, formed from former Whigs and Know Nothings who wanted to bury the slavery question and "preserve the Union" — nominated John Bell of Tennessee. The Republicans, meeting in Chicago, nominated Abraham Lincoln.

The Four-Way Race

The 1860 election produced four sectional candidates with four sectional bases:

Lincoln became the only president in American history to win an Electoral College majority with under 40% of the popular vote. The arithmetic of his victory is the textbook Duverger's Law demonstration. The combined Democratic popular vote — Douglas's 29.46% plus Breckinridge's 18.10% — came to 47.6%, more than Lincoln's 39.82%. But split between two tickets, that combined Democratic support translated into only 84 electoral votes against Lincoln's 180. A unified Democratic ticket might well have denied Lincoln the presidency.

A Lesson Burned In

The cost of the Democratic split was not measured in one election. Within weeks of Lincoln's inauguration, eleven Southern states had seceded. Four years of civil war followed. Six hundred thousand Americans died. The lesson was burned in.

Since 1860, no major American party has fielded more than one presidential candidate. The Republican–Democratic binary that emerged from that election has held continuously to this day. The names have not changed. The coalitions have shifted dramatically — the Republicans of Lincoln are not the Republicans of the 21st century, and the Democrats of Breckinridge are not the Democrats of today. But the structure has not moved: two parties, single tickets, Single Choice Voting, and the iron rule that you run one candidate or you lose.

The lesson the Whigs learned from 1836 was not the lesson that needed to be learned. They concluded: coordinate or lose. The correct lesson was: single-choice voting produces this dynamic, and the voting method should be changed. No one proposed changing the voting method. The Knowledge-Application Gap operated at the deepest level — the disease was understood well enough to navigate, but its root was never addressed.

The two elections produced one further irony. The Whigs chose Harrison over Clay precisely for electability. Harrison won — then died of pneumonia 31 days after his inauguration. The vice-presidential slot had gone to John Tyler — a Virginia states-rights man added to the ticket to win Southern votes — who turned out to be no Whig at all. He had been a Democrat, and his convictions remained those of one. He promptly vetoed the entire Whig legislative agenda and was expelled from the party. The coordination solution worked. The governance problem remained.

In the more than a century and a half since 1860, no third party has won a presidential election. Nearly every major reform movement — from the Progressive Era to the New Deal to the civil rights movement, from the Tea Party revolt to the MAGA movement — was ultimately absorbed into one of the two existing parties rather than breaking through as an independent political force.

The Duverger Syndrome was not an accident of American political culture. It was engineered, confirmed in two elections, and perpetuated by the same Single Choice Voting method that made it inevitable in the first place. The cure — replacing single-choice voting with a method that lets voters express genuine preferences across the full range of candidates — was always available. Two centuries later, it has still not been applied.