Can Proportional Representation Create Better Governance?

Policy outcomes, accountability, and stability under different electoral systems

Introduction

The United States faces a crisis of governance. Budget standoffs and threats of government shutdowns are now standard practice, as are partisan impeachment inquiries. Leadership instability has paralyzed the House of Representatives. Even when Congress legislates, it often produces policies at odds with public opinion¹. Partisan polarization yields gridlock when government is divided and potential policy oscillation under unified control.

Elections are democracy’s solution to dysfunctional governance, so the current state of American democracy raises the question: Is there a better way to elect our representatives?

Like many other scholars and reform advocates, we think proportional representation (PR) for electing House members would bring about significant improvements to American politics by doing away with zero-sum dynamics and better capturing the views of a diverse electorate. But proposals to switch to some sort of proportional electoral rules are often met with fears that PR elections would compromise governance. Proportional representation tends to evoke the chaotic legislatures of countries like Brazil and Israel, both of which have highly fragmented party systems, and raises questions about how to identify who is responsible for policy and how to hold them accountable.

Read the full white paper as a PDF Read the full white paper as a PDF

We address these concerns by showing how PR can improve governance in the United States. By governance, we mean three things: congruence (policy outcomes that people want); accountability (the ability to course correct by voting out bad politicians); and stability (consistency and predictability of government).

On all three criteria, research suggests a combination of mixed results and reasons to believe that PR could be an improvement for the U.S. compared to winner-take-all (WTA) elections. Particularly in a highly polarized society, PR could help improve congruence and accountability by reducing the impact of gerrymandering; making elections more competitive; giving voters more choice within and across parties; allowing for a greater representation of societal diversity; and promoting more consensus-seeking in the legislative process.

There is little reason to believe PR would create instability or make it difficult to hold politicians accountable. In contrast, there are many reasons to believe the WTA model in the U.S. is failing to ensure good governance. The conditions that allowed WTA to work alongside a presidential system—mainly, significant overlap between Democrats and Republicans—no longer exist. Today, the combination of presidentialism and WTA in the House of Representatives is proving dangerous and destabilizing, leading to gridlock in times of divided government and to unilateral—even extreme—policymaking in times of unified government². PR is better suited to the country’s conditions and could work well in combination with presidentialism, without sacrificing—and perhaps even enhancing—governance.³

Key Takeaways

How proportional representation affects governance, accountability, and stability

Proportional electoral systems are better at achieving what voters want.

Proportional systems are better at promoting consensus – especially in polarized societies – and achieving public policies that better reflect what majorities of citizens want.

Proportional electoral systems can enhance government accountability.

In polarized societies, some forms of proportional systems represent improvements over winner-take-all elections when it comes to holding politicians accountable at the ballot box by giving voters more choices within and across parties.

Proportional electoral systems can reduce political instability.

Proportional systems that maintain the number of parties at a moderate level can reduce the risk of political instability currently posed by the winner-take-all system in the United States.

Balance is possible between two parties and too many parties.

The choice of electoral reform is not a binary one between the two-party system that winner-take-all encourages and the fragmentation seen in very permissive proportional systems. PR systems can be designed to ensure a healthy balance between too few and too many parties, and can also facilitate good governance.

A Closer Look

Presidentialism and the Senate

Scholarship on PR focuses overwhelmingly on using it to elect national and subnational legislatures (variously known as assemblies, parliaments, congresses, etc.). When we draw from that scholarship to discuss prospects for PR in the United States, we are thinking about how PR would affect elections and representation in the U.S. House of Representatives and in the state legislatures. There are no constitutional impediments to PR in these contexts, so electoral reforms could be adopted by statute. But of course, the U.S. political environment is populated by other key actors as well, most notably the presidency and the U.S. Senate, both of which are constitutionally bound to election by winner-take-all methods.

Later in this essay, we discuss the combination of presidentialism with PR elections for the legislature specifically. For now, we just note that multiparty coalitions can form and operate under presidentialism. Indeed, most presidential democracies use PR elections for their legislative assemblies.⁴ Where legislative elections are proportional, winner-take-all competition for the presidency motivates multiparty coalitions to form in support of presidential candidacies. Presidents elected with the support of multiple parties, in turn, tend to allocate cabinet ministries and other appointed positions to coalition partners, and presidents tend to rely on support from legislators whose parties hold cabinet ministries. Incentives for cross-party cooperation are not as strong as in a parliamentary system, where coalition fracture can cause the government to fall, but they are not absent under presidentialism. Thus, we would expect coalition politics under multipartism to be more fluid than under a parliamentary system, but for coalitions to be meaningful political actors nevertheless.

Similarly for the U.S. Senate, the constitutional requirement that each state’s senators are elected in staggered terms effectively locks Senate elections into a winner-take-all format, which should constrain multipartism in the Senate but not prevent it altogether. Parties that demonstrate viability in House (or state legislature) elections could gain traction in Senate contests. Brazil combines list PR for its House elections with winner-take-all contests for Senate seats, yet 11 parties are currently represented in its Senate. Like the presidency, we anticipate the U.S. Senate to impose a centripetal pull within multiparty politics but not to obviate altogether the multiparty dynamics that PR would foster./p>

Achieving What People Want

We begin with congruence. As the political scientist G. Bingham Powell put it, “democracy should systematically induce the policymakers to do what citizens want them to do and avoid what the citizens dislike”⁵. More technically, researchers define congruence as the distance between the preferences or ideology of the median voter and the policies or positions taken by the government.

A long-standing piece of conventional wisdom among social scientists was that WTA encourages political parties to move to the center to maximize votes, converging on the preferences of the median voter⁶⁷⁸. Voters in WTA systems were also presumed to be well positioned to hold their representatives accountable, encouraging good performance among politicians and ensuring they do not deviate too much from what the people want⁹¹⁰. With only one representative from each district, the cognitive load of keeping track of a representative’s performance is minimized, making it easier for citizens to “throw the rascals out.”

