Do you consider yourself an informed voter? If so, understanding Duverger’s Law can empower you to make strategic choices in upcoming elections, including the November midterms, regardless of your party affiliation. In many undergraduate American Politics 101 courses, students learn that the U nited States has a “two-party system.” They study the histories of the Democratic and Republican parties, the shifting coalitions each has represented, their partisans’ general policy preferences, and the ideological differences between them. Yet considerably less attention is given to the electoral mechanics that preserve the dominance of these two large parties. So-called “third parties,” of course, are not outlawed from participating in the democratic processes of our constitutional republic. They may nominate candidates, organize supporters, appear on ballots, and compete for public office. Why, then, have third parties so consistently remained minor parties? The obdurate U.S. party “duopoly” does not endure because the electorate naturally sorts itself into two tidy political camps. On the contrary, the national motto, e pluribus unum (“out of many, one”), describes a nation forged in the crucible of profound social, cultural, and political diversity —not one whose individual citizens have ever fit squarely into two partisan categories.  This diversity is currently reflected, at least partly, in Americans’ growing reluctance to identify with either major party. Gallup reported in January that a record-high 45% of American adults identified as political independents in 2025, including 56% of Gen Z (born between 1997 and 2007) and 54% of Millennials (born between 1981 and 1996). Although many of these self-identified independents still lean toward one of the two major parties, these figures hardly suggest widespread attachment to the existing partisan choices. The continued dominance of only two nationally viable parties, then, cannot be explained solely by voter satisfaction, tradition, political culture, or media coverage. Instead, it is sustained and continually reinforced by the rules under which U.S. voters elect their public officials. This is where Duverger’s Law, named for the French political scientist Maurice Duverger (1917–2014), offers valuable insight into one of the most durable patterns in electoral politics: single-member districts using plurality voting tend to produce competition between two major parties. Let us begin unpacking this claim because, in its compressed form, it sounds much more complicated and perhaps less intuitive than it actually is. Duverger developed his principle during the mid-20th century, most famously in his 1951 book Les Partis Politiques, published in English in 1954 as Political Parties. Under plurality voting in single-member districts (also called first-past-the-post or choose-one voting), the candidate who receives more votes than any other candidate wins, even without securing an absolute majority. In a four-candidate race, for instance, a candidate could win with only 35% of the vote if their opponents received 30%, 20%, and 15%, respectively. At this point, the word “law” may still create some confusion. Duverger’s Law is not a legal rule enacted by a legislature or an electoral management body, nor is it an exceptionless iron command comparable to a mathematical certainty. Rather, it is a scientific law in the social-scientific sense: a concise statement describing a repeatedly observed relationship between particular conditions and a likely result. Scientific laws identify regular patterns, but they do not necessarily imply that every individual case will unfold identically. Newton’s law of gravitation, for example, continues to describe the gravitational attraction between objects even when other forces, such as air resistance, affect their motion. Duverger’s Law likewise identifies a persistent tendency in certain electoral systems while allowing for exceptions arising from regional parties, unusual political divisions, electoral alliances, or other institutional arrangements. Duverger’s central prediction thus remains remarkably useful for understanding U.S. electoral politics and the two-party system. Within this system, we have the following conditions: a) only one candidate can win a district, and b) victory requires only a plurality of votes (as described above, despite our majoritarian intuitions about electoral fairness, a candidate need not receive more than 50% of the vote to win). The result: political competition usually consolidates around two viable choices. The logic hinges on two related forces. The first is a mechanical effect. In “winner-take-all” elections, such as the presidential contests in nearly every state under the Electoral College, finishing second, even by a slim margin, yields no electoral votes from that state. Similarly, a political party could receive 40% of the vote in every single-member congressional district and still win no seats. Its support would be substantial but insufficient to secure office, as it never finished first. Proportional representation, one of the most widely used electoral systems among democratic nations today, works differently. If a party receives 15% of the vote under a genuinely proportional system, it may receive approximately 15% of the seats, typically through multi-member districts. Winner-take-all elections in single-member districts provide no comparable reward or representation. The second force is a psychological effect. Voters, candidates, donors, and political organizations anticipate the mechanical effect and adjust their behavior accordingly. For example, voters may strategically vote for a major-party candidate to avoid “wasting their vote,” and candidates may align with one of the two major parties to increase their chances of winning, thereby reinforcing the two-party system. For this reason, promising candidates often join one of the two major parties rather than build a new one. Donors follow suit, hesitating to support campaigns regarded as incapable of winning from the get-go. Journalists concentrate on the two leading candidates. Voters may even abandon a preferred third candidate in favor of a less-preferred but more viable alternative. The result becomes a self-reinforcing dialectic. Third parties struggle because voters doubt their viability, while voters doubt their viability because third parties struggle. The 2000 presidential election provides the most famous example of Duverger’s Law at work in the United States. George W. Bush’s certified margin over Al Gore in Florida was just 537 votes (roughly 0.009%) of the nearly six million votes cast in the state. Because Bush finished first under Florida’s winner-take-all rules, he received all 25 of the state’s electoral votes. Meanwhile, Green Party candidate Ralph Nader received 97,488 votes (1.64%) of the statewide total —far more than Bush’s margin of victory. If one assumes that most Green Party voters were ideologically to the left of the Democratic Party and hence preferred Gore to Bush, their votes for Nader may have helped elect the candidate they preferred least. In other words, voting for their first choice may have actually contributed to their least-preferred outcome. This is commonly known as the “spoiler effect,” in which support for a candidate with little to no chance of winning changes the outcome between the leading candidates. Nevertheless, it would be inaccurate to claim that every Nader voter would otherwise have supported