Posted this question on Reddit. I am crossposting it here as well: https://www.reddit.com/r/AskSocialScience/comments/vpnviu/polarization_and_electoral_systems/

Is there research linking polarization of the electorate and the electoral system employed (FPTP, preferential vote, proportional representation)? How a particular electoral system causes or prevents polarization in the medium to long run? Can anyone provide some references, theoretical or empirical?

  • Please edit the question to limit it to a specific problem with enough detail to identify an adequate answer.
    – Community Bot
    Commented Jul 2, 2022 at 13:48
  • 2
    Not sure if it's even possible to measure polarization in PR systems. Since it would require partitioning voters into two diametric poles. Yet with so many political parties in PR systems with different orientation, it's hard to measure where they are on a single spectrum. Commented Jul 2, 2022 at 14:15
  • 1
    I would recommend reading "Why We're Polarized" by Ezra Klein though. Commented Jul 2, 2022 at 14:16
  • 1
    Pretty related politics.stackexchange.com/questions/33069/… Commented Jul 2, 2022 at 21:16
  • 1
    @amrods I think he talks about electoral systems in terms of cause and remedy to polarization. The description might be more qualitative rather than quantitative. Commented Jul 3, 2022 at 1:20

6 Answers 6


Some that I know of off the top of my head:

These model the polarization of voters and their elected representatives, and show that the adoption of top-two runoff voting did not improve polarization:

  • Kousser, T., Phillips, J., & Shor, B. (2018). Reform and Representation: A New Method Applied to Recent Electoral Changes. Political Science Research and Methods, 6(4), 809–827. https://doi.org/10.1017/psrm.2016.43
  • Ahler, D. J., Citrin, J., & Lenz, G. S. (2016). Do Open Primaries Improve Representation? An Experimental Test of California’s 2012 Top-Two Primary. Legislative Studies Quarterly, 41(2), 237–268. https://doi.org/10.1111/lsq.12113

While the voters' ideological ideal points are distributed in a rough bell curve shape, the representatives are polarized (I took the graphs from the paper and summed them to make the bell curve shape):

A graph of political ideology of voters and their elected representatives

“Our analysis provides a clear lesson for the immediate impact of California’s twin electoral reforms of 2012: neither the Citizens Redistricting Commission nor the top-two primary immediately halted the continuing partisan polarization of California’s elected lawmakers or their drift away from the average voter in each district.” (Kousser et al., 2018, p. 16)

“By comparing the congruence of voter positions with legislator positions before and after the implementation of the new primary and redistricting rules, we can assess whether these reforms improved representation in the state’s assembly, senate, and congressional delegation. In an analysis that focuses on California’s congressional delegation, we find no evidence that reform improved representation. If anything, California’s congressional candidates and eventual lawmakers became a bit more ideologically extreme in 2012, thus moving further apart from the average voter in their district.” (Kousser et al., 2018, p. 2)

“To determine whether open primaries lead voters to choose ideologically proximate candidates, we conducted a statewide experiment just before California’s 2012 primaries, the first conducted under a new top-two format. We find that voters failed to distinguish moderate and extreme candidates. As a consequence, voters actually chose more ideologically distant candidates on the new ballot, and the reform failed to improve the fortunes of moderate congressional and state senate candidates.” (Ahler et al., 2016, p. 237)

These talk about the differences in political violence found in countries that have majoritarian systems vs proportional representation:

“In the empirical analysis, I show how the level of inclusiveness of the political system is an important factor in reducing the probability of civil war even for democratic countries. To capture the notion of inclusiveness, I have used data on Checks and Balances (CHECKS) from the DPI database constructed by Keefer et al. (2001). Using this dataset, I find that the more inclusive system has a lower probability of civil war. Moreover, I find that proportional representation systems are associated with high levels of inclusiveness and majoritarian and presidential systems to low levels of inclusiveness.” (Reynal-Querol, 2005, p. 462)