These theoretical predictions are not playing out in the U.S. today, and they fit awkwardly with our frustrations with American politics. Gerrymandering and partisan sorting have created party strongholds that preclude partisan competition from disciplining and moderating politicians. Our system of primary elections selects candidates most appealing to highly mobilized partisans. Extreme partisan and affective polarization prevent voters from defecting from their party and voting for another one even if their representative is underperforming, weakening the threat of voting representatives out of office. The lack of partisan options beyond the Democratic and Republican parties means the country’s ideological, cultural, and demographic diversity gets flattened into a binary that fails to appropriately capture the array of views in the country. In short, the U.S. currently does not have the conditions for a WTA electoral system to fulfill its potential of linking citizen preferences with politicians and policies.

In this context, many scholars and activists propose that the United States reform its electoral rules to be more permissive, allowing multiple parties to obtain seats in the legislature in proportion to their vote shares. Advocates for proportional rules argue that a PR system would allow for more compromises in the policymaking process in a way that better reconciles diverse perspectives. But skeptics worry that PR could lead to instability and indecisiveness in Congress, and allow for small and fringe parties to have undue influence. They also claim that PR could weaken linkages between politicians and citizens, and that an increase in the number of representatives per district would weaken mechanisms of accountability by increasing the information demands on voters and making it harder to monitor politician performance.

The promise of representative democracy is that it is able to achieve a close correspondence between citizens and politicians. Democracies are configured quite differently, and plenty of research efforts have been devoted to understanding what electoral institutions—whether WTA or PR—result in greater congruence, both in terms of ideology and policy outputs. Theoretically, both families of electoral rules have properties that under ideal conditions should ensure congruence among citizens, politicians, and policy outcomes. Competition between two parties in majoritarian systems may motivate parties to appeal to the median voter, while competition among multiple parties reflecting a wider assortment of views in a proportional system can make the median party a pivotal player.

Both winner-take-all and proportional representation systems are capable of delivering policy outcomes in line with public opinion.

Early empirical studies of congruence gave an edge to PR systems in terms of achieving greater congruence between citizens and politicians¹¹¹²¹³. Most of these studies compared the difference between the median ideological placement of citizens in surveys and the ideological placement of political elites across different electoral systems.

More sophisticated approaches have looked at the overlap between the ideological distributions of citizens and elected representatives, and have found that parliamentary legislatures elected through PR tend to be more representative than those elected with WTA rules¹⁴. Focusing on comparisons across districts of different magnitudes – the number of representatives elected from each district – within Switzerland, which helps alleviate concerns that country characteristics other than their electoral systems are driving the effects, Carey and Hix¹⁵ demonstrate that Swiss legislators from cantons with low to moderate district magnitude (in the range of three to eight) more effectively represent district median voters than those from single-member districts (that is, WTA) or from larger multimember districts.

In short, the U.S. currently does not have the conditions for a WTA electoral system to fulfill its potential of linking citizen preferences with politicians and policies.

While these studies are mostly based on parliamentary systems, which lack a separately elected chief executive, they highlight the importance of coalition formation as a moderating force for achieving congruence between public opinion and policy outcomes. In parliamentary proportional systems, the need to negotiate among different parties in the legislature and include their diverse priorities tends to provide a safeguard against maximalist, unilateral policies. Similar dynamics are at play in proportional systems that work alongside presidentialism. In these systems, presidents usually need to negotiate with multiple parties in the legislature, and even include members of other parties in their cabinets as part of coalition governments, which prevents extremism in policymaking.

The experience of the United States shows that WTA models have historically worked at achieving a correspondence between policies and public opinion when polarization was low and there was some overlap between the main parties, but these conditions are increasingly disappearing¹⁶.

In polarized contexts, proportional representation systems are better suited at achieving congruence than winner-take-all.

Some research has challenged the advantages of PR systems at achieving better congruence, showing that both PR and WTA achieve similar levels of congruence. In making sense of these findings, Powell¹⁷¹⁸ points out that the similar performance of both families of electoral systems is largely due to the low levels of party-system polarization in the nineties in countries that use single-member districts. Powell argues that in countries with low levels of polarization, we should expect both WTA and proportional systems to achieve congruence and that polarization is especially pernicious in systems with WTA rules, as it discourages parties from shifting their stances to the political middle. This is an important insight for the United States, which currently combines high levels of partisan and affective polarization with single-member districts, creating conditions that are not suitable for WTA rules to fulfill their potential to moderate the parties, and achieve congruence between citizens and their politicians. Proportional systems, in contrast, seem to be better at achieving congruence even with a polarized electorate because moderate parties become pivotal players in coalitions¹⁹.

Proportional systems seem to be better at achieving congruence even with a polarized electorate.

Other than polarization, other contextual and institutional design features might complicate any simple binary WTA vs. PR comparisons and policy congruence. The design of PR systems can moderate the effect of electoral rules on congruence and representation. PR systems, for instance, vary in the number of representatives elected from each electoral district – their district magnitude. Carey and Hix²⁰ showed that district magnitude can affect congruence in surprising ways. In Switzerland, congruence improves as district magnitude increases from a single-member district but then diminishes as the district magnitude keeps increasing above six representatives per district.

District magnitude, and the size of the legislature, can also affect the number of parties in an electoral system, with higher magnitudes and larger legislatures associated with more parties to the extent that there are underlying cleavages in a District magnitude and the size of the legislature can also affect the number of parties in an electoral system, with higher magnitudes and larger legislatures associated with more parties to the extent that there are underlying cleavages in a society²¹²²²³. A large number of parties can strengthen the forces pulling parties outward in PR systems, resulting in a greater divergence between public opinion and political elites. To counteract an ever-growing number of parties, many electoral systems impose vote share thresholds that parties have to meet to continue competing in elections and these thresholds vary from country to country.

The choice of open or closed lists in a proportional representation system affects the responsiveness of representativeness and the types of policies they pursue.