Gore. Some would have voted for Bush, chosen another candidate, or simply stayed home. The election also notoriously involved ballot-design problems, disputed recount procedures, voter-access controversies, and the Supreme Court’s consequential 5–4 decision in Bush v. Gore (2000). Still, Florida demonstrates the strategic risk created by plurality voting. When only a few hundred votes separate two leading candidates, support for a third candidate can help determine which of the leading candidates wins —even when that third candidate has no realistic path to victory. Moreover, the ultimate victor may be further removed from third-party voters’ political preferences than the major-party candidate they chose not to support. The 2016 election supplied another example. Donald Trump defeated Hillary Clinton in the swing state of Michigan by a margin of 10,704 votes (0.23%). Green Party candidate Jill Stein received 51,463 votes (1.07%) of the nearly 5 million votes cast in the state, while Libertarian Party candidate Gary Johnson received 172,136 votes (3.59%). The Michigan ballot also included candidates from the U.S. Taxpayers’ Party and Michigan’s Natural Law Party, who together received less than 0.40% of the statewide vote. Those third-party voters by no means belonged to Clinton. Nor can analysts assign all the possibly right-leaning Libertarian votes to Trump. The figures instead reveal the razor’s edge upon which presidential elections may be decided. Of course, the strategic problem can affect either major party. In 2020, Joe Biden defeated Trump by 10,457 votes (0.3%) in Arizona, 11,779 votes (0.23%) in Georgia, and 20,682 votes (0.63%) in Wisconsin. Libertarian Party candidate Jo Jorgensen received more votes than Biden’s margin in each of those states. Again, one cannot assume that every Libertarian voter would have chosen Trump in a two-candidate race. Yet a voter deciding among a favored third-party candidate, an acceptable major-party candidate, and an unacceptable major-party candidate should understand the possible consequences. The worst consequence is helping elect the candidate the voter opposes most. Be that as it may, Duverger’s Law does not prove that third-party voting is foolish, immoral, or illegitimate. In a political system animated by deliberation, third-party campaigns can introduce neglected ideas, pressure major parties, provide political homes for dissatisfied citizens, and build organizations for future elections. A protest vote can also serve a defensible purpose. Some voters use the ballot not merely to select an officeholder but to register opposition to the available choices. In a state or district where the outcome is virtually certain, the immediate risk of a “spoiler candidate” may be small. However, in a closely contested election, such a vote may be criticized as politically reckless or insufficiently attentive to its likely consequences. Indeed, some political scientists might call this sort of sincere voting behavior “naive” rather than “strategic.” Therefore, voters should distinguish between two different uses of a ballot. An expressive vote says, “This candidate most closely represents my beliefs.” A strategic vote asks, “Which available choice gives me the best chance of producing an acceptable outcome or preventing an unacceptable one?” Sometimes the same candidate satisfies both purposes. Sometimes the voter must choose between them. Someone whose highest priority is building a third party may rationally vote for a candidate who is unlikely to win. Someone whose highest priority is defeating a particular major-party candidate, however, is better off voting for the most viable alternative. This argument offers no defense of either major party. But it does explain why many voters feel as though they are stuck choosing between “the lesser of two evils.” Democrats and Republicans both benefit from a structure that enables their duopoly by disadvantaging emerging competitors. Yet both have also lost close elections in which minor-party candidates attracted potentially pivotal votes. Now, Connecticut voters reading this in anticipation of August 11 should note that Duverger’s classic model does not apply as neatly to party primaries. Plurality primaries may encourage strategic voting: a voter may support a second-choice candidate if the preferred candidate appears unlikely to defeat a disliked rival. At the same time, crowded primaries may sustain several competing ideological or regional factions within the same party, making the outcome much harder to predict. Americans who want more than two viable parties in general elections, however, need not wait for millions of partisan voters to become more adventurous. They can change the electoral rules that sustain the two-party structure. Ranked-choice voting allows citizens to rank candidates in order of preference. A voter can support a third-party candidate first while identifying an acceptable major-party candidate as a backup. If the first-choice candidate is eliminated, the ballot transfers to the next-ranked candidate. This approach can reduce the spoiler problem, although it does not guarantee the development of a multiparty system. Proportional representation offers a more direct route. Legislators may be elected from multi-member districts, with seats distributed according to each party’s share of the vote. A party winning 40% or 15% of the vote could then gain meaningful representation rather than repeatedly finishing second or third and receiving no representation at all. Other reforms could include runoff elections, easier ballot access, more inclusive debates, fusion voting, and public financing to support competitive campaigns by political outsiders. Connecticut already has experience with one alternative, namely “cross-endorsement” or “fusion voting.” Under this system, a minor party may endorse the same candidate as a major party while maintaining its own ballot line. For example, a voter can support the Working Families Party, the Independent Party, or another qualifying organization without splitting the total vote received by the endorsed candidate. The Connecticut State Elections Enforcement Commission (SEEC) also administers the Citizens’ Election Program, “allowing candidates to compete without reliance on special interest money.” That said, while such arrangements demonstrate that the current system is neither natural nor inevitable, they do little to create a genuinely competitive multiparty democracy in which third-party candidates have a realistic chance of winning office and their supporters have a meaningful chance of gaining representation. In a word, electoral systems shape political behavior. Winner-take-all plurality elections reward large coalitions capable of finishing first and punish smaller organizations whose support is widely dispersed. Voters respond to those incentives whether they recognize Duverger’s Law or not. Still, before entering the voting booth, voters should ask more than, “Which candidate do I like best?” They should also ask, “What am I trying to accomplish with this vote, and how do the rules translate my choice into political power?” Duverger’s Law cannot answer that question for them. But it can help them understand the decision they are making. Robert T.F. Downes is a PhD candidate in the Department of Political Science at the University of Connecticut, specializing in political theory and public law. He is also an Adjunct Professor at Rhode Island College. ...read more read less