The Case for Proportional Voting

Lee Drutman

Winter 2018

America's two-party system is an increasingly tough place to be a conservative. The Republican Party under President Donald Trump has come unmoored from classic conservative values and shows no interest in returning to them. The Democratic Party is certainly no place for a conservative, with the Bernie Sanders wing of the party increasingly ascendant.

If only there were another option — a party of principled conservatism that might help move our politics in its direction. In the American system of plurality-winner elections, held across 435 congressional districts and 50 states, such a party would likely accomplish exactly one thing: deliver Democrats wide majorities in Congress. Thanks to our electoral laws, third parties face almost hopeless odds, leaving our political system without an obvious mechanism to diffuse the zero-sum partisan warfare that is consuming our politics.

This would not be true under a proportional representation system. Such a system would allow multiple parties to truly compete to win seats in the legislature, and their representation would be in proportion to their electoral support. In this system, a new conservative party would likely win about 15% of the seats in the legislature — and without boosting Democrats' fortunes. Right-of-center voters who wanted to vote for a non-Trump conservative party would have that option, and their representatives could bargain and negotiate as an independent force in what would be a multi-party legislature.

By passing electoral reforms to open up the party system, conservative lawmakers now living in fear of right-wing primary challengers could carve out a new role for themselves. Instead of being forced out one by one by the powerful and angry Republican base that our totalizing partisan conflict has created, they could find an opportunity to lead a pivotal center-right party. Such reforms would restore representation to some disenfranchised Republicans and many other Americans whose political views are widely shared yet not dominant enough to command one of our two major parties. Most important, it would help to restore all citizens' faith in the democratic process.

Fundamental electoral reform may seem improbable, but it's not as unlikely as it first appears. During the first half of the 20th century, the vast majority of Western democracies changed their electoral systems from single-member majority- or plurality-winner elections like ours to proportional systems. About four in five of the world's democracies now have a proportional representation system with multiple parties, and they have benefited overwhelmingly from the change. Proportional voting democracies with multi-party systems consistently have more centrist governments, more political stability, higher voter turnout, and citizens who are more satisfied with democracy. It's also what Americans want: In 2017, Gallup found that a record-high 61% of respondents said they wanted a third-party option.

A new, more representative system is possible, and, crucially, it could be achieved within the bounds of our constitutional system. In an era of pernicious politics and dwindling faith in the institutions of democracy, such reforms should not seem so much radical as necessary.


Two pressing fears of democratic instability loomed in the mind of James Madison as he set out to design a representative system of government: the problem of majority tyranny and the problem of faction.

Madison's studies of smaller confederacies led him to conclude that when a dominant majority could form, it would. It would then use its power to punish the minority, ultimately creating political instability. Therefore, he wanted to spread power around, balancing ambition and making it hard for any permanent majority to take shape.

Faction presented a different problem. If majority tyranny was a problem of too much concentrated power, factional warfare was a problem of too much diffusion of power. Factions would pull and tug at unity, tearing the republic apart. The problem here was even more devilish. Power could be shared and diffused. But to limit faction was to limit liberty, especially freedom of expression. Factions needed to be able to make their cases, so political leaders could then deliberate over their claims and find compromises. (Madison hopefully described these leaders as "a chosen body of citizens, whose wisdom may best discern the true interest of their country, and whose patriotism and love of justice will be least likely to sacrifice it to temporary or partial considerations.") Madison's hope was that in an "extended republic" there would be enough competing factions that none would be permanently dominant, and instead public-spirited lawmakers would balance out their interests.

Madison and his compatriots hoped to avoid political parties, which they feared might become permanent factions that would tear the country apart. Without attachment to parties, he and his contemporaries hoped that whatever coalitions existed would be fluid, with no permanent majorities, and the process would be consensus-oriented. Avoiding political parties was not an unreasonable expectation at the time. The first Congress was made up of 26 senators and 65 representatives, a small enough group that everybody could get to know each other well, and potentially operate in a consensus-oriented way.

Yet, despite the best hopes of the founders, parties quickly emerged. American politics organized around the competing Hamiltonian Federalist and Jeffersonian Democratic-Republican parties, and the election of 1800 proved to be one of the nastiest in American history. Partisan politics had arrived, and it never left.

Though the founders' fantasy of nonpartisan consensus still has a unique hold on American political thought, no major modern democracy has ever functioned without parties. The basic problem is this: At a large scale, it is very difficult for voters to judge the character and quality of candidates, and to figure out what exactly they stand for. (This would be true even if Madison had succeeded in his goal of capping House districts at 30,000 constituents.) Parties ideally function as intermediaries; they are supposed to vet candidates for quality in order to simplify political choices for voters, who have more important things to do than follow politics closely. Voters need shortcuts to figure out who shares their values and interests. Parties emerged to solve these problems. In doing so, they make electoral choices meaningful for voters, and make governing a little more stable by creating more regular voting coalitions.

While parties certainly serve important functions in a large democracy like ours, it's easy to see why Madison was skeptical of them given the political fracturing that has occurred in our own country. The problem may not be so much the existence of parties, but the fact that we have only two.


For the first 70 years or so of American democracy, the American party system was still somewhat fluid. Many states used multi-member districts, meaning that comparatively large congressional districts were served by more than one representative. A few major parties came and went, and third parties generally played a more important role, most notably, the Republican Party in the fateful election of 1860, which precipitated the Civil War. But since the election of 1860 (with the exception of Theodore Roosevelt's 1912 run on the Bull Moose ticket), Republicans and Democrats have been locked in a two-party dance for control of American democracy, sustained by a winner-take-all electoral system that makes defection very costly.

By contrast, across the Atlantic, two-party democracy never quite took hold in the same way. In 1899, Belgium became the first country to adopt a proportional voting system. The system was created to ensure that all significant political groups were represented in the legislature in proportion to their support among the electorate. Over the next several decades, proportional voting systems replaced plurality systems across Western Europe. Though each country had a slightly different path to proportional voting, two broad generalizations can help explain how it came to be adopted so broadly.

First, a growing consensus emerged around the fundamental fairness of proportional voting. Whereas single-member majority or plurality winner-take-all approaches to elections resulted in a lot of wasted votes and produced highly disproportionate translations of votes into seats (frequently depending on quirks of geography), a proportional system counted all votes as roughly equal and translated votes to seats much more smoothly.

Second, proportional voting gave both parties and politicians more predictability, since results from single-member districts could vary widely depending on narrow swings, especially when there were more than two parties competing (as there often were). In some countries, the political right was fractured and thus feared that plurality voting would aid a rising socialist party. Passing proportional voting rules ensured that, even if the political right was split among multiple parties, voters who wanted right-leaning parties representing them would get that representation without having to agree on a single party line.

The case for proportional voting was considerably weakened after Adolf Hitler came to power in Germany, and for a while, the standard argument against proportional voting was that it made it easier for extremist parties like the Nazis to gain a foothold in the legislature. Thus mid-20th-century American democracy, with its moderate two-party politics, looked like a model of stability. In retrospect, however, that stability depended on a series of conditions that were particular to the time. It was in fact thanks more to a period of post-war affluence and the population having broadly shared values and experiences — plus a common enemy in the Soviets — than to any virtue inherent in a two-party system.

Defenders of the two-party system have long claimed it possesses two major advantages in particular. The primary long-standing argument was that, because both parties were theoretically competing for the median voter, both parties should hew to the center. This theory is commonly called the "median-voter theorem." No party, it was theorized, could hope to win national elections without a broad appeal and a big-tent coalition. In a multi-party system, by contrast, different parties carve out different niches, creating a much broader ideological range of parties, and potentially introducing more extremism into politics.

A second long-standing case for a two-party system was that it could provide clearer choices to voters, who could choose one of two teams to run the government. Then the winning team would have both a mandate and clear responsibility, and if voters were unhappy by the next election, they could hold the party in power accountable by voting them out. By contrast, when citizens vote in a multi-party election, they are not selecting between two competing visions for the country, but rather among many visions, with no clear understanding of what role those smaller parties might play in a governing or opposition coalition. In theory, this muddles accountability, since it's hard to hold a single party responsible in a governing coalition.

Both of these arguments have been undermined by a lack of corresponding evidence. The accountability argument works only if a significant number of voters are willing to switch parties based on performance. This assumes that they are otherwise indifferent between the two parties. But when 90% of partisans will blindly support their own party no matter what happens (as they do now), the whole concept of "accountability" collapses. Moreover, this logic always worked better for the U.K., where the parliamentary system empowers even narrow majorities, than it did for the U.S., where the more complicated system of checks and balances and a filibuster makes majority-party control harder.

The logic of accountability also leads to over-promising. In a multi-party system, parties can only promise to advocate particular positions, since governing by coalition is more complicated. In a two-party system, parties have to promise to deliver better outcomes, which leads to a politics of over-promising and disappointment. Over decades, this has devolved into deep distrust of politicians (who never seem to deliver on these promises) and of political institutions. The growing distrust has led to ever more radical politics. It has created a self-cannibalizing energy on both sides of the political spectrum, though its influence is most clear in our time in the burn-it-all-down rhetoric of the populist wing of the Republican Party.

The median-voter theorem also has aged poorly. The two parties have not converged in the middle. They've pulled to their extremes. Part of the reason is that, with 435 separate House electorates, 50 different Senate electorates, and 50 different presidential electorates, there is no single median. In most states, the median voter is either considerably to the left or to the right of the national middle. (In a sense, the very thing that reduces competition to two parties — the single-member plurality-winner district election — also makes the median-voter theorem a terrible guide to understanding national elections.) Add in primary electorates, which loom especially large in the presidential election, and the electorate is pulled even closer to the margins.


And yet, many would argue, American politics once functioned quite well as a two-party system, with Democrats and Republicans working out plenty of historic bipartisan compromises to accomplish landmark legislation — particularly in the mid- to late-20th century. What's wrong with the two-party system that can't be restored by recovering the lost art of political compromise?

Such nostalgic arguments are quite common among elder statesmen in Washington, the kind who attend panels and write op-eds about the need for "political courage" and "regular order." What they fail to understand is that the bipartisanship of yore was not just a matter of political character, but a matter of political incentives, party organization, and genuine common ground.

Bipartisanship flourished because voting coalitions split parties on an issue-by-issue basis. Liberal (Northern) Republicans and liberal (Northern) Democrats had many positions in common, as did conservative (Southern) Democrats and conservative (Western) Republicans. There were few permanent enemies and few permanent allies. Both parties also held a broadly shared consensus on American values, largely united against a shared enemy: the evil empire of the Soviet Union.

In retrospect, the pre-1980s era of American politics was not really a two-party system at all, but instead a four-party system within the broader container of a two-party system. Both party coalitions held together liberals and conservatives, who operated as independent factions within the parties. As a result, both parties looked modestly centrist as a whole, and could compete everywhere because their brands were capacious enough to take on different forms depending on local values.

Because partisan identities were less distinct, and complicated by other, cross-cutting regional and ethnic identities, politics lacked the militaristic us-versus-them dynamic it has now fallen into. It was perfectly reasonable for Democrats to sometimes vote Republican, or Republicans to sometimes vote Democrat, because they liked a particular candidate or liked the idea of parties checking each other. Though cross-partisan presidential-support scores have fallen into the single digits and split-ticket voting is a rare phenomenon, such things were common during the mid- to late-20th century.

This bipartisanship began to unravel as the parties realigned in the 1980s and 1990s. New cultural fissures that had emerged out of the '60s and '70s reshaped the dividing lines of American politics. And in an era of growing affluence, post-materialist "values voting" replaced pocketbook voting for many voters, and battles over abortion, religion, and social justice took center stage.

As a result, the culturally conservative South moved from solidly Democratic territory to predominantly Republican territory, turning the Republican Party into a much more culturally conservative coalition. Meanwhile, as Democrats gained dominance on the coasts and in the big cities and lost their Southern conservative "Blue Dogs," the Democrats became much more uniformly culturally liberal. Liberal Republicans and conservative Democrats essentially went extinct. The overlapping ideologies and loyalties that used to cross-cut the parties realigned along party lines. Politics became regionalized, and without any cross-cutting dimension, partisanship became totalizing.

As this process continued, the Democratic stranglehold on Congress loosened, which further polarized our politics. Since 1980, and especially since the mid-1990s, almost every national legislative election has put majorities in both houses up for grabs, and this tight partisan competition has transformed American politics into trench warfare. Political scientist Frances Lee has looked closely at the devastating consequences of this intense national partisan competition, most recently in her masterful book, Insecure Majorities. The conclusion of her research is dispiriting: The more closely contested the control of institutions, the more politics devolves into zero-sum partisanship with all its dysfunctional consequences. Every vote becomes a party-line vote. Party leaders raise the stakes of every potential dispute to draw clear contrasts for activists, donors, and voters. Each party denies the other any small victory that might be useful in the next election, and looks primarily to embarrass the other side rather than to collaborate. The permanent campaign takes over.

Both parties have their internal divisions, of course. But the key difference is that now parties are divided in a non-overlapping way, the result of decades of sorting and intense, team-driven leadership focused on winning close national elections. And when there is no overlap, there is no template for policy- or issue-oriented compromise — just winning and losing.

This is the logic that led to Trump's policy-agnostic version of politics, in which everything is cast in terms of winning and losing, and in which Trump's supporters thrill to the president's words and actions primarily to the extent that they upset and anger the hated liberals and "political establishment." Democrats are now in danger of mirroring this logic.

This dynamic creates a devilish paradox for two-party politics. On the one hand, considerable democratic theory argues that partisan competition is essential to democracy, since we depend on the "out" party to hold the "in" party accountable. But Lee's research makes a convincing case that close two-party competition makes things worse.

Two-party politics could work at a time of weak partisanship, overlapping coalitions across multiple issue dimensions, broadly national but weakly led parties, and a sizable number of citizens being willing to go back and forth between the two parties to hold them accountable for their performance (or possibly even their extremism). This is no longer the case. We now have totalizing partisan conflict and unstable two-party politics.


Since World War II, hundreds of political scientists have gathered and analyzed extensive data on cross-national democratic performance. The almost universal verdict has been that, on almost every metric, proportional voting performs better than plurality-winner, single-member-district systems. Especially in recent decades, the evidence is now solidly on the side of better outcomes in proportional voting systems.

Countries with proportional voting have shown three notable advantages over our two-party plurality-winner system: Their governments do a better job of representing the median voter, and politics is generally more stable; voting rates are higher, and support for democracy is higher; and it is easier to marginalize extremism. It is worth examining how each of these works in turn.

First, countries with proportional voting systems have more stable, more representative governments. Political scientist G. Bingham Powell, Jr., has spent a career comparing political systems and demonstrating through exhaustive evidence that proportional voting, multi-party systems consistently produce governments that are closer to the median voter than single-winner-district, two-party systems. The basic logic is pretty straightforward: Because two parties rarely converge in the middle in practice, winning governments in two-party systems are likely to swing back and forth — too far to the left, then too far to the right, like a thermostat that can only be set to sweltering or freezing.

In a multi-party system, by contrast, there are far more settings for voters to choose. Continuing the thermostat metaphor, some voters might prefer it to be very cold, while others might prefer it very hot. However, there is usually a pivotal group of voters who prefer the middle. In a multi-party system, it is possible for them to register that preference and elect a representative party. This is why multi-party elections tend to produce governing coalitions that include the party or parties in the middle. As a result, governing coalitions tend to stay oriented toward the center in proportional voting systems, and policy tends to stay more stable.

In short, by giving diverse voters a broader range of settings on the policy dial, this system makes it easier to collectively agree on what the midpoint should be, and the midpoint remains somewhat stable as a result. When there are only two extreme choices, it becomes much harder to agree on a midpoint, leading to more wild swings back and forth depending on who wins control.

Second, proportional voting systems have higher voter-participation rates and higher political efficacy. Among advanced democracies, voting rates in the U.S. are relatively low — only 55.7% of voting-age citizens cast votes in the 2016 election. Voting rates are consistently higher in proportional voting systems for two main reasons.

The first reason is that, in a proportional voting system, all votes matter — not just those in swing states or districts. In the U.S., parties focus almost all of their attention on a limited number of swing districts and swing states, hoping to boost turnout for their side where it can make a difference. The rest of the electorate is mostly ignored because its behavior is easily predicted and unlikely to make a difference. And voters who know their votes are unlikely to matter are less motivated to make the effort to vote. In a proportional voting system, there are no swing districts or states. As a result, parties have more incentive to boost turnout everywhere, and voters are motivated to go to the polls knowing their votes actually matter.

The second reason is that, because a proportional system produces more parties, voters have more choices and are therefore more likely to find a party that they feel represents them. Certainly, there are limits to this — too many parties can lead to a paradox-of-choice problem, where voters feel overwhelmed. (Too many parties can also lead to legislative dysfunction.) But in general, a proportional system with a decent threshold will limit parties to a reasonable number (ideally four to six), giving voters adequate but not overwhelming choice, and will make every vote meaningful and therefore worth mobilizing for. As a result, more citizens feel represented, because they are represented.

It follows, then, that support for democracy is also higher in multi-party systems. Recent work on "democratic deconsolidation" by political scientists Roberto Foa and Yascha Mounk has demonstrated that younger generations are far more likely than their elders to say in surveys that "having a democratic political system" is a bad way of governing, creating real concern that support for democracy is weakening. However, when political scientist Pippa Norris compared such surveys across 24 Western democracies, she found that "a statistically significant fall in democratic approval by birth cohort" is largely (though not exclusively) limited to Anglo-American countries — the U.S., the U.K., Canada, Australia, and New Zealand — in which two parties dominate. In the rest of the countries, which all have multi-party systems and some form of proportional voting, there is generally no decline.

Given the deficiencies of two-party democracy, it's not surprising that young people are turning away from democracy primarily in two-party systems. Because of the representational and mobilization deficiencies of two-party systems, a large number of citizens, especially younger ones, are likely to feel politically disconnected and disengaged. And the more politically disengaged citizens are, the less support they evince for their political system. Given the increasingly extremist anti-establishment resentment that powered Donald Trump to the presidency, this is not a merely academic argument.

Moreover, because losing parties can be shut out of power entirely in a two-party system, there is considerable evidence that losers tend to seethe much more in majoritarian as opposed to proportional systems. Because proportional systems promote coalition government and power-sharing, even minority parties often get some role in governing, making their supporters feel more consistently invested in the political process. Majoritarian two-party systems, by contrast, instead produce losing partisans who rail aggressively against the system as fundamentally broken and corrupt — as Republicans did after the election of Obama, and as Democrats are doing now — because they are entirely shut out of power and desperately want to get it back. The anger and resentment now consuming both parties are the products of a mentality that sees politics as scorched-earth warfare.

The third advantage of a multi-party system is its handling of extremism. Certainly, extremist parties are on the rise across almost all Western democracies. And again, one standard argument against multi-party systems has long been that they make it easier for extremist anti-system parties to gain a foothold they otherwise would not have. But there are two other factors to consider. First, while a proportional representation system makes it more likely for a populist party to gain some support, it can also provide an outlet for discontent. Political scientist Powell, for example, has made a convincing case that, in divided societies, this outlet actually reduces the likelihood of large-scale political violence. In a multi-party system, there are plenty of ways for other parties to form coalition governments that marginalize extremist parties.

In a two-party system, an extremist movement is more likely to concentrate its energy on taking over one of the two major parties, since everything else is meaningless. This is a much tougher fight, but the prize is far more consequential. Consider that only 43% of registered voters are Republicans, and, in 2016, less than half of registered Republicans actually made it to the polls. While Trump won the support of the vast majority of those voters (over 90%), they made up less than a fifth of the total electorate. That's about the same level of support populist parties have been winning in some European countries. The extremist Jobbik party, for example, won approximately 20% of the popular vote in Hungary's 2014 election.

The main difference is that by winning the Republican Party's nomination, Trump could tap into the power of partisan loyalty. Many Republican voters assumed that if he was the Republican Party's nominee, he couldn't be that bad. And besides, what were they going to do? Vote for the Democrat? Or stay home and let Hillary Clinton win? And so, just as in every election since 2000, 90% of self-identified Republicans voted for the Republican, and close to 90% of Democrats voted for the Democrat — as they would have done seemingly regardless of whom the parties had nominated.


What would be involved in advancing this kind of change in our politics? Full proportional representation would require a constitutional amendment. But because the Constitution does not specify how states should choose the members they send to the House of Representatives, significant steps in this direction could be taken within the existing constitutional framework. For instance, the Fair Representation Act, sponsored in this Congress by Representative Donald Beyer of Virginia, would shift House elections from single-member to larger, multi-member districts with ranked-choice voting. This would create districts of up to five members, with the top five vote-getters in each district going to Washington.

In practice, this would mean that on Election Day, voters in a five-winner district would see on their ballots a few different Democrats, a few different Republicans, and a few candidates from other parties. They would then rank the candidates in order of preference. If their preferred choice did poorly, their votes would get redistributed to their second choice, and the overall tally would update, eliminating candidates from the bottom up and redistributing their votes until just five candidates remained.

This would mean opportunities for third and fourth parties, and considerably more diversity among both congressional Democrats and congressional Republicans, since Oklahoma Democrats and New York City Republicans would be able to elect representatives despite not being a majority in every single district. This would cut against the geographical partisan sorting of politics, and take away the winner-take-all two-party competition by making it less likely that any one party could achieve a congressional majority.

And for voters, it would mean that every vote really would matter, because there would be almost no wasted votes. Voters could register their sincere preferences, rather than having to hold their noses and choose the lesser of two evils. Voters will also be more likely to have someone who actually represents their values in Congress.

Changing the nation's electoral system would be a gamble, but not pursuing change may be even riskier. Certainly, one can hope that the fever will break, that Trump's takeover of the Republican Party will lose steam, that the Democrats will moderate, that a new center will emerge. But our electoral laws make these possibilities very unlikely. Changing our electoral laws may be our best hope for finding a new center.

Lee Drutman is a senior fellow in the Political Reform program at New America. 


from the


A weekly newsletter with free essays from past issues of National Affairs and The Public Interest that shed light on the week's pressing issues.


to your National Affairs subscriber account.

Already a subscriber? Activate your account.


Unlimited access to intelligent essays on the nation’s affairs.

Subscribe to National Affairs.