This one isn't directly empirical, and isn't directly about polarization, but shows the typical representativeness of the winners under different multi-winner systems based on simulations and a 2D spatial model of voter behavior (which has been shown to produce realistic distributions of ballots):

  • Elkind, E., Faliszewski, P., Laslier, J.-F., Skowron, P., Slinko, A., & Talmon, N. (2019). What Do Multiwinner Voting Rules Do? An Experiment Over the Two-Dimensional Euclidean Domain (arXiv:1901.09217). arXiv. http://arxiv.org/abs/1901.09217

Probability distributions of winner committees under various multi-winner voting systems in a 2D space

This is my own amateur original research, but in the same vein as the above paper, you can simulate the likelihood of candidates winning across the 1D political spectrum under different single-winner systems, and see that the cruder systems that suffer most heavily from vote-splitting tend to elect less representative candidates, and can even have a pit in the middle showing a bias against the most representative candidates in the middle of the voter ideology spectrum:

Probability distributions of winners under various single-winner voting systems in a 1D space

  • This is very illustrative. Thank you for taking the time to answer my question in a thorough way. Can you include what the acronyms in the last 2 graphs mean?
    – amrods
    Commented Jul 3, 2022 at 10:47
  • 1
    @amrods Added them
    – endolith
    Commented Jul 3, 2022 at 10:58
  • 1
    @amrods Updated the last diagram and added some more abbreviations
    – endolith
    Commented Jul 8, 2022 at 14:09

There are some Yale lectures by Ian Shapiro discussing how primaries increase polarisation, and single seat electorates (compared to proportional representation) reduce polarisation.

A working paper details most of the same content, with academic citations.

The basic ideas are:

  • A winner takes all system incentivises presenting more moderate platforms (as a race for the median vote) analogous to final offer arbitration. In contrast, proportional representation leads to conventional bargaining (in parliament), whereby exaggerated initial stances are advantageous (leaving more space to negotiate).
  • Governance involves balancing competing goals (for example: reducing debt, improving services, and reducing tax each sound popular when framed in isolation, despite being collectively inconsistent). In winner takes all the party has to present a compromise and is held wholly responsible for the trade-offs, whereas it is viable to hold extreme positions (on particular issues) in proportional representation or direct democracy (and no entity has long-term accountability for the interactions between issues).
  • Primaries bias candidates to be less moderate (representing the party base more than the electorate's median voter).
  • Healthy parties should have leaders that are responsive to voters (such as if the backbench can oust failing leaders) and backbenchers that are well disciplined (so that the leadership can negotiate with credibility and can deliver on difficult compromises). Discipline requires mechanisms for the leadership to reward and punish individual backbenchers (e.g. with promotion into executive or vetoing from pre-selection). Primaries undermine the leadership's ability to discipline their own members of parliament, letting some instead promote more populist positions.

Of course, a two-party (winner takes all) system provides less obvious avenues for an interest group to bring their own issue to attention.

  • 1
    I think you may be confusing polarisation with fragmentation, at least in that last regard. He's talking about Duverger's law etc., which is not about polarization but fragmentation. The relationship between polarization and fragmentation is alas not so straightforward (or studied), see the linked Q. Commented Jul 6, 2022 at 5:47
  • 2
    I see later he says (around 24:30) that in a two-party systems the positions of the parties (on issues) will converge (reduce polarization), but obviously that doesn't hold much water empirically (see the studies in the linked Q). So it may be held true in a political science 101 class, but not under more scrutiny. Commented Jul 6, 2022 at 5:57
  • 1
    He does say later "that is true up to a point"... and gives some example and counter-examples. The counter-examples actually exceed the examples on the slide at 28:15, but that's not exactly a scientific method as they are cherry picked (in both categories). See the linked Q for some attempts to rectify that approach. Commented Jul 6, 2022 at 6:03
  • 1
    @Fizz that link you provided above talks about polarization and fractionalization of parties. I am not sure how much those aspects relate to polarization of the electorate. I am not a political scientist, are they closely related? On one hand, parties should reflect the opinions of the electorate, but on the other I think that parties tend to amplify differences.
    – amrods
    Commented Jul 6, 2022 at 6:27
  • Well, in most studies on polarization (that I know of) the position of the electorate is taken as fixed, and the reflection thereof on the political parties is studied, e.g. as additionally a function of electoral system etc. I think there are very few studies that touch on the evolution of the positions of the electorate itself. The mechanisms of transmission of polarization "the other way around" i.e. from parties (or mass media etc.) to the electorate are far less studied, as far as I know. Commented Jul 8, 2022 at 9:45

I have to agree with another answer's frame challenge although not with the conclusion.