The most common proportional systems used internationally are list systems. Within these systems, one of the most important design features is whether representatives are elected from open or closed lists. Seats are won by the party in proportion to their vote share and are then assigned to candidates based on their ranking in the list. In open lists, voters can indicate a preference for individual candidates in the party list and seats are allocated first to parties (in proportion to their overall tallies), then to the party candidates based on the preference votes they receive. In closed lists, parties put forward a list of candidates and voters vote for the party list without expressing any preference among the candidates and any seats won by a party are filled based on the ranking of the list created by the party.

The choice of open or closed list has important ramifications for congruence, representation, and policy outcomes. Under closed lists, competition for votes happens only among parties. Open lists foster simultaneous competition among parties and within parties, as candidates work to distinguish themselves from fellow party members. Competition within open lists to build a personal reputation can be in tension with maintaining a consistent party brand and with party cohesiveness²⁴²⁵. When combined with high district magnitudes, in particular, these incentives can encourage patronage-based politics²⁶²⁷. Open lists encourage attentiveness to district-level concerns, but incentives for localism should be balanced against broader policy priorities in the design of PR systems²⁸.

A Closer Look

Transferable Vote Systems and Proportional Representation

The range of electoral systems used worldwide is extensive, and the corresponding nomenclature can be confusing. Among advocates of electoral reform, the Single Transferable Vote (STV) system has substantial support. STV is a version of Ranked-Choice Voting (RCV), which has gained traction in recent years at the state and local level in the U.S. RCV generally refers to transferable vote elections in single-seat districts, whereas STV refers to transferable vote elections in districts where multiple seats are awarded. To voters, RCV and STV look the same: All candidates are listed on the ballot, and the voter indicates a ranked preference among candidates (first choice, second choice, third choice, and so on).

Transferable vote systems operate by establishing a threshold for first-choice preferences needed to win (50% in single-seat districts; a lower percentage in multiple-seat districts) and by sequentially eliminating candidates with the lowest levels of support, transferring the ballots indicating top preferences for those candidates to the columns of surviving candidates marked for lower-order preferences. Under STV, if a popular candidate wins more first-choice preferences than required to secure a seat, surplus support is reallocated according to the lower-order preferences on those ballots, so that surplus support may affect who wins other seats.

STV is sometimes described as proportional because its operation in multiseat districts means that if voters rank within their preferred parties, then candidates from multiple parties can win representation in proportion to their share of the votes. The mechanics of STV are not proportional; they are sequential elimination of candidates. A ballot for a given candidate benefits other candidates from the same party only if the voter ranks those specific candidates next. Under list PR, by contrast, the first step in translating ballots into representation is to determine the proportion of ballots for each party, and only after the partisan (and proportional) allocation of seats is determined are seats distributed to specific candidates within lists.

In practice, STV can provide representation that looks and feels something like list PR insofar as districts often elect representatives from more than one party, with their relative numbers reflecting the distribution of voter support across candidates. Under list PR, the pooling of votes across all the candidates from each party encourages team-oriented behavior among co-partisan politicians, whereas STV competition is every candidate for themself, which encourages individualism. But voters can, if they wish, rank all the candidates from their most favored party in succession on their ballots. And many list PR systems—particularly open list varieties—can encourage individualism, steering the incentives of politicians, and the kind of representation they deliver, in a similar direction as STV.

In short, STV is mechanically distinct from list PR but—depending on the details of how each system is configured—they could deliver similar incentives for politicians and similar results for citizens.

When does proportional representation make a difference?

Both WTA and PR can achieve similar levels of congruence given the right conditions. However, the choice of electoral systems can lead to different policy outcomes because of the different effects they exert on citizen preferences and the composition of party coalitions.

Citizen preferences that drive voting behavior are themselves shaped by electoral rules²⁹. To win in WTA systems, parties need to build a broad winning coalition, which means they are better off activating large social cleavages. In PR systems, parties can activate various smaller cleavages and be successful electorally. Depending on the nature of the cleavages, citizens in WTA systems may vote based on different preferences from citizens in PR systems. Moreover, the activation of a few cleavages in WTA systems may pull apart the electorate, resulting in greater polarization and a diminished role for the political center. In contrast, the activation of multiple cleavages in proportional systems can strengthen the position of center parties as pivotal players.

The different types of coalitions that WTA and PR systems encourage lead to different policy outcomes. Most of the historical and contemporary research on the topic indicates that the choice of electoral system makes a difference when it comes to levels of taxation and redistribution. Because PR systems tend to empower center parties in coalitions and encourage politicians to reach out to a wider population, PR systems are associated with higher levels of redistribution, with more broadly targeted (as opposed to those with regional or group-specific benefits) and higher levels of social welfare spending.

Proportional representation systems favor the median voter and programmatic policies, resulting in higher levels of progressive redistribution.

The variation in citizen preferences can affect the composition of party coalitions across electoral systems. Iversen and Soskice³⁰ argue that the structure of coalition-building in multiparty systems strengthens the bargaining power of the party representing the median voter relative to a two-party system. The logic begins with the observation that wealth distributions are universally skewed such that the median voter’s wealth is below the mean level, and the median voter should favor moderately progressive redistribution of wealth. In a two-party system, however, centrists will be on the margins of each major party, either of which might govern more radically than the centrists prefer. Iversen and Soskice posit that centrists fear the prospect of radical redistribution more than no redistribution, driving them toward a center-right alliance in a two-party environment. Under PR, by contrast, centrists can control a moderate party which, in turn, can be the pivotal partner in coalition governments, opening the door to reliable alliances with pro-redistributionist forces on the left.

Another feature of PR systems that favors redistribution over majoritarian systems is district magnitude and the need to reach wider constituencies. In WTA systems, as long as elections are sufficiently competitive, legislators ought to be motivated to deliver benefits to their local districts, and the election of only one representative per electoral district facilitates credit-claiming for policies that can be geographically targeted.

