Skip to content

Two-Party Dominance — the case for it, the case against, and what actually changes it

"Break the two-party system" is a rallying cry for a lot of voting reformers — but two-party dominance is not self-evidently bad, and plenty of thoughtful people defend it. A reform library that only prosecutes it isn't teaching; it's campaigning. So this page steelmans both sides, then draws the one distinction that actually matters for voting methods: what makes two-party dominance change, and what doesn't.

The case for two-party systems

These aren't strawmen — they're the serious arguments, and some are backed by real political science:

  • Governability. A two-party system usually hands one party a working majority, so a government can pass a coherent program instead of haggling a fragile coalition into existence. Fragmented multiparty parliaments can produce paralysis, frequent collapses, or caretaker governments (the cautionary tales: interwar Weimar, post-war Italy, some periods in Israel).
  • Accountability. Voters know who is in charge and whom to blame, and power alternates cleanly. No voter wakes up to a coalition government that no one actually voted for, assembled in back-room deals after the election.
  • Moderation (the median-voter pull). In classic Downsian theory, two big parties compete for the center, forcing each toward the median voter — a centripetal, moderating force. Both parties are broad coalitions that must aggregate diverse interests before the election rather than after.
  • No extremist kingmakers. In a fragmented legislature a small radical party can hold the balance of power and extract outsized concessions. Two-party systems deny fringe parties that leverage.
  • Simplicity. Two clear choices are low-information and easy to reason about.

The case against

  • Limited choice / "lesser of two evils." Large blocs of voters have no party that fits them; many feel they're voting against, not for.
  • Underrepresentation. Third parties and independents can win substantial vote shares yet almost no seats — a persistent votes-to-seats gap (e.g. Australia's Greens under the Alternative Vote).
  • Duopoly and safe seats. With only two viable brands and many uncompetitive districts, the two parties face little real competition and can quietly collude to keep others out.
  • Polarization. The moderating pull is a tendency, not a law — modern US politics shows a two-party system can instead calcify into intense polarization, negative partisanship, and gridlock, with primaries dragging candidates away from the center.
  • It can be manufactured, not chosen. When two-party dominance is propped up by the spoiler effect and vote-splitting, it may reflect the voting method punishing alternatives rather than the electorate's genuine preferences.

Where the scholarship lands

There's no free lunch. The standard framing (Arend Lijphart, Patterns of Democracy) contrasts majoritarian democracies (two-party, single-party governments, clear accountability) with consensus democracies (multiparty, proportional, broad power-sharing). Each buys something and pays for it: majoritarian systems trade breadth of representation for decisiveness and clarity; consensus systems trade decisiveness for inclusion and (Lijphart argues) at least as good policy outcomes on many measures. Which trade-off is "better" is a values question, not a settled fact — which is exactly why this page doesn't rule on it.

The part that's about voting methods (and the honest caveat)

Here is the distinction reformers most often blur. Two-party dominance has two different sources, and different fixes touch different sources:

  1. Spoiler-enforced dominance. Under choose-one plurality, a third-party vote can help elect your worst choice, so voters and donors abandon alternatives pre-emptively. This part is a voting-method artifact — and better single-winner methods remove it. STAR, Approval, Ranked Robin, and IRV all let you support a favorite without spoiling, so new parties can run and grow without being blamed as spoilers.

  2. Structural (Duverger) dominance. Even with the spoiler gone, single-member districts are majoritarian: each seat has exactly one winner, so the system still tends toward two big blocs. Australia's century under the Alternative Vote is the proof — no spoiler problem, and still two-party-preferred in the lower house.

So the honest claim, which cuts against a common overreach:

Better single-winner methods end the spoiler-enforced part of two-party dominance — they do not, by themselves, end two-party dominance. What reliably produces a multiparty legislature is proportional representation (STV, MMP, party-list, proportional STAR) — multi-winner methods, not single-winner ones.

This is why Australia is two-party in its AV House but multiparty in its STV Senate. If your goal is specifically to break two-party dominance, single-winner reform is necessary-but-not-sufficient; the lever is proportional multi-winner elections. If your goal is honest, spoiler-free single-winner elections (a worthy goal on its own), STAR/Approval/Ranked Robin deliver that whether or not the party system ever changes.

How do you count parties? (ENP)

"Two-party" and "multiparty" sound obvious until you try to measure them — a country with two giants and five specks isn't a seven-party system in any meaningful sense. The standard fix is the Effective Number of Parties (Laakso–Taagepera):

ENP = 1 / Σ(sᵢ²)      where sᵢ is party i's share of votes or seats

Squaring the shares makes big parties dominate the sum, so tiny parties barely move it. Four equal parties give ENP = 4; one dominant party with three specks gives barely above 1. In practice the US sits near 2, while highly proportional systems like Israel exceed 6.

Two things this buys you. First, it makes the claim above testable rather than rhetorical: you can check whether a reform actually changed the party system, instead of arguing about it. Second, it reframes the trade-off on this page as a dial, not a binary — the Lijphart argument is really about where on that dial a country wants to sit. Jameson Quinn has argued the sweet spot is around 3–4: enough parties to escape zero-sum, us-versus-them politics, but few enough that each must still assemble a broad coalition rather than campaign on one narrow issue. That figure is a judgment call, not a finding — but the metric is standard.

Takeaway

Two-party dominance is a genuine trade-off, not a self-evident evil — it buys governability and accountability at the cost of choice and representation, and reasonable people weigh those differently. And whatever you think of it, be precise about the mechanism: the spoiler effect is method-caused and fixable with better single-winner voting; the deeper two-party tendency of single-member districts is structural and only proportional representation changes it.