Essentially an all-or-nothing voting model can pave a way towards polarization (by enabling it), but it doesn't have to force it. In other words, it doesn't have to make it inevitable.

Even in 2-party system there is a lot of room for compromise if the minority party has enough actual power to force the majority party to make concessions. It is the lack of incentives for concessions that raises the stakes. When the stakes are too high, the fight gets more personal and it gets much dirtier.

Charles Krauthammer, who was a big critic of earmarks, conceded (shortly before his death) that he was wrong because these bribes, that the 2 parties had to pay each other to get bills passed, that allowed for the parties' cooperation (a cynic would say "collusion").

As a simple study, one may look at what happened to the Supreme Court confirmations. Removing filibuster from the judges' confirmations meant that the majority party no longer had to compromise with the minority party.

Before the filibuster was removed, the confirmations (in most cases) were passed with much higher margin than the 60/40 required to break the filibuster. After the filibuster was removed, the passage margins had actually shrank while the rhetoric got much more personal.


Regarding the theses of the Shapiro lecture (which doesn't cite much studies), the one about primaries increasing polarisation is somewhat known, see e.g.

Also, the simple model that predicts that two parties will converge on a single issue is called the Downsian model, and goes back to 1957. This is sometimes incorrectly used to infer that two parties will thus so converge on any number of issues.


Essentially all of the other answers basically assume political science models in which polarization is mostly a product of voting systems, essentially prejudging the question being asked.

But, the historical evidence isn't so clear.

For example, in the U.S. from the end of the U.S. Civil War in 1865 until at least sometime in the early 21st century, white voters outside the South have been much less polarized than in the South, where "racial block voting" in which native Southern white voters overwhelmingly vote for one political party which is deeply divided from another political party that captures almost all non-white voters and a significant share of white migrants to the South from outside the South. The election system was essentially the same in both thee South and outside it. But, what differed was the cultures, ethnicities and religions of the people participating in that same electoral machinery in different U.S. regions.

Recent papers discussing this is in the U.S. context include Yongren Shi, et al., "Cultural Fault Lines and Political Polarization" (2017) and Duane A. Alwin, "The Changing Dynamics of Class and Culture in American Politics: A Test of the Polarization Hypothesis" Annals of the American Academy of Political and Social Science (December 10, 2015). Supporting a mixed cultural and institutional analysis in the U.S., cf., David Blakenhorn, "The Top 14 Causes of Political Polarization" The American Interest (May 25, 2018).

Internationally, see, e.g., Simon A. Levin, et al., "The dynamics of political polarization" PNAS (December 6, 2021) (supported by 11 separate articles in the same issue reviewing the question). Narratives of polarization focused on the role of the mass media and social media are also common. See, e.g., JungHwan Yang, et al., "Why Are “Others” So Polarized? Perceived Political Polarization and Media Use in 10 Countries" 21(5) Journal of Computer-Mediated Communication 349–367 (September 13, 2016).

The polarization or lack thereof in different regions and demographics of the United States fundamentally reflects the deep cultural and economic divisions (often traceable back to the populations involve in the initial colonial settlement of different U.S. regions, see David Hackett, "Albion's Seed: Four British Folkways In America" (1989)) under the underlying populations and are far less a result of moulding of the political landscape as a result of the electoral systems that were in place.

