Duverger’s Law Is the Founders’ Revenge

Political scientists have a name for something the Founders built without realizing it: Duverger's law. Usually, it is cited as an indictment of the American two-party system-proof that our elections are too narrow, too rigged, too "undemocratic." I want to turn that argument on its head. Duverger's law is not a bug in our constitutional republic. It is one of the strongest confirmations that the Founders' design works exactly as Madison intended-and that attempts to "fix" it are central planning in its purest form.

The Mechanism Most Critics Misread

Duverger's law, named for the French political scientist Maurice Duverger, holds that single-member district plurality elections tend to produce two-party systems. In plain English: if only one candidate can win each district and the winner needs only the most votes, voters do not want to throw their ballots away on a candidate who cannot win. Over time, they gravitate toward the two viable options. Third parties are squeezed out-not by conspiracy, but by arithmetic. By contrast, proportional representation systems allow small parties to win seats in proportion to their vote share, so many parties survive.

The United States, with single-member House districts and a winner-take-all Electoral College, is the textbook case. Duverger's law is a tendency, not a law of physics-Canada and the United Kingdom have had durable regional third parties-but the American structure makes the pull toward two parties especially strong.

Here is what the critics miss: the Founders did not need Duverger to understand the logic. They had Madison.

In Federalist 10, Madison argued that an extended republic with many competing interests would make it harder for any one faction to gain a majority and oppress others. His exact words: "Extend the sphere, and you take in a greater variety of parties and interests; you make it less probable that a majority of the whole will have a common motive to invade the rights of other citizens." Duverger's law is the missing mechanism. It forces those many interests to come together into two broad coalitions if they want to win power. A small, intense faction cannot simply take a plurality in a few districts and govern; it must persuade a majority coalition across an enormous, diverse country. That is exactly the safeguard Madison designed.

The Safety Valve Nobody Planned

But the real story-the part too rarely discussed-is what Duverger's law did inside the American political machine: it gave us the primary system, and the primary is the safety valve of the republic.

Duverger's law does not mean there are only two viewpoints. It means there are two governing coalitions. The real competition-the place where ideological diversity actually lives-is inside those coalitions, in the primaries. Because third parties cannot win general elections, ambitious politicians and insurgent movements fight for control of one of the two major parties. The Republican Party is not a monolith; it is a coalition of nationalists, free-marketeers, social conservatives, and populists. The Democratic Party is a coalition of progressives, labor unions, environmentalists, and identity groups. These factions battle in open, often brutal primaries, and the voters decide which faction leads the coalition.

The primary system was not designed by the Founders. It emerged over more than a century as a spontaneous response to the constraints of Duverger's law and the demand of ordinary Americans for more control over party nominations. That makes it a perfect example of what Friedrich Hayek called spontaneous order-an institution no central planner designed, but which evolved because it solved a real problem: how to keep ideological diversity alive inside a two-party structure.

Look at the last decade. Donald Trump did not build a third party. He routed the Republican establishment in the 2016 primaries from inside. Bernie Sanders did not form a third party; he pulled the Democratic Party left by running in its primaries. In both cases, disruptive movements were channeled into the two-party structure, forcing the major parties to respond to voters rather than splitting off into impotent fringe parties. That is not suppression. That is representation-of the most direct kind.

When Reformers Promise More and Deliver Less

Now consider the reformers who want to abolish the Electoral College, adopt ranked-choice voting, or move to multi-member districts. Their stated goal is to make government more "representative" by breaking the two-party system. But they ignore the seen and the unseen, as Henry Hazlitt would say.

In proportional systems, elections do not produce a government. They produce a legislature full of small parties, and then party leaders negotiate a coalition in back rooms after the votes are counted. Voters do not choose who governs; they choose a party list, and the real decisions are made by professional politicians trading ministries for votes. Small parties with five or ten percent of the vote become kingmakers, extracting concessions no majority ever approved. Accountability collapses. When the coalition falls apart, the government falls too-weeks of crisis, new elections, more instability.

Belgium went 541 days without a full government after its 2010 election because the coalition talks broke down. That is what "more choice" looks like in practice: more parties on the ballot, less control by the voters, and a government formed by horse-trading among elites rather than by the people's decision at the polls.

This is the knowledge problem applied to politics. Hayek taught that no central planner can gather the dispersed knowledge of millions of individuals to run an economy. The same is true of elections. No central planner can design an electoral system that perfectly translates the scattered preferences of millions of voters into a governing majority. The two-party system is a spontaneous order. It emerged from the incentives created by the constitutional structure, not from a blueprint. It works because it forces voters to make a binary choice between two broad packages, and it gives them the power to throw the rascals out wholesale. Proportional representation promises more "choice" but delivers less control.

Third Parties Still Matter by Threat, Not by Winning

Does the two-party system silence dissent? No. It disciplines it. Third parties still matter-not by winning, but by threatening to take votes. The Populists of the 1890s forced the Democratic Party to adopt many of their ideas. The Progressive movement worked inside both parties. Ross Perot's 1992 campaign won nearly 19 percent of the popular vote but zero electoral votes-a perfect illustration of Duverger's law-and forced both parties to take deficit reduction seriously in the 1990s.

The difference is that third-party ideas must win a majority coalition to govern, rather than holding the entire system hostage as a small faction in a multiparty parliament. That is the discipline of the republic: if an idea cannot persuade a broad cross-section of Americans, it should not be handed the levers of power by a backroom coalition agreement.

The Real Trade-Off

Thomas Sowell reminds us that there are no solutions, only trade-offs. The trade-off of the two-party system is that you must compromise within a broad coalition. The trade-off of proportional representation is that you get more parties on the ballot but less accountability and more instability in government. I know which trade-off the Founders would choose.

The ordinary American does not need a parliament of a dozen parties negotiating away his liberty behind closed doors. He needs a constitutional republic where elections produce clear winners, where a majority coalition must be broad enough to respect minority rights, and where he can throw the bums out if they fail. Duverger's law is not a defect of that republic. It is the proof that the Founders' design-limited, federal, and republican-is still doing its job.