Why Do We Only Have Two Major Parties?

For the last 150 years people in the United States have complained about the fact that American politics is dominated by the two major parties. Hundreds of third parties have popped up over the years, only to fizzle out without gaining much traction. Why is that? Some people blame their struggle to fund raise, recruit candidates, and get attention through debates. While these factors play a role, there is a much deeper, more systematic reason why they cannot break out. This page will first discuss voting systems that reinforce two-party domination (plurality, Instant Runoff Voting), and then discuss voting systems that lead to multi-party democracy (Top-2 Runoff, proportional representation, and scored voting).

Duopoly

Why Plurality Leads to Duopoly

Our current voting system, known as single-winner plurality, virtually guarantees a duopoly, or a domination by two major parties. This fact has been established both theoretically and through broad historical evidence. In political science this phenomenon even has a name - Duverger's Law.

The 20th century French sociologist Maurice Duverger observed the fact that almost all countries using a plurality voting system tend to have a two-party system, while those using other voting methods such as Proportional Representation or a Two-Round Runoff tend to have multi-party systems. This can be explained by the following attributes of first-past-the-post voting systems.

Vote Splitting and the Spoiler Effect

When two ideologically similar candidates run, they effectively split the vote that otherwise would have gone to only one of them. In this situation, one of the candidates is said to have "spoiled" the election for the other. You will see later that this is a problem with any voting system that does not allow voters to support candidates equally.

Wasted Votes

Voting for a candidate that is seen as having little chance of winning is considered a "wasted vote" since the voter is giving up the chance to express their preference between the candidates that are seen as having a chance of winning.

Polarization

Voters tend to gravitate to either the candidate most likely to win, or the opposition candidate most likely to keep the front-runner from winning. Any kind of division of a voting block into factions virtually guarantees defeat.

Is Ranked-Choice Voting The Answer?

Over the past few decades, the most common form of single-winner Ranked-Choice Voting (RCV), known as Instant Runoff Voting (IRV), has grown in popularity in the United States. Proponents of this method claim that it is the most viable path to ending the two-party domination of American politics. There are really two major schools of thought among RCV proponents who seek to end the duopoly: those who see it as a stepping stone to proportional multi-winner RCV known as the Single Transferrable Vote (STV), and those who see IRV as the end goal. 

Instant Runoff Voting As A Stepping Stone To STV

The most well-informed proponents of IRV are largely in this first camp (see Fair Vote). Ultimately, they see Proportional Representation (PR) voting methods as the best way to break up the duopoly, and they see the Single Transferrable Vote (STV) as the PR method that is most politically viable in the US given America's general distrust of political parties. It's very difficult to go straight from the simplicity of single-winner plurality to the complexity of STV, so many in this camp see IRV as a good stepping stone. Unfortunately, in their well-intentioned long-term push for STV, which really does break two-party dominance, they must advocate for IRV, which on it's own is a voting method that reinforces two-party dominance.

Currently, the largest push toward implementing STV in the US is a bill called the Fair Representation Act. This bill would combine single-seat House districts into larger multi-winner districts with 3-5 seats each. STV would then be used to elect candidates according to their proportion of support within each district.

Instant Runoff Voting As The End Goal

The largest group of IRV proponents are those who see it as the ideal voting system which will end the duopoly and significantly improve American politics. This is a largely naïve claim based on common misconceptions and is not supported by historical or scientific evidence.

Historically IRV Has Never Broken The Duopoly

The first most important point to make is that IRV has been used for periods of time in various places around the world (Australia, Irleland, Fiji, Canada, Sri Lanka) for over a hundred years, and it has preserved two-party dominance in every case. This is largely because IRV is just multiple rounds of plurality voting with candidate elimination in between each round, so it suffers from many of the same psychological factors as regular plurality voting. A perfect illustration of this can be found in Australia which uses proportional STV for elections to its upper legislative body, and IRV (known there as the Alternative Vote) for the elections of its lower legislative body. While various political parties compete for seats in the upper chamber, the lower house is always dominated by two political groups.

Breaking Down Why IRV Doesn't End The Duopoly

The biggest misleading claim made by IRV proponents is that it eliminates the spoiler effect. This is true only if you define spoiler candidates as those who don't have a real chance of winning. Once there are three or more competitive candidates, IRV suffers from the same vote splitting issues as plurality because, as previously stated, IRV is really just multiple rounds of plurality voting. There are various mathematical and psychological factors that come into play here, but the two most prominent are called the "Center-Squeeze" effect and the "Favorite Betrayal" phenomenon. These two issues with IRV illustrate why neither moderate nor conservative/liberal third parties frequently win in an IRV system.