Political polarization also tends to decline during times of extreme crisis that is viewed as external in nature, like FDR's Democratic coalition during the Great Depression and World War II, and the Grand Coalition in the U.K. during World War II, and then to resurface when those outside threats pass. Frequently, as was the case on the eve of the U.S. Civil War, the deep polarizations reflect deep division over important issues that are rooted in cultural, economic, religious, and ethnic differences and fall fairly cleanly along cultural lines.

The U.K., Canada, and the U.S. all had single member district plurality voting systems for almost all of their modern political histories.

But in the U.K. and Canada, the deepest divisions have involved regional autonomy (Ireland prior to the Republic of Ireland, Northern Ireland, Scottish Independence, Brexit in the U.K.; Quebec independence in Canada), with a fair amount of consensus on political process issues and a fair amount of policy continuity with a change in partisan control. In contrast, in the U.S., deep polarization has taken a more partisan character.

Some of the most polarized societies that have attempted democratic systems, like Lebanon, Iraq, Northern Ireland, Bosnia, and Nigeria (and to a lesser extent, for example, Belgium) have reflected ethnic and religious divisions, to which electoral arrangements are at most a response rather than a cause.

The apparent association between less polarized political systems and proportional representation electoral systems may simply reflect the fact that proportional representation electoral systems are largely Western European, which is a region in which national boundaries and ethnic, religious and linguistic boundaries are mostly reasonably aligned, while this is not the case in most former European colonies that did not have wholesale population replacement in the way that Australia, New Zealand, and Canada did.

Ultimately, any electoral system funnels the preferences of voters into either a majority coalition (in parliamentary systems and lots of the time in strong Presidential systems), or into a "cohabitation" as the French call it, in which no one party or coalition has clean majority control (as happens frequently in the U.S. and France, as is designed into the political systems of Lebanon and Iraq as the usual outcome).

In a U.S. style political system with two dominant parties, that coalition arises before the election. In multiple party parliamentary political systems with proportional representation styled elections, that coalition is formed after the election. But, either way, the ultimate aim is for a political party or a coalition of political parties to control the institutions of government to the greatest extent possible, so it makes a certain amount of sense that polarization would only be weakly influenced by the electoral system.

What drives polarization is ultimately how clean the separation is between different factions and how much common ground the contending factions have when setting policy. When there are few swing voters or people who have mixed views not captured by any one party or coalition, and when the cleanly divided factions have little common ground, you have polarization. This isn't something that a political science model of generic political parties in a generic electoral structure can tell you.

When polarization has a strong regional character, you get secessionist movements. When polarized factions are hard to divide into cleanly divided contiguous regions, as is in the case in Northern Ireland, for example, you just get a highly polarized political environment.

  • You address the root causes of polarization. Can different electoral systems lower the level of polarization, if so which one? Presumably that depends on what root cause dominates, for example for factions hard to divide geographically.
    – amrods
    Commented Sep 15, 2022 at 11:48

I wrote a short article some years ago related to this question. I think that the two-party duopoly is largely responsible for the absurd level of polarization in the country and recently published a somewhat longer article on how that problem could be addressed by adopting a better voting system. In very brief form, the problem with our voting system is that it favors the big political parties and it demand a high level of name recognition. The problem is that voters are only asked which candidate they like and they have no opportunity to express themselves about which candidate or candidates they oppose. There is of course a middle position voters might have. A voter may be indifferent and perhaps has never even heard of a candidate.

A better voting system would allow voters to express opposition as well as support and it would allow voters to express these feelings about many, perhaps all, candidates. In a two-party duopoly, the two candidates would likely get nearly the same level of opposition as support, but when opposition is not express, both candidates have much more support than any new or minor party candidate could have. Minor party candidates might get support from voters who know them but would often have the advantage of relatively little opposition.

  • 3
    Whilst this may theoretically answer the question, it would be preferable to include the essential parts of the answer here, and provide the link for reference.
    – JJJ
    Commented Sep 13, 2022 at 19:15
  • 3
    Is this answer just here to advertise your published work?
    – Joe W
    Commented Sep 13, 2022 at 19:30

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .