# Duverger's Law **The model:** Plurality-rule elections in single-member districts (the "first-past-the-post" system used in U.S. House and Senate races, U.K. parliamentary elections, and most U.S. state offices) tend to produce two-party systems. Proportional-representation systems tend to produce multi-party systems. The mechanism runs through both elite incentives (small parties have no hope of winning seats and rationally merge or fold) and voter incentives (voting for a third party is a wasted vote that may help your worst-preferred candidate win). Duverger himself called it a sociological law rather than a deterministic one — there are exceptions (Canada has had viable third parties under FPTP; India is multi-party despite FPTP). But the strong tendency is well-established empirically and matters enormously for understanding why American politics looks the way it does. ## The Mechanism **Mechanical effect.** Under FPTP, only the plurality winner wins. A party that gets 15% of the vote in every district wins zero seats. A party that gets 30% in every district probably also wins zero. Only parties that can place first somewhere translate votes into seats. Small parties either build to plurality status or die. **Psychological effect.** Voters anticipate the mechanical effect and avoid "wasting" votes on parties that can't win. This drives third-party support down further than its sincere preference base. The Green Party and Libertarian Party in U.S. presidential elections persistently poll worse than they would in head-to-head questions about their platforms. **Center of gravity.** Both major parties under Duverger's Law have strong incentives to compete for the median voter (cf. [[Median Voter Theorem]]) within the constituency — except where structural features (closed primaries, gerrymandering, geographic sorting) reposition the median for general elections. ## KB Applications ### Why "Just Vote Third Party" Is Strategically Limited The recurring frustration with the two major parties produces calls for third-party voting (libertarian, Green, "neither of the above"). Duverger's Law is the explanation for why such calls almost never produce a sustained third-party coalition: the system structurally punishes the strategy. The Reform Party (1996), the Tea Party caucus (which absorbed *into* the GOP rather than forming a new party), the [[Lincoln Project]]'s "anti-Trump GOP" coalition (which has not become a party), and the various "No Labels" efforts all illustrate the pattern. ### Why Capture-and-Hold Beats Splitting The right-wing populist movement chose to capture the GOP rather than form a third party. This choice was strategically sound under Duverger's Law: a captured major party retains all the structural advantages of two-party-ness, including ballot access, fundraising infrastructure, and the implicit guarantee that anti-major-party-Y votes will flow to it. The same calculation produced the choice — visible in real time — to consolidate inside the GOP rather than build outside it. Recognizing the Duverger-driven incentive is the first step in predicting whether dissident factions will leave or stay. ### Primary Capture as the Real Battlefield Under Duverger's Law, intra-party primaries become the locus of meaningful ideological competition. The [[Federalist Society]]'s judicial pipeline, [[Heritage Foundation]]'s policy infrastructure, MAGA's primary-challenge infrastructure — all are battles for the *party*, not battles for new parties, because Duverger's Law makes the latter pointless. This is why understanding GOP donors, primary turnout patterns, and primary-challenger funding is more analytically valuable than understanding general-election polling: the action that decides what the party stands for happens in the primary. ### Reform Proposals That Address the Law Ranked-choice voting (RCV), proportional representation, multi-member districts, and approval voting are all attempts to weaken Duverger's Law's grip. Maine and Alaska have adopted RCV at the state level; their elections are now empirically testable cases of whether the mechanism produces the predicted multi-party effects (so far: modest evidence yes). RCV-positive arguments routinely cite Duverger's Law as the diagnostic naming the problem to be solved. ### Caveats Duverger's Law has known exceptions. In Canada, regional concentration sustained the Bloc Québécois and the NDP. In India, regional and caste-based parties survive FPTP. The key variable is whether a party can be *first* somewhere — local concentration can substitute for national coalition. The law is a strong tendency, not a deterministic constraint. ## See Also - [[Median Voter Theorem]] — the companion model about *where* the two parties position themselves - [[Iron Law of Oligarchy]] — the within-party version of the structural-coalescence dynamic - [[Gerrymandering]] — the structural tool that distorts which median Duverger-driven parties chase