Many studies have documented that proportional systems experience less inequality than WTA systems—and that as the degree of proportionality increases, inequality tends to fall.

Higher-magnitude districts in proportional systems, by contrast, make credit-claiming for geographically targeted benefits challenging because many legislators represent any given geographical district. This favors broader, programmatic policies that benefit a wider population. As Alesina, Glaeser, and Sacerdote³¹ put it, in the extreme example of a proportional election in a single national district, there is no need for geographic targeting at all. Moreover, because parties win more seats with additional votes, their motivation is greater than in WTA systems to reach out to a broad constituency within and across electoral districts, instead of focusing only on pivotal voters in swing districts. This is more easily achieved with policies that benefit a wider group of people than with more targeted policies.³²

A number of studies have provided evidence that proportional systems provide policies more aligned with what the median voter wants: more progressive redistribution, higher levels of social welfare spending, lower overall levels of income inequality, and more programmatic policy spending than electoral systems with majoritarian rules. Iversen and Soskice³³ show that countries using proportional systems have higher levels of redistribution towards the poor. Higher redistribution contributes to lower levels of income inequality, and many studies have documented that proportional systems experience less inequality than WTA systems—and that as the degree of proportionality increases, inequality tends to fall³⁴³⁵. Alesina, Glaeser, and Sacerdote³⁶ argue that the absence of a generous welfare system in the United States is due, in part, to the combination of racial animosity towards Black citizens, and WTA electoral institutions that have discouraged the rise of left-wing parties and diluted the electoral power of the poor more generally. Proportional systems are also theorized to produce more public goods than WTA systems³⁷, and some research has shown that proportional systems have better health outcomes than WTA ones³⁸.

A methodological challenge present in all these studies, and in claiming that PR leads to more progressive redistribution, is that the choice of proportional rules in Western European democracies has historically been due to pressure from left-wing parties or where the structure of the economy encourages negotiations across economic sectors³⁹⁴⁰⁴¹. An important challenge to the narrative linking PR to progressive redistribution contends that the same set of forces produced both the adoption of PR and the differences in redistribution across electoral systems. Paulsen⁴² addresses this challenge by focusing on a reform in Norway that kept magnitude constant while requiring municipalities with more than one representative to adopt proportional representation. Paulsen finds that the municipalities forced to switch to proportional representation increased taxation and resources invested in the poor, and attributes this change to increases in political mobilization and moderate politicians trying to preempt threats from the left. Recent work from Górecki and Pierzgalski (2023) reinforce this conclusion based on cross-national patterns of redistribution and New Zealand’s switch from WTA to PR in the 1990s.

As with research comparing congruence across electoral systems, research on the effects of electoral rules on redistribution and other outcomes tend to flatten the various institutional design features of proportional systems, such as district magnitude or the choice of open or closed lists. The comparison between proportional systems and majoritarian systems, and the incentives that the latter creates for targeted geographic policies, would suggest that open list systems would have more incentives for targeted and pork-barrel politics relative to closed list systems. The experience of Brazil and Italy suggests this is the case⁴³⁴⁴⁴⁵ , but recent work questions whether open list systems always provide more pork at the expense of public goods. Kselman⁴⁶, for example, contends that open list systems encourage all politicians to put more effort towards public goods in order to increase their vote share and ensure their reelection. In closed list systems, politicians ranked higher in a party list can take advantage of the effort of lower ranked politicians to ensure their reelection, resulting in a lower level of public goods being delivered to constituents overall.

When Losers Win

That all votes should count equally is a fundamental principle of democracy and one that a majority of Americans embrace⁴⁷ The equal impact of votes ensures that the party or candidate who wins the greatest number of votes has a mandate that represents the majority. WTA elections, however, open the door to situations in which a party or candidate wins fewer votes than another and yet is still able to capture a greater share of seats. Under parliamentarism, where executive authority follows directly from control of the legislature, inversions can lead to a second-place party capturing control of government as well. If such electoral inversions occur often, they threaten the ability of elections to reflect voters’ expressed preferences accurately, and can create mismatches between government policy and public opinion.

Electoral inversions are unheard of in PR systems but not uncommon in WTA systems, where the geographical distribution of votes—not just their total number—drives results⁴⁸. The United Kingdom, for example—where members of parliament are elected from single seat districts—experienced inversions in 1951 and 1974. In New Zealand, the National Party won successive elections in 1978 and 1981 despite losing the nationwide popular vote to the Labour Party—results that fueled support for the adoption of a PR system in that country. Canada experienced an inversion in 2019, when the Conservative Party won 34% of the vote to 33% for the Liberals, yet the Liberals went on to form a (minority) government.

Inversions clearly violate the principle that all votes should count equally. That they are nearly impossible under PR is a democratic asset.

In the United States, electoral inversions at the congressional level have occurred twice since the 1990s—once in 1996 and again in 2012.⁴⁹ In 2012, Democrats exceeded the Republican popular vote total by 1 percentage point for the House of Representatives, but the GOP captured 54% of the seats⁵⁰. Electoral inversions are a more severe problem in the Electoral College and for the Senate because of malapportionment, but it is increasingly becoming an issue for the House because of residential sorting and gerrymandering.

Inversions clearly violate the principle that all votes should count equally. That they are nearly impossible under PR is a democratic asset. In WTA systems, they are unusual but not rare. With each election, there is a non-negligible chance that a majority is given to a party or candidate that lost the majority of the votes, flipping the most basic principle of democracy on its head. When inversions do occur, they can compromise the legitimacy of elections among voters and hinder the will of the majority, both of which can contribute to democratic instability. This is a risk baked into WTA and is among its greatest shortcomings.

Implications for the United States

The experience of countries around the world with PR provides little evidence that the policymaking process would be hijacked by fringe or small parties if the United States were to switch to a more permissive electoral system. That does not mean these extreme parties would not exist in proportional systems, but PR systems do provide certain safeguards that prevent extreme parties from taking over the system.

In a PR system, extreme parties can run independently and do not have to take over one of the two main parties to advance their goals, as has happened in the United States with the Republican Party. If an extreme party does win enough votes to win control of government, PR systems are better equipped to deal with them in power: The opposition is not entirely shut out of power, and multipartism can provide flexibility for new coalitions to form. The recent case of Poland illustrates the importance of having an electoral system that promotes coalition building across multiple parties for dealing with antidemocratic forces.⁵¹ There is, after all, a reason that would-be authoritarians elected in PR systems have tried to make the electoral system less proportional by creating more single-member districts, as Viktor Orbán did in Hungary.⁵²

The greater likelihood of having more than two viable parties in PR systems also gives voters options on the ballot. This may make it easier for voters to defect from a party engaging in antidemocratic or extreme behavior, something that is quite rare in the U.S. given the distance between the two parties and the lack of third-party options⁵³. A PR system would also mean parties will be responsive to a broader constituency—not just swing voters or voters in swing districts. Because each additional vote can help parties obtain a seat in the legislature, parties have incentives to prioritize a wider set of issues. A change to a PR system would dramatically change campaign dynamics, and it could allow historically neglected segments of the electorate to have a stronger voice, bringing policies more in line with the diversity of public opinion in the United States. We see these possibilities as substantial improvements from the current WTA system in the United States, in which so many voices are left unheard.

Enhancing Electoral Accountability

A commonly held concern among skeptics of proportional systems is that improvements in representation from having more representatives per district would come at the expense of electoral accountability, the ability to vote politicians out of office when they are underperforming. While all electoral systems have tradeoffs⁵⁴, the added representational benefits of PR do not come at the cost of accountability. There are forms of PR that could enhance accountability in the American context, primarily by reducing the incentives for gerrymandering; increasing electoral competition; and giving voters more realistic voting choices within their preferred parties and in other parties.

The fears that proportional representation would reduce accountability and responsiveness stem from the increase in the informational demands on citizens that more representatives would entail. For electoral accountability to work, voters need information about the performance of their representatives; distinguish between politicians; and assign credit and blame accordingly⁵⁵. This process is presumed to be easier in single-member districts, where having only one representative per district grants voters more clarity with respect to who is responsible for government policies. From this logic, scholars have posited that WTA elections should ultimately facilitate either rewarding or punishing the governing party⁵⁶⁵⁷⁵⁸ and individual politicians⁵⁹⁶⁰ at the ballot box.

However, successfully holding politicians accountable requires more than just information about politician performance. That information needs to factor into the vote choice of citizens, and citizens need to be able to vote for a politician other than the incumbent in order to punish them electorally. The conditions that allow for such accountability are becoming increasingly rare in the U.S. Thanks to partisan and demographic sorting and gerrymandering, very few of the 435 congressional districts are actually competitive.⁶¹⁶². Even where a viable challenger exists, partisan polarization makes voting for an opposition candidate to punish the incumbent unpalatable to more and more voters.

Although healthy levels of polarization help voters distinguish between parties and candidates, the extreme partisan polarization that characterizes American politics today makes it almost certain that voters give more weight to partisanship versus politician performance or integrity. While there is strong evidence that politicians at all levels are rewarded with more votes in good economic times and punished with fewer votes following economic downturn⁶³, these swings are rarely big enough to flip an election given the majoritarian, winner-take-all nature of American elections.⁶⁴

Contrary to common assumptions, proportional representation—particularly an open list system with moderate district magnitude—could facilitate holding American representatives accountable.

Even with the added cognitive demands that multiple representatives impose on voters, there are reasons to believe a PR system would enhance channels of accountability relative to a majoritarian one. Recent research affirms that both families of electoral rules are equally capable of fostering a clear link for voters between parties and responsibility for government decisions, particularly when parties are grouped into distinct policy camps and when district magnitude in PR systems is kept in the low to moderate range of between four to 10 seats per electoral district⁶⁵⁶⁶⁶⁷. Moreover, open list proportional systems that allow voters to indicate their preferences for a candidate can foster individual accountability, provided that the district size is sufficiently low such that the number of viable candidates remains cognitively manageable for voters⁶⁸⁶⁹. A low to moderate district magnitude also has the added benefit of improving representation while keeping the party system from becoming fragmented and government coalitions from becoming too complex—both conditions that can improve the clarity of responsibility among voters⁷⁰⁷¹.

Proportional representation can facilitate accountability in polarized contexts by providing voters with more choices within their party.

Proportional systems also have the potential of improving accountability relative to WTA systems, particularly in a context of high partisan polarization. An advantage of PR systems is that PR makes it more difficult to gerrymander districts in favor of one particular party because electoral districts are larger, both in geographic size and in the number of people who can win in each district. Districts that are less gerrymandered allow for more electoral competition, a crucial ingredient for accountability. With increased competition, politicians face stronger motivation to perform well in office for fears of getting voted out in the next election, and voters have more meaningful choices on the ballot to express their frustration with an incumbent.

Besides the opportunities for gerrymandering, another serious concern with WTA is that it fuses the party with the candidate, making electoral punishment less likely in a highly polarized context. To the extent that voters care more about partisanship than about candidate characteristics like integrity, they will continue voting for the incumbent. Rudolph and Daubler⁷² formalize this logic showing that even when voters learn about politician malfeasance, they will continue voting for corrupt politicians if voters place more value on their party holding onto office than on politicians’ integrity.

These dynamics were present following a 2009 expenses scandal in the United Kingdom, which uses WTA rules like the U.S., and where electoral accountability for the scandal was weakest in districts where partisanship was stronger in the electorate⁷³. An open list proportional system would decouple the party from the candidate and enable partisan voters to continue voting for their preferred party while voting for an alternative candidate within the party. In line with this logic, Rudolph and Daubler⁷⁴ find that politicians involved in a public scandal in Germany, where candidates can contest elections in both single-member districts and in open list PR, were more likely to be held accountable under open list PR than in single-member districts.

Party leaders can tame intraparty competition and promote party cohesion, keeping the system tractable for voters.

While open list proportional systems can decouple candidates from parties, allowing for more flexibility, there are concerns that an open list system would lead to fragmented party systems and a lack of party cohesion. In open lists, the conventional wisdom is that candidates have incentives to prioritize their reputation above that of their party and that parties lose their disciplining power that comes from the ability to demote candidates in closed lists⁷⁵. Fragmented party systems and politicians deviating from the party line can make the political landscape confusing for voters and muddle accountability.

Keeping the district size in the low to moderate range can help mitigate this outcome. But open list systems also have properties that keep the parties cohesive, and the party system from becoming unwieldy and unmanageable with too many candidates. Under open list PR, parties lose control over the ranking of candidate lists, an important tool for disciplining their members. Parties, however, are not rendered powerless: Carroll and Nalepa⁷⁶ argue that instead of using the threat of demotion in a party list, parties can promote party unity in open list systems by seeking and proposing policies that all their members can agree to. In doing so, party leaders achieve more policy agreement within their parties relative to a closed list system, in which demoting party members in the list provides a less costly disciplining mechanism but weaker incentives to foster party unity. Increased party cohesion helps voters hold politicians accountable by projecting clear party positions and allowing voters to distinguish among parties.

Open list systems also do not necessarily devolve into all-out intraparty candidate competition that would weaken parties as many fear. Cheibub and Sin⁷⁷ explain that party leaders have various tools at their disposal to control the level of competition within their parties. Party leaders can draft their candidate lists strategically in a way that avoids candidates competing for the same voters, and they can devote campaign resources to their preferred candidates so as to strengthen their position and weaken the threat of competition from other party candidates. Cheibub and Sin⁷⁸ provide evidence from the Brazilian context—a system seen as intensely competitive—that parties are able to regulate internal competition. Taming intraparty competition further facilitates electoral accountability, both by promoting more party cohesion and by keeping the number of viable candidates manageable enough for voters to monitor.

A Closer Look

District Magnitude – The Key Component of Proportional Representation

To achieve proportionality between the share of votes cast for a party and the seats it wins, PR systems allocate multiple seats to each electoral district. We refer to the number of seats as a district’s magnitude. Magnitude can vary from one to the total number of seats in the legislature. The US’s winner-take-all, single-member districts are an example of the former. At the other end of the spectrum are Israel and the Netherlands, which each use a single, nationwide district to allocate all seats. Most PR systems rely on magnitudes somewhere in between.

Magnitude affects how big a party must be to be viable. In a 10-seat district, a party that commands 10% support is viable, but in a 5-seat district the same party probably is not. With higher magnitudes, smaller parties – and more of them – win representation and seat shares track vote shares more closely. But there is a trade-off: affording representation to more parties can also complicate the process of building and sustaining legislative coalitions – of governing.

In contemplating the ideal magnitude, two considerations are critical. First, as magnitude increases above one to allow for proportionality, there are diminishing returns. That is, the increase in proportionality when moving from single-member districts to four-member districts is much greater than when moving from seven-member to ten-member districts, or from 21 to 24. Most of the benefits of proportionality can be achieved without moving to high-magnitude districts.

Second, as magnitude grows and the threshold for party viability shrinks, more parties will run lists and the number of candidates on each list rises because there are more seats in the district. As magnitude increases, then, the total number of candidates tends to increase multiplicatively, not linearly. And if the PR rules allow voters to indicate preferences for individual candidates, this can impose a cognitive burden on voters.

This combination of factors means there are advantages to low-to-moderate district magnitudes, in the range of 3 to 8.

Given the rich variation in how proportional representation systems can be designed and how these design features interact with a variety of contextual factors, scholarly work on accountability and electoral institutions is increasingly leaning away from a strict dichotomy between proportional and majoritarian systems. A lot of research is now focused on exploring a wider range of design features—such as federalism, list structures, term limits, and legislative committee structures—and how they interact with electoral rules in affecting processes of accountability⁷⁹⁸⁰⁸¹⁸². Federalism, for instance, may bring the government closer to voters compared to a centralized regime, but it also introduces various layers of government that can make it difficult for voters to assign responsibility. For instance, Mexico—which combines a federalism with a mixed-member majoritarian electoral system and where criminal violence is a salient issue—politicians are only held accountable for spikes in crime when there is partisan alignment across all levels of government⁸³. Term limits take away the ability to sanction individual politicians electorally from voters, and prevent parties from disciplining their members with incentives like reelections or long-term careers in the party⁸⁴. How these institutional design features interact with electoral rules is an area ripe for further research.

Implications for the United States

Achieving electoral accountability is increasingly difficult in the United States because of gerrymandering, partisan polarization, and the lack of viable alternatives on the ballot. Under the current system, most partisan voters will continue voting for their party regardless of their performance in office because the distance between the Democratic and Republican parties means the only competitive alternative is considered far worse. Even if politicians engage in antidemocratic behavior, many Americans are unwilling to sanction them if they belong to their party⁸⁵. Primary elections are one of the few spaces left for politicians to be held accountable. But in these polarized times, primaries can encourage politicians to engage in more polarizing behavior to avoid drawing a challenger from within their party⁸⁶.

Contrary to common assumptions that PR would weaken accountability, PR— particularly an open list system with moderate district magnitude—could facilitate holding American representatives accountable. This system would allow voters to continue voting for their preferred party while voting for a different candidate. PR would also encourage the creation of more parties, which would improve accountability through two main mechanisms: First, more parties could provide more realistic alternatives to voters, allowing them to defect from their original party choice if they are dissatisfied by their representatives’ performance. Second, multiple parties could increase electoral competition, which would keep politicians more disciplined in the first place.

Keeping the System Stable

Much of the literature on stability and electoral rules has focused on parliamentary systems and the stability of government coalitions. Government stability in this context refers to the longevity of a government and its vulnerability to votes of no confidence.

In parliamentary systems, a loss of majority support often means the removal of the prime minister from office. Research comparing stability across WTA and PR systems has generally found that governments in WTA tend to be more stable because they tend to generate prodigious winners’ bonuses for the largest party, thereby manufacturing single-party majorities even when no party wins a majority of votes⁸⁷⁸⁸. Delving deeper into the mechanics driving government stability, Taagepera and Sikk⁸⁹ show that government duration is a function of district and assembly size. As district magnitude and assembly size decrease, the share of seats held by the largest party increases, resulting in more stability. This generalizes the discussion away from the majoritarian versus proportional dichotomy and underscores how the design of the system matters: proportional systems with low district magnitude and moderate assembly size are more likely to have larger parties, and therefore experience levels of government stability close to those under majoritarian systems.

In presidential systems, the president is elected separately from the legislature, and these elections follow a fixed electoral timetable—so an executive can remain in office until the end of their term even if they have minority support in the legislature. Instability in presidential systems comes from the forceful removal of the president through military coups, which are becoming increasingly rare, or from impeachment procedures and conflict between the president and the legislature, which are becoming more common⁹⁰.

Scholars had long expressed concerns that presidential systems are prone to instability⁹¹ and that the combination of presidentialism with multiparty democracy is a “difficult combination” that threatens the stability of democracies⁹²⁹³. The fears were that a president would have a difficult time building and sustaining coalitions in a multiparty legislature and would be unsuccessful in passing through their policies, leading to gridlock and immobilism. Presidents, then, would have incentives to amass power and engage in unilateral policymaking, threatening democratic stability.

However, the conventional wisdom on presidentialism and multiparty systems, which are facilitated by proportional rules, is changing. (For a more detailed argument of how presidentialism and PR can be stable combinations, see Mainwaring and Drutman 2023.) Despite the greater difficulties in forming governing coalitions in multiparty systems, most presidents in multiparty democracies finish their term, and many countries with presidential systems and multiple parties have developed into stable democracies. To assess what a multiparty system would mean for American democracy, we should consider the conditions under which presidentialism operates with multiple parties. Comparisons between WTA and proportional rules in parliamentary regimes are informative—and the experience of Latin American countries, most of which combine PR with presidentialism, provides valuable insights.

Presidentialism and multiparty systems can promote stability.

It is more challenging to address questions about electoral rules and government stability in presidential systems because relatively few long-term democracies combine WTA with presidentialism. Beyond the United States, only Ghana, Liberia, and Sierra Leone feature this combination. (For more details, see Drutman and Mainwaring 2023; Mainwaring and Drutman 2023.) For a long time, the two-party presidential system in the U.S. was considered a paragon of stability⁹⁴ This stability, however, stemmed from the moderate nature of the two parties in the mid-20th century and the cross-cutting cleavages across the two. Conditions today are different: The parties no longer gravitate toward the middle; ticket-splitting is rare; and governing coalitions in Congress increasingly rely on razor-thin majorities. If stability characterized the U.S. party system in the 20th century, the historic ousting of the speaker of the house—the first time a speaker has been voted out— in October 2023, may presage a period of instability. Without comparable examples of systems that combine presidential and majoritarian rules, the relevant question then becomes under what conditions do multiparty presidential systems achieve stability and whether the U.S. could meet those conditions.

Since the 1980s, when most Latin American countries transitioned to democracy, various countries in the region have had presidents impeached and deposed, leading to periods of political instability and uncertainty. The vast majority of presidents, however, finish their terms even though the most common scenario in the region is a president who does not have a majority from a single party in the legislature⁹⁵⁹⁶. Countries like Costa Rica, Chile, and Uruguay have consolidated their democracies, despite the predicted risks of presidentialism and multiparty systems. This suggests that factors other than the electoral rules account for variation in stability in the region.

Noting that presidents in Latin America mired in crisis sometimes get removed and other times do not, Pérez-Liñán⁹⁷ contends that the removal of a president depends on the interaction of social movements and legislative majorities. Presidents are more likely to be removed if they face massive social protests calling for their resignation and have no support in the legislature, or when they face a legislature calling for resignation and there is no widespread social mobilization that would defend them⁹⁸.

Zooming away from impeachments and presidential resignations, Pereira and Melo⁹⁹ argue, based on their observations of Latin American cases, that the stability of multiparty presidential democracies depends on effective institutions of checks and balances, and on presidents who craft coalitions by having parties in the legislature that can be motivated to work together by cabinet positions, or by the exchange of goods and resources. In the Latin American countries Pereira and Melo studied, a president’s strength derives from clear constitutional authority that empowers them to lead negotiations with the legislature, build coalitions, and prevent gridlock. Importantly, the presence of multiple parties and the ensuing political competition help strengthen institutions of checks and balances. In the United States, the Constitution does not endow the president with extensive powers. Rather, the executive branch has grown in capacity and its purview has expanded through a series of executive actions, judicial decisions, and changing norms.

Proposals for a switch to PR are also often met with concern that PR could lead to democratic instability because of the potential for the proliferation of parties and the fragmentation of the party system. PR evokes contemporary cases like Brazil and Israel, where threats to democracy have happened in the context of a fragmented party system, or historical examples like the Chilean party system leading to the 1973 coup d’etat, or the Weimar Republic. However, a systematic analysis of 320 elections in 38 countries finds that party system fragmentation has no effect on democratic outcomes, except in highly polarized elections¹⁰⁰. Democratic instability may be more a function of other explanatory variables—like state capacity or economic shocks—than the number of political parties.

Despite fears that proportional representation would lead to an unwieldy multiparty system that would create legislative chaos and contribute to clashes with the executive, stability is possible in multiparty presidential democracies when facilitated by a presidential figure who can act as a dealmaker with the legislative body. The U.S. has a president with sufficient constitutional authority and resources to act as a central coalition builder; clear formal rules that have developed from long-standing legislative practices and judicial precedents, and institutions for horizontal accountability—but it is lacking multiparty competition.