Moderate Third Parties And The Center-Squeeze Effect

The Center-Squeeze Effect is the phenomenon under various voting systems where moderate candidates are disadvantaged by losing support from the right and left while candidates on each extreme only lose votes from one side. This effect is most clearly illustrated by a hypothetical election in which three strong candidates are evenly distributed along a typical left-right spectrum. In the example below, candidate B would win head-to-head against either candidate A or C, but in an IRV election they are the first to be eliminated because they have the fewest first place votes. In such a scenario, the moderate candidate would have to get to a point where they would beat both of the other candidates head-to-head by huge margins before they even survive the IRV elimination round. While IRV is still better for moderate parties than plurality, it's only slightly better.

Center Squeeze Illustration 

The following video discusses a few different voting methods including IRV (4:47), and gives a great illustration of the center-squeeze effect (6:23). Note, near the end of the section discussing IRV, the presenter shows Australian election results to show that IRV does help third parties win legislative seats. Unfortunately, the results shown are from the Australian Senate which actually uses proportional multi-winner RCV (STV) and not IRV. In the Australian House, where they do use IRV, third parties frequently don't win any seats at all despite the proportionality of the Senate.

Conservative/Liberal Third Parties And The Favorite Betrayal Phenomenon

The "Favorite Betrayal" phenomenon occurs when a voting bloc supporting a preferred candidate over an acceptable one actually causes the bloc's least preferred candidate to win. This is most commonly seen when a third party on the far right or left grows enough to cause the major party that is ideologically closest to them to be eliminated first in IRV. 

To understand this better, consider that any third party on the right or left must go through 3 stages of growth under IRV.

  1. The first stage is when a third party's support is small enough for it to be eliminated before the two major parties. In this stage, third parties can be safely ignored by the major parties since they no longer spoil elections, allowing them to grow. This is the stage that IRV proponents point to when they say IRV helps third parties by allowing them to grow, and helps major parties by preventing spoiled results.
  2. The second stage is when the third party successfully beats one major party (surviving elimination) but ultimately loses head-to-head against the other major party. For a third party on the right or left edge of the political spectrum to grow, it typically must pull almost all of its support from the major party ideologically closest to it, making that major party the one most likely for the third party to beat first.
  3. The third and final stage is where the third party successfully beats both major parties and wins the election. This stage is almost always preceded by the previous two stages.

The second stage is where the favorite betrayal phenomenon is seen, and it acts as a firewall against the growth of third parties on the right or left. Consider the voting scenario below in which the Green Party (third party) has grown enough to eclipse the Blue Party (major party) in support. The Blue Party is eliminated, and Blue voters' ballots are redistributed between the remaining two candidates. Since these votes split between the two remaining candidates, the Red Party receives enough secondary support to get over the 50% threshold and win the election. The key point here is that had just a few more Green voters held their nose and voted for the Blue candidate, they would have elected someone acceptable to them. By supporting their favorite (Green) candidate, they caused the relatively moderate Blue party to be eliminated first and the election to be thrown to the Red candidate, since Green hadn't yet grown enough to enter stage three and win head to head. At best, this kind of election result understandably causes Green voters to return to the closest major party in the next election, making the Green Party's continued growth unlikely. At worst, such results cause voters to lose faith in IRV (and voting reform) all together and return to the voting system that they know best - plurality.

First Round IRV Election Results

FavoriteBetrayalRound1.png     

Second Round IRV Election Results

FavoriteBetrayalRound2.png

Summary of IRV and the Duopoly

As illustrated by examples above, IRV is not friendly to third parties either in the middle or on the edges of the political spectrum. Moderate parties have to get to a point where they would win in a landslide head-to-head against both major parties before they even make it to the final round, and more conservative/liberal third parties have to continue to grow despite potentially throwing elections to their supporters' least favorite major party. These are both different manifestations of the spoiler effect which is not eliminated by IRV. 

Ultimately, single-winner RCV (IRV) is not a voting method that leads to breaking the duopoly. While it is a good stepping stone to multi-winner RCV (STV), which does break two party domination, STV proponents are not typically transparent about their long-term goals, and have thus misled many IRV activists into thinking that IRV itself will lead to a thriving multi-party democracy in the US, while in the reality it maintains the status quo. Whether IRV's unfulfilled promises will lead most people to take the next step to proportional STV or become disillusioned by voting reform in general remains to be seen.

Breaking the Duopoly

Ending the two party domination of American politics has long been priority among those who feel that the current system stifles competition and limits voter choice. If the current political duopoly is a result of our method of voting, what changes would allow the US to move to a thriving multi-party system? Luckily there is a large body of research and hundreds of years of electoral history from democracies around the world to give us insights on how it has been done in the past, and how it could be done in the United States. Ultimately there are various possible paths: first, there is strong historical evidence that multi-party democracies emerge when either a single-winner top-two runoff or a multi-winner proportional representation method of voting is used; and second there is strong theoretical evidence that a single-winner rated voting system (Approval, Score, STAR) would also lead to a multi-party system.

Single-Winner Top-Two Runoff Voting

This voting method is familiar to most people in the United States. It simply uses plurality voting for the first round, and if a candidate gets over 50% of the vote then they win. If no candidate reaches this threshold, the top two finishers advance to a runoff round that happens some weeks or months later which ultimately decides the winner.