Implications for the United States

There is no reason to conclude that a PR system would be destabilizing for the United States. PR could, in fact, provide forces for stability by increasing the legitimacy of democracy. Delivering policies in line with public opinion; preventing the overrepresentation of parties that do not win the majority of votes; and empowering voters to keep their politicians accountable can all prevent voters from souring on democracy, and supporting politicians who capitalize on that discontent and seek to destabilize the system.

Multiple parties could also promote policy stability. Currently, the policy distance between the Democratic and Republican parties on major issues means that when there is divided government, there is policy gridlock and that when there is unified government, policies swing to the extremes. Multiple parties could facilitate more consensus in the policymaking process—and while consensus building may take longer, it would not be worse than the alternation between gridlock and debt-ceiling gamesmanship that characterizes divided government under the current system.

The design of the PR system, however, is key for ensuring stability. One important consideration is keeping the number of parties at a reasonable level and avoiding excessive party proliferation. If the United States remains as polarized as it is today or becomes moreso, existing research suggests that party system fragmentation could pose a risk to democratic stability¹⁰¹. Certain institutional features of the United States would help guard against party proliferation, such as establishing multimember districts that elect only modest numbers of representatives (known among election scholars as moderate district magnitudes). Moderate magnitudes prevent very small parties from winning seats, effectively striking a balance between the more inclusive representation PR offers and requiring broad support to enter the legislature. U.S. federalism imposes a natural moderation on district magnitude in House elections, as 37 of 50 states have fewer than 10 House seats to begin with and the federal nature of the system precludes nationwide districts. Larger states could subdivide into districts with moderate magnitudes (i.e., five to seven seats), maintaining the obstacle to extremist and splinter parties.¹⁰²

Conclusion

Both WTA and PR have mechanisms that help voters achieve their preferred policies, hold their politicians accountable, and keep the political system stable. We argue that it is important to look beyond a simple dichotomy between the two types of systems and consider the conditions under which each system functions properly. Today, conditions in the United States are not conducive for a WTA system to deliver effective governance, accountability, or stability. Instead, the country’s WTA system is accentuating societal divisions, often producing outcomes inconsistent with the will of the majority, making it difficult for voters to punish politicians electorally, and increasing the risk of political and democratic instability.

A PR system is better suited to channel the diversity of views in the United States, and we do not foresee a proportional system undermining congruence, accountability, and stability. In contrast, we see reasons to be confident that PR would bring about improvements to these three areas by strengthening the position of median parties, making all votes count equally, and giving voters realistic alternatives on the ballot.

The virtues of PR come from the system facilitating the creation and survival of more than two political parties. WTA can also give rise to multiple parties when these are concentrated in certain regions, like in the United Kingdom, or when parties are less organized and do not provide strong incentives for politicians to remain in the party¹⁰³. An electoral system with multiple parties encourages more competition, which makes elections meaningful and strengthens accountability by giving voters more choices and encouraging politicians to be responsive to the electorate. The presence of multiple parties encourages compromise and consensus within legislative bodies.

Skeptics of PR see multiple parties as a liability of proportional systems because multiple parties can be confusing for voters and make negotiations in Congress difficult. Multipartism is sometimes portrayed as synonymous with instability and chaos. We acknowledge that too many parties can often, but not always, be destabilizing. But the choice is not a binary one between the two-party system that WTA encourages in the U.S. or the fragmentation a very permissive PR system encourages. PR systems can be designed to ensure a healthy balance between too few and too many parties. This should encourage thinking about electoral system reform beyond a PR versus WTA dichotomy and consider the many other institutional design features that interact with electoral rules in ways that can affect outcomes we care about.

We do not know what the party system would look like in the U.S. with a shift to PR or what parties would prove successful in a more permissive party system. Preferences are not stable; coalitions are malleable; and factions can come and go. Particularly because PR would activate different societal cleavages if political entrepreneurs deem it advantageous, it is difficult to predict the party landscape. What we do know is that a PR system would be more amenable to new coalitions without sacrificing governance. We also know the status quo is unsustainable. The WTA system in the U.S. is no longer able to ensure good governance and is making the country increasingly vulnerable to authoritarian threats.

Policy outcomes, accountability, and stability under different electoral systems

Introduction

Key Takeaways

How proportional representation affects governance, accountability, and stability

Proportional electoral systems are better at achieving what voters want.

Proportional electoral systems can enhance government accountability.

Proportional electoral systems can reduce political instability.

Balance is possible between two parties and too many parties.

A Closer Look

Presidentialism and the Senate

Achieving What People Want

Both winner-take-all and proportional representation systems are capable of delivering policy outcomes in line with public opinion.

In polarized contexts, proportional representation systems are better suited at achieving congruence than winner-take-all.

The choice of open or closed lists in a proportional representation system affects the responsiveness of representativeness and the types of policies they pursue.

A Closer Look

Transferable Vote Systems and Proportional Representation

When does proportional representation make a difference?

Proportional representation systems favor the median voter and programmatic policies, resulting in higher levels of progressive redistribution.

When Losers Win

Implications for the United States

Enhancing Electoral Accountability

Proportional representation can facilitate accountability in polarized contexts by providing voters with more choices within their party.

Party leaders can tame intraparty competition and promote party cohesion, keeping the system tractable for voters.

A Closer Look

District Magnitude – The Key Component of Proportional Representation

Implications for the United States

Keeping the System Stable

Presidentialism and multiparty systems can promote stability.

Implications for the United States

Conclusion

About the Authors

John M. Carey

Department of Government, Dartmouth College

Oscar Pocasangre

Political Reform Program, New America

Related Content

Why do we need political parties?

Checking the Pardon Power: Areas of Abuse

Checking the Pardon Power: Preventing & Responding to Abuse

Checking the Pardon Power

It can happen here. We can stop it.

It can happen here.
We can stop it.