Many voting reform advocates believe that this voting system is just a simplified version of Instant Runoff Voting (IRV) but with more expense and hassle. This is not true. A recent analysis of elections using IRV and the Top-Two Runoff (T2R) method show a stark difference in how often the ultimate winner differs from the first round (plurality) winner. It found that while IRV changed the outcome only 5% of the time, T2R changed the outcome 17% of the time. This is due to a key psychological difference between the methods. With T2R, voters can honestly support their favorite candidate in the first round, knowing that they will be able to later vote again for their honest preference between the top two in the runoff. Conversely, with IRV, there is strong motivation for voters to rank their least favorite candidate lowest, and the candidate most likely to beat that least favorite candidate highest in an effort to magnify the effect of their vote (the NESD property). This difference, along with the center-squeeze and favorite betrayal effects described above explain why IRV maintains two-party domination but T2R does not.

These factors explain the party dynamics that exist in countries that use these two different methods. Using the T2R voting method, France recently elected centrist third-party candidate Emmanuel Macron as president. In comparison, Australia uses IRV for electing members of its lower legislative body and in three recent consecutive election cycles comprising 450 individual elections, not a single third-party candidate won.

Proportional Representation

Proportional Representation (PR) describes a family of multi-winner voting methods used by a majority of representative democracies around the world that elects representatives according to vote percentages. For example, in a 5-winner district with roughly 40% support for party A, 40% support for party B, and 20% support for party C, a PR voting method would allocate 2 seats to party A, 2 seats to party B, and 1 seat to party C. There are many factors at play when considering what flavor of PR to use, such as how many seats are given to each district (district magnitude), whether voters vote for parties or candidates, whether minimum support thresholds should be used to avoid giving seats to small extremist groups, and what mathematical formula is used to actually allocate the seats. Setting these factors aside, all PR methods generally accomplish the goal of allocating seats according to vote percentage.

PR voting methods have led to healthy multi-party democracies around the world simply because third parties do not have to reach a majority in any one region to successfully obtain representation in government. Since multiple parties win seats in the legislature, it's uncommon for one party to win a majority of seats, requiring parties to form governing coalitions to accomplish legislative goals.

Single Transferrable Vote

The most popular PR voting method among electoral reform advocates in the US is a form of multi-winner ranked-choice voting called the Single Transferrable Vote (STV). Like with IRV, STV uses a series of rounds where the candidate with least first-choice support is eliminated and candidates reaching a certain winning threshold (called a quota) are elected. The major difference between IRV and STV is that once a candidate is elected, any surplus vote over the winning quota is also reallocated to other candidates. For example, let's say we are using STV to elect representatives in a 4-member district. The winning quota is typically calculated as 100% divided by the number of seats plus one: 100%/(#seats + 1). In this district the winning quota would come out to be 100%/(4+1) = 20%. Just like with IRV, voters would rank all of the candidates according to preference. Let's say that none of the candidates reaches the winning quota in the first round, and so an elimination would occur just like with IRV. The candidate with least first-choice support is eliminated, and their support is reallocated to their voters' second choice candidates. Let's say that in the second round a candidate finally surpasses the winning quota with 23% of the vote. That candidate is now elected to a seat, and the 3% of the vote by which they surpassed the winning quota is reallocated to those voters' second choice candidates. These rounds of either losing-candidate elimination or winning-candidate election with surplus reallocation continue until all seats have been filled.

STV is the preferred PR voting method by many in the US because it allows voters to more fully express their preferences using a ranked ballot. It also limits the influence of political parties because voters support candidates directly instead of just voting for their preferred party. In fact, there is a bill in the US House of Representatives right now called the Fair Representation Act that would require all House Representatives to be elected in multi-winner districts using STV. It also requires that US Senators and the President be elected using IRV. Even though only House members would be elected using a PR voting method, it would still be a good step toward breaking the duopoly.

Single-Winner Score Voting Methods

Score voting methods (also known as rated, range, or cardinal voting methods) are those that allow voters to independently rate each candidate like they might rate a movie or a restaurant. The ratings for each candidate are added up, and the highest scoring candidate wins. Popular variations of this voting method family include regular Score Voting, Approval Voting which uses a scale of 0-1, and Score Then Automatic Runoff (STAR) voting which uses a 0-5 scale with an instant top-two runoff round.

Unlike other single-winner methods, score voting methods help break the duopoly because they allow voters to give maximum support to both their favorite as well as the lesser-of-two-evils candidate. This ensures that despite most voters employing the Naïve Exaggeration Strategy (giving minimal support to hated candidate A, and maximal support to candidate B who is most likely to beat candidate A), candidates seen as less viable can still receive full support from voters.

While scored voting methods are relatively new, there is a growing body of academic research showing that scored voting methods eliminate the mathematical and psychological factors that maintain two